To make sure it's clear to people watching this video Carlsen was not on a 53 game win streak. He was on a 53 game no loss streak. The record at high level chess for consecutive wins is 20 or 21 unless someone broke that record.
That idea that he sacrifices the queen like a bot might do reminds me of a time I sat in with the chess club in high school, and there was this puzzle about how to achieve checkmate from some seemingly hopeless position, which did happen in some big game. Everyone was throwing out ideas and the instructor kept shooting them down, and eventually it went quiet. I just said, "sacrifice the queen." The guy next to me was like, "NOOOOO" and the instructor was like, "yes", and walked us through how this sacrifice offset one other piece by a single square, opening up some diagonal move allowing mate to happen after a bunch of other moves. I sat there looking like this galaxy-brained genius who'd figured out this thing nobody else could, but I only said it because it was the only idea left on the table after everyone else had already said their thing, and it seemed like a taboo. I had no idea how to actually do it. I played against the guy who yelled "NOOOOO" one time. He checkmated me in like a couple of moves then said I was a good player. Thanks, man.
I've coded a few games, decades ago. You do see computers making these "unusual" moves. When it happens with games you've coded, you can think it's messed up, but then things suddenly turn around many moves later (exactly like one of the games in the AlphaGo tournament). When I've analysed them I've found that the only reason that these moves make sense is if you can do a deep enough analysis to determine that there is a guaranteed potential technical advantage, which means they cannot be successfully replicated by humans without the deep search capability of a computer. I've also noticed that you often never observe from subsequent play just what advantage the computer originally saw when it made the move. The reason is that computers don't always make the compensating payoff move as soon as they can, they just move to _keep that option open._ They often keep rolling forward that option without exercising it, until the human makes a mistake that creates a better option for the computer (or the computer just sees a better option), and it then moves to roll that option forward, and so on. That makes it harder to learn just by playing the computer, because there may be a lot of moves between sacrifice and realised payoff, and the observed payoff may not be the one the computer originally saw when it made the sacrifice. You can make the computer show its current best line of play, and I assume that is what these players do when they use the engines to practice, but humans are always up against their limited search depth compared with computers if they try to replicate that except in the exact same position.
That DOES compute!....terrific answer from a suspected Trolling chess 'bot!...😁....trying to gain a psychological advantage by convincing future human opponents... Remember ..you're beaten BEFOREHAND.... "Nice try carbon based skinsuit".....DIABOLICAL chessbot!👺
As a beginner, there have definitely been times where I played what was evidently a mistake and the analysis tool was like "wow, yes, that's what I would've done too! You're so smart :)" because it saw a sequence of winning moves that I couldn't have predicted - and indeed, because I _didn't_ predict it, my follow up move is guaranteed to get a "...aaaand you're losing now" from the engine. Always makes me laugh when that happens 😅
@@SloverOfTeuth in modern chess computers are a huge component of preparation. A2/H2 are super common from computers and having a chess brain you can see these moves then begin to see situations when these games are made despite not being able to see so far into the future. Also, the randomness it injects into the game hurts prepared players. Magnus himself says he plays a young player strategy where he plays suboptimal lines to throw opponents out of their preparation.
I like her nerdy sass. "If your child was so smart then why doesnt he just spend 5 mins to do his homework and then go sleep." I feel like she been tempted say this to people directly. lol
Note: The "first move advantage" thing is not purely psychological, because when chess engines play chess engines, white wins more often than black (of the 0.01% of games that *aren't* draws, which, fair enough.)
To be clear, a lot more than 0.01% of games are wins for one computer or the other. You're off by several orders of magnitude in fact, in the 2020 edition of the Computer Chess Championship for instance there were 26 decisive games out of 200.
The number of draws depends on how powerful the hardware is and how much time is allowed to "think" per move. The longer the time between moves, the more likely a draw is. So if you want "interesting" games between computers that don't end in a draw all you have to do is reduce the time available until they stop drawing their games. I couldn't find the rules for the Computer Chess Championship but I'm pretty sure it's set in such a way so as to minimize the number of draws instead of setting it up so as to get the "best" possible games.
@@oscarprieto9013 The Computer Chess Championships are a bit of an outlier for our purposes because in those, the engines are intentionally given specific openings to play, which artificially reduces the number of draws far below what happens when chess engines are freestyling. Anyway, the goal here isn't to find interesting games. The goal here is to determine whether white having a higher winrate than black is caused by human psychology vs. by white being objectively (although by only a tiny amount, to be fair) stronger than black positionally.
@@tudornaconecinii3609 The advantage of playing first is actually very substantial relative to the small differences between top players. If you look up common openings like the Ruy Lopez or Queen's Gambit on Wikipedia the diagrams should make plain why this is to anyone who knows the moves: basically the early initiative the player of the white pieces possesses can be converted to enduring structural advantages. Win rates for the first mover are much greater at high levels and significantly greater even at much more modest skill levels.
That anecdote about playing checkers in the office was infuriating, I lived through so many similar situations myself. Physics students can be quite something sometimes.
@@acmhfmggru The difference is that in tic tac toe you can just know by heart what to do in each situation. In checkers you can't. You would still need a computer. For humans it will still be an interesting game. It's like it if said in the news that someone finally solved chess and has an AI that will always win (or hold a draw or whatever). That wouldn't mean that humans playing chess with other humans suddenly becomes pointless.
@@luxshokk Some of the top chess players often play checkers casually in between games. If the top chess players with their extremely good calculation abilities still find checkers difficult enough to play and not trivial to draw, then I doubt random "I don't play checkers" guy (and "she's just bitter" over here) can easily win.
@@luxshokk Finding a game unapealing =/= being a condescending prick. Elon comment was spot on, I prefered turn based and real time strategy games over chess since forever for many reasons.
I dream of a day where people stop having this exact conversation: 1) "Chess has more positions than ___." 2) "Well that's only if you count wrong." 3) "This other game has more." Like, come on. Please stop doing this.
My understanding of moves "looking computer-y" is that there's a difference in risk/reward calculation between players and computers. Computers will regularly make moves that humans, even very good ones, would probably consider either too dangerous or too slow. A strong human player can see a "risky" move which could (say) sacrifice material to generate a strong attack, but they might be unable to confidently calculate forward enough to know for sure that it will work. A computer would be able to see the same move and know its not risky at all, and go for it. Conversely computers will often do things that just... *look* like a pointless waste of time, but actually aren't? A subtle slight repositioning of a piece that doesnt obviously generate attacking potential or neutralize danger, but has some strong importance far down the road. Something a human, even a very strong one, would probably consider a waste of time, failing to see the long term value that move could provide. At least this is what ive learned from watching chess streamers -- im not very good at chess myself.
Actually, sacrificing material for an advqntage can be a very human thing to do. _Taking_ material and then defending perfectly for 20 moves in a passive position until they can untangle is the kind of thing that humans try to avoid and computers love to do.
It's true and a lot of it lies in the justification for the moves. See, if a player was to completely change what they're doing after every single move made on the board, then they're going to find themselves trying to do an insane number of calculations. Meaning that they're very likely to make mistakes or run out of time, so you're better off playing with (or forming during the game) a solid plan on how to win and trying to make it work as best you can (all the while, trying to figure out what your opponent is plotting to do against you). This is why even grandmasters constantly study and practice various positions, so that they already know what the plan is should they find themselves in that very situation during a live match. (When you see them taking 30 minutes to make a move, that's not a great sign; you know their plan has gone awry and they're now having to improvise). When spectating games (and especially when watching them in hindsight) you can often find the reasoning behind their moves when you consider how these moves effect those plans of theirs, as they're going to be presented with a lot of questions from their opponents during the game. "Do I want to trade these pieces?", "he's attacking my knight, how am I going to react?" etc. You'll find time and time again that players will often try to make the moves that will best aid their plans, whilst avoiding things that are counterproductive to them. Chess bots, of course, do not work like that at all. They will happily recalculate the position after every single move that they and their opponent make, completely changing what they're doing on the fly. They'll make the best move, simply because it's the best move. They don't care if a move is gonna trash their plans. They don't even have plans. What they're doing is making a lot of calculations to figure out every possible combination of moves, 20 steps ahead or more, so that it can simply make whichever one brings them closer to winning whilst causing the least damage to their position. This lack of adherent to any sort of plan leads to some very bizarre-looking behaviour. As an example, a real player typically wouldn't set themselves up for a flawless attack and then as everything is going to plan, scrap the whole thing at the last minute, and push a random pawn for what is (to us) seemingly no reason, because a completely different way to win in 10 moves, rather than the 11 moves your plan would have taken, has presented itself. Real humans would just go ahead with their already calculated plan of attack, it doesn't matter if it's not the absolute optimal, 100% efficient way to win, so long as you're comfortable that your plan is going to work. You're not even going to waste the time trying to figure that out. In the human's mind: The clock is ticking, and I'm winning, so let's go!
One of the key factors in Niemann's cheating was just how unreasonably perfect his play was. From recollection he made less than a third of the mistakes of the world's best players, yet kept that ridiculously high standard through every single game no matter how exhausted. It was particularly on show when it mattered, and particularly obvious in unusual or difficult situations. People just can't play every move absolutely perfectly like that, computers very much can. And that's a very key factor in play looking computery. As well as the more typical (especially to lower players) awkward moves. They will switch rapidly between playstyles and do moves no player would ever dream of because the learned experience over thousands of games says "don't do x play". But computers looking at billions of potential moves can say when that move is optimal, and doesn't have a specific playstyle beyond just doing the best move.
Well then sometimes you get rookie players beating more experienced ones because the rookies have no clue what they’re doing, which throws off the experienced guy. Suppose an experienced player had the idea to replicate this, throw in a couple of bold and risky movies that just don’t seem like the optimal choice to a human but that computers might do, that would probably throw off their opponents
3:05 I run a math tutoring business and my job these days consists mostly of talking with parents who have no idea what their child is even learning, and I cannot possibly describe how cathartic this mini-rant about Aiden's mom was.
As a chess lover but hopefully not a snob (I haven't solved checkers) who recently started binging your videos, I just wanted to address a couple of your kinda open questions/statements: 1. Regarding making a surprise move when someone has been studying your game: this is totally a thing! It's just called a "variation" or, if specifically prepared, well...a "prepared variation." It's less exciting though because *usually* this happens like 20 moves in, between top players. (In a weird way chess sort of is solved, for like, the first move. And then kinda less solved for the 2nd move. And much less solved for the 3rd move. And so on. It's all theory and "book" [memorized] moves for crazy good GMs until a variation happens). 2. The first move advantage is definitely a real thing: Playing as white is sort of like being...the home team? You are "expected" to go for the win, get maximum points for your tournament game, while when playing as black, players often intentionally try to play for the draw. Here is why: There's a concept called "tempo" - if playing with tempo, it just means, the other player needs to "respond" to your moves. You are in control and "asking questions of your opponent." You are like hey I am attacking this piece, what will you do now? Getting this "tempo" concept down is really key - I have tutored kids at chess quite a bit and I introduce the idea pretty early on. Well, with mediocre players like me, this doesn't make a huge difference. But at the very top level of the game, the basic idea is, white starts with tempo, and if everyone keeps playing super accurately, they can maintain their tempo. Even though playing 1. e4 looks innocuous enough, you are actually "attacking" both the empty c5 and d5 squares. You are saying to black, "these are mine now, what are you going to do about it?" So a very good, very aggressive player, can maintain this tempo throughout the opening and into the later stages of the game, which is why white does have a first move advantage. Anyways, longer than I intended. Kinda rambly. Hope it made sense. I really like your channel, keep up the great content! =)
Even for mediocre players (I say this as one) white can be a big advantage because you're going to play an opening you know, whereas f you're black you might need a defense against something you've never seen and just have to reason through it.
One of Magnus's positions that he said in an interview is that the level of competition is so stiff, and the difference between winning and losing is so close, that all a player might need is someone to have a binary signal that "something about this turn has a good move available" and that would be enough to turn the game on its head. Like, Niemann didn't need to have a full computer engine available to him, but rather he'd just need someone to cough in the audience on turn 55 to know that something needs a closer look.
Resigning because you suspect somebody cheated with a single cough and no concrete evidence is still pretty extreme. I feel like a normal response would be to play it out and examine all the video footage later. In a best case scenario he would be able to flex that he won despite the cheating, in the worst case he just loses. It sounds like he made a mistake and was mentally compromised, and subconsciously wanted something to blame, and the easy scapegoat is his opponent who has been caught in the past.
@@gogokowai Yeah, he definitely got tilted, the threat of someone cheating got to him I think, to some degree. But I think the general consensus in chess is that at the highest level when you are playing black, you are going for the draw at best, not for an outright win. So this was a very surprising result. It almost sounds as if he was excessively paranoid...I guess being at the top gets to you sometimes
@@gogokowai some important context: - Carlsen didn't resign that game or the next game, he played this game out and decided to pull out of the tournament after it. Then in a different tournament he had to play against Niemann again and decided to just not play that one game, resigning after 1 move because he couldn't just forfeit that single game otherwise. - The decision of Carlsen to pull out of the tournament came only the next day, after having reviewed the full game with his team, and after having contacted the tournament organization about his suspicion and after being told by them that nothing could be done about it and there would be no investigation, but they would start doing the absolute basics to prevent cheating for the rest of the tournament (metal detectors, radio wave detectors) - The "single cough" thing is not a suggestion of how Niemann would have cheated, it is just to show how little a top player needs to get a significant advantage in a game. At the top level, if there is a winning move, and you tell a superGM there is a winning move, they will find it. So it's not "I think he cheated because someone coughed on move 50", it's "I think he cheated, I don't know how, but even a cough at the right time is enough for a big advantage so he's probably found a way". If you know you'll be alerted when there's a good move, that even gives you a double benefit as a top player: 1) you will find the good move and 2) you don't need to spend time trying to figure out how good the position is (which is a large part of top level chess, and is something a lot of their allotted time is spent on) - Finally: Cheating has been an open secret in top level chess for a while. It's just extremely hard to prove someone is cheating unless you catch them literally with their pants down (Like with Igor Rausis).
@@nolifeorname5731Are you implying that a content creator would leave out details for the sake of a more cohesive narrative and that I assumed the worst and started spreading misinformation in the comment section? That would never happen.
Except in tournaments it is “touch move”. Once you touch a piece you have to move it. So if your partner wants to tell you not to move the piece he can’t give you a signal when you touch it. I never thought of this but that’s probably to circumvent cheating as you described. I realize it could still be done by say gesturing towards the piece but it wouldn’t be easy.
I'm not sure if you covered it, but iirc the deep reinforcement learning-based chess engines like AlphaZero have a distinct flavor to their moves vs. the ones with handcrafted evaluation engines like Stockfish. both will beat a human every time, but AlphaZero is known for being ruthlessly aggressive, playing dramatic sacrifices and doing a lot with pawn position. You can even tweak the engines so that they toy with their opponent, or play defensively, etc. So I think maybe there's so much room above human play for the machines that the concept of "top engine move" isn't so well defined, in terms of catching cheaters. Not that I think Niemann cheated, like you said there's no tangible evidence, just a fun fact on chess software.
You might find it interesting to know that Stockfish nowadays also uses a neural network in its evaluation function, i.e. it is no longer handcrafted. It is not a deep network like that of AlphaZero and it does not need a GPU to evaluate (its uses the CPU's SSE.AVX/AVX2/AVX512 instructions).
So we worship the engine now as if it's omniscient and always producing the best moves, but we also did that for engines 10 years ago. Engines now crush the engines of yesteryear. Over the years I would expect the same to happen to the current engines. We won't know if the recommended moves are best unless chess is solved
The same with evaluation, there is no gradual evaluation of a position: Either it is winning or losing or a draw for best play. All numbers in between mean that the engine is too weak or shallow to determine. So a game to guess the evaluation of a position would make no sense, as the engines of tomorrow could have a different evaluation.
I went through a chess phase in my 20s which lasted a couple of months. Turns out, I liked the idea of chess way more than I liked playing chess. Also, as somebody who used to give piano lessons, parents and grandparents are often terrible judges of how inherently talented their kid is at something.
Lol same. I figured out a way that worked for me to improve really quickly and then promptly lost motivation. Still find high level chess really fascinating and I like how you don't have to be very good to appreciate calculations even at the Grand Master level (unlike a lot of sports/games which often require a level of proficiency/understanding to grasp how good people in the top level really are).
@@민정-q3m I heard Hikaru talking about how he could calculate multiple lines at once then file them away and come back to them like tabs on a computer. So I tried replicating exactly that. I played multiple games at once (on between separate tabs) against bots that were at or slightly above my elo. I'd start with like 3 games at once, then as I got better I quickly went to like 6 or 7, then I upped the bot elo and went back to like 4 games, and so on and so on. So yeah, that's my secret: play a bunch of games at once. Then when you go back to playing one game it feels like you have all this extra memory space and processing power compared to before. I also did zero studying on openings and theory, I'm sure that if I coupled the multiple tabs/games practice with studying some actual theory I could break 2000 elo in well under a year (I started at under half that). Maybe it was just a strategy that happened to work really well for me and other strategies tend to work better for other people, but it was some of the fastest I can remember ever improving at a skill. I just don't have interest in playing chess any more. I've gone through phases of interest in other games but this is the only one I've felt like I found a cheat code. (In poker for instance I'm convinced to get to the professional level you need to brute force learn how to do the statistics on a dime coupled with some decision theory study). And I kinda have other hobbies I'm obsessed with which take priority over mastering games that I'm already better than the average person at (but nowhere near the level of a more serious enthusiast).
I'm glad you mentioned "The Turk" , that's a really weird piece of chess history. The guy who built that thing was a real inventor, he just built it for fun to show off at a party. Much later though, some royal wanted him to bring it back out to show his friends at another party. He was busy working on steam engines, he didn't actually want to bother with the fraudulent thing, but they were SO enamored with his "invention' that he was practically commanded to fix it up and bring it.
The vibrating anal beads story was the most hilarious thing of the year to me, I could picture Niemann moving slightly on his chair to better "feel the move"
It resulted in a 19 year old being singled out and having a metal detector waved over his ass in front of a world stage. It was an awful thing to suggest and it had a serious impact on public perception of him, and that's regardless of whether or not he was cheating.
@@acollierastro OMG, they have no idea you're taking the piss in between jokes do they? Lmfao, the sarcastic conversational narrative is priceless. Be appreciated, consider emigration.
@@robertvarner9519 I was thinking this when listening to her talk about doing adjunct teaching. I wish I could have taken my astro class with her as the lecturer.
Yeah it’s like “I dunno I played the game and won, that’s how chess do.” I don’t think I could come up with a more analytic answer than that if I were asked the same question. Granted I’m no chess player but still
Thank you for defending checkers! I had an argument about how "chess was superior" and me saying that I enjoy chess and checkers was somehow unacceptable.
Magnus was mad because he lost in an endgame as black. Practically never happens to him. It was a big deal because he's the greatest endgame player of all time. So for him to lose to someone rated below 2700 in an endgame as black sent the chess community over the edge. Also... Hans gave some of the worst commentary ever on his chess game. So it made a lot of people suspicious...
@@thechesssavage6400 What if he was playing purple vs mauve? What's your belief on that? How about if he was playing banana green vs crystal pink? How about if chess pieces were donuts with chess symbols on them and he was playing glazed pink with sprinkles vs chocolate filled with powdered sugar? Hint: In all those cases it is up to the player, not the paint or type of wood or pigment in the plastic. Or the glaze.
@@d3nza482 If purple went before mauve it definitely makes a difference. The opening repertoire is so different between black and white because there is a functional difference between the sides
@@d3nza482 So, you don't really know anything about what people are talking about right now, and I'm gonna clue you in. When you are talking about chess "white" and "black" are also ways of saying "the player that goes first" and "the player that goes second". Pigment is irrelevant, because no one is talking about pigment of the chess pieces when they're talking about chess.
There was also postgame interviews with Niemann where he struggled to explain his moves against Carlsen and againstbAli Reza - i.e. wasn't able to tell the interviewer the lines he saw that would have justified the moves he made
One thing that complicated the image of niemann was that in a later interview, following another game, he was not able to explain some of His best moves.
This reminds me of the cheater sleuths analyzing speedruns in gaming, where they count frames and calculate odds of certain patterns occurring. It's so much fun to learn how they do it.
I wonder if the Chess Federation ever offered Niemann amnesty in return for disclosing his method. I know that happens occasionally with speedrunning as well as in motorsport like Formula One.
Magnus based his assessment not on the moves, but on Niemann looking unconcentrated during the game. Also his post game interview (where the player and a commentator analyse the game) contained some rather egregious errors. If someone plays like they're possessed by stockfish and then in the interview say they could also have done this other thing (confidently blundering a piece and the game) it makes people wonder.
When I played chess competitively as a teen I remember us all agreeing that the only people who think chess players are smart are non-chess players (i.e. people don't play in a club or in tournaments). Stereotypes suggest that there could be a cultural difference between Sweden and the US on such a matter, I guess. It's not orange Pepsi, it's orange juice in a Pepsi bottle. Magnus used to always drink OJ during games until some sport physiologist told him it was a bad idea.
I think most very good chess players in the states have such weird huge egos they think they're smart as well, especially since a lot of them have never really had to do anything else, and the public generally thinks they're smart, but people who play a little or encounter them peripherally are typically very aware that not all competence is generalizable.
You can't possibly know how spot on you are on the "amount of cheating". Cheating like Nieman has been found to cheat is so obvious that it is like the guys is screaming out loud "I am a cheater, get me". But actually, any elite player knows that they just the slight hint of " you are better in this position" twice or three times in a gamewould make them almost unbeatable. This is what makes cheating so scary at chess; smart cheating is virtually undetectable
"Is Aiden so smart, or have you been calling him so smart for his entire life, and for the first time, he's confronted with the fact that he can't get stuff right away. And maybe math is a little harder than he thought and it doesn't just come easily, and he's too scared to fail. Too scared to look at the worksheet and not understand, so he just doesn't do it." Brutal truth. Love it.
I knew kids who were plenty smart, but lazy. They didn't have anything wrong with them, they were just slackers who didn't want to be bothered living up to their mother's dreams for them. I know this because I myself am one. Gifted classed all through school. But I never had to study. I would read the text book during the first week of class and then go off memory the rest of the semester. I'd do homework on the bus to school. Then when it got hard and I actually needed some study habits, I didn't have them, and I didn't care enough to start trying at that point. Her comment hit home with me because it summed up my own experience quite well. I just don't have any ambition. I like to write, and have written 3 novels, but never bothered to try and query them with an agent or even self-pub them. I get a few beta-readers, and if they like it, that's enough satisfaction for me. But I know I'm lazy and I also know her words perfectly describe people like me.@@falquicao8331
@@falquicao8331 yeah i had this experience. i was put in 3rd grade math during second grade an voila! my grades were better. I wasn't a struggling perfectionist. I would just rather play with markers and crafts if you aren't teaching me anything. it is also possible the 2nd grade teacher just didn't like me/had a weird bias against me/had a weird teaching style, and the 3rd grade teachers were grading me accurately. who knows. I was a very bullied neurodivergent little kid; i dont think i realized what school was for. it was just the place where we did fun puzzles but the people were mean and loud.
With so few comments, maybe I will be the first commenter who notes that an early version of a chess computer immediately sacrificed its queen for no reason because it had been trained on GM games where that sacrifice was frequently a winning line.
24:00 that's not why they think he cheated. some very good chess players have said his moves seemed unnatural (the type a computer would find but a human wouldn't), and on top of that niemann showed he couldn't explain his thought process in the post game interviews
Well they say it but apparently an algorithm specifically designed to catch cheaters doesn't think so. I believe they believe he cheated. But just because a guy looks guilty doesn't mean he is. I think the whole 'scandal' was pretty disgusting. Even if eventually it turns out he cheated they certainly accused him of doing it without actually knowing it which demonstrates some really poor judgement.
My brother was one of the students sho was too stifled by the school environment to do well... He dropped out, immediately got his ged at like 15 then taught himself how to program before joining the airforce a few years later. Hes now by far the most financially and imo socially successful person in ny family. Yeah that whole too bored/smart for school is a real thing.
Around 9:00, when you're discussing the magnitude of cheating, I think it's important to remember that the people that are bad at cheating are the ones that get caught.
Super controversial I know, but I'll put this out there as someone diagnosed long before it was a dime a dozen diagnosis... But A.D.D. is a thing. Your comment about "Aiden is so smart, he's just bored". I 100% get your point, and some kids are just brats, but that's something I've always heard and dealt with my entire life. It's not necessarily boredom it's like an impassible brick wall that'll ruin your life unless you find your own ways to circumvent your brain. No amount of "Just knock it out in 5 mins and take a nap", It's sitting there for hours, trying your hardest to just get it done, and your brain kicking and screaming and doing it's damndest to push you in another direction. I've always heard the same thing, even now in my professional life "Oh, you're so smart, if you just applied yourself!". Oh sure, I'd love to, let me jump right into that and just knock out everything with my perfect motivation and focus. It's nuanced and complicated. Anyways, Rant over. Just throwing it out there.
As someone with ADHD, I had the exact same thought during that section. I really wish I had been diagnosed and treated for it as a kid, because it's exactly this sort of "if you're so smart why don't you just sit down and do it?" attitude from my parents and teachers that made school absolute hell for me :( (And not to mention the anxiety and depression which is often comorbid with ADHD, and gets triggered by struggling with completing work...)
@@miss-magic-maya I was diagnosed at 13. To put that into perspective of how long ago THAT was... I was in the last phase clinical trials of Adderall. As in, It was just barely approved for use by physicians and I had to have regular checkups and bloodwork done to make sure it wasn't screwing something up. Anyways, To my point. Even diagnosed and on meds it doesn't make it go away. It just takes the edge off. I mean, Sure, the meds help you focus but it's up to you to focus on THE RIGHT THINGS. I take a different med now but there are millions of times I took my adderall dose and then hyperfocused on a video game. But I have done some amazingly creative stuff when my mind behaves too. tl;dr Diagnoses and medication don't change the core issue
Same. With adhd. I would test really well but never had care nor motivation to do daily work as it didn't interest me. It always manifested as a lack of motivation for that which isnt fun rather than being slow or dumb. Sucks fr. Luckily I always enjoyed tests elsewise id have never pass High school. FWIW I had excellent grades in AP classes and did extremely poorly in my middle range classes, because AP classes were interesting and so i chose them. Wish I stayed on my meds
I've had this problem my whole life. Never went to a doctor about it unfortunately. This problem still halts me from time to time. My undergrad years were very good but it took me 18 days to apply to one grad school because I just couldn't get myself to do it.
Even if perfectly play by both sides would result in a draw (which, I agree, seems most likely), it's still possible for white to have a real, non-psychological advantage due to having the first move. Note, for instance, that opening theory is highly asymmetrical between white and black - there are many openings in which white is able to make a strong attack and seek an early checkmate, whereas the strategic considerations for black usually focus on equalizing first and then setting up a counterattack. I suspect that if one looked at the set of games featuring only moves that a reasonably strong chess player would be able to calculate as "good moves", more of those games end up as a win for white than as a win for black.
Sorry, immediately after writing that I realized it might read rather like "man explains things to women". I didn't mean it that way; I was just musing to myself about whether the first move advantage is real or psychological, and if the former, in what way it could be real.
@@stevenklinden As I understand it, the modern chess computers learn by "AI". The computer just plays itself countless times, and slowly improves. If that's the case, I would assume you could look at the results, and see if white has a real advantage. I assume it does, but very small.
@@ronald3836 Only on a very broad understanding of the word "psychological". Since the players aren't perfect calculating machines, there can be, for instance, achieving a position in which it is - objectively - more difficult for your opponent to calculate optimal play is advantageous, and it seems a bit misleading to me to call that merely a psychological advantage.
@@stevenklinden I agree that psychology only really comes into play when you are close to perfect play. E.g. a computer with a 32-men endgame database would have to be taught to distinguish moves that easy for moves from moves that are difficult. Otherwise the computer might play a series of drawing moves that even an amateur chess player might be able to draw against.
re: the first move advantage thing.. general rule of thumb is that in the opening of a chess game, white is playing for an advantage while black is playing for equality. you can see this reflected in the way that a lot of theoretically established openings for white have the word "attack" in their names (the keres attack, the english attack) but never "defense," while for black it's exactly the opposite (sicilian defense, king's indian defense etc, but no "attacks"). white's advantage is very important and real on some level, but it exists in an interesting sort of dissonance with your 100% correct hunch that chess is basically definitely a draw.
“Math was a little harder than he thought, and he’s just too scared to look at the worksheet and not understand so he just doesn’t do it, and so he’s failing.” I’ll admit it, that cut super deep.
I think a nuance that might be missed here is that when chess players are this good, any sort of indication is an amazing, amazing help to them. Sometimes, there are these incredibly hard to spot errors that the other person does. Since it's so much about pattern recognizing, it means that a blunder that one player at the top misses, the other might miss to, because it's happening somewhere obscured to them. If Niemann's plug vibrated once when it was a crucial move, that would likely be enough for any of these players to find the right move. The knowledge that "there's only one good move, and it's winning" just does something ephemeral at that level, to where they'll just 99% of the time find the right move if they have that knowledge, as compared to maybe 20% if not.
One thing important to mention in how a chess engine works is evaluation functions and search. Right, when the board is in some position, the computer will give it a score, like "this position is +0.837," or "that position is -1.936." These evaluation functions do not have to be very good. In fact, the best computer engines often use fairly cheap and easy to calculate functions and then they search the game tree in an order that roughly looks like "best first." For complicated reasons of memory efficiency it's actually a form of branch and bound search in most cases. It plays for both sides until some depth is reached or it looks really bad for one side. If it does, that branch is abandoned because obviously neither side wants to play bad moves if they can play good moves instead. Effectively it propogates the acceptable limits of cutoffs down the tree and keeps evaluating positions for every possible move. So the computer doesn't actually know better than you about a position. It just thinks about millions of promising future positions faster than you could and lets that information decide what move it should make now. There is however another kind of engine that has recently become prominent. Neural network engines. This is where the engine uses some giant mess of linear algebra to learn how it plays when it's thinking really deeply, and uses it to train it to play like that in one move. Then this can be used to train an even bigger mess of linear algebra to play like that one did at really high search depth but in one move. These neural networks are far more complex and slow to compute than the handcrafted, relatively simple evaluation functions of traditional engines, and it's harder to know if they actually "understand" chess or not, but the benefit is that their evaluation function without even thinking multiple moves deep is already master level, and they can still search down the tree to improve on that. Finally, it's off topic, but there are LLMs. Basically, ChatGPT can help you cheat at chess, but it's more like the level of a good human player than a godlike incomprehensible superbeing the way the others are.
It's also worth noting that when it comes to Chat GPT, as with a lot of things, it will also occasionally (and confidently) produce something that's an outright fabrication. An entirely invalid move, like moving a pawn three spaces. Chat GPT doesn't seem to have the strongest grasp on the actual rules of the game and will occasionally attempt to break them unless you plug it into software that locks it into viable moves.
@@JaMaAuWright I think "strongest grasp" is an overstatement. Chat GPT doesn't know the rules of chess at all. It doesn't even know what a rule is. What it does is things that look similar to chess moves. Chat GPT wouldn't help you cheat at chess, it's considerably worse at it than even basic computers from 20 years ago.
Strong chess players (GMs , IMs, National Masters like myself) regard physicists in such high regard but some professionals that I know hold excellent chess players in very high regard too. A funny anecdote is that when I had a major (medical) operation back in 2007, the doctor was flabbergasted that he was performing an operation on a chess master. I've been using chess engines a long time, since the mid 1980s when they were much weaker. It took awhile before desktop computers were supplied with multiple microprocessors. Once that happened, the combination of fast hardware with good software made chess engines much stronger. I have not been able to handle them since around 2005 or so, when old Fritz 6 and Fritz 7 were state of the art, rated maybe FIDE 2700 or so. Nowadays a recent version of stockfish running on some kind of a monstrous piece of hardware is a little over 3600 strength. It could spot Magnus Carlsen a pawn and a match played at such odds would be fairly close.
Here's a funny story about a chess player and a physicist. Albert Einstein and Emmanuel Lasker (the second ever world chess champion) were good friends and they would visit each other to talk all the time. The problem was that Lasker only ever wanted to talk about physics while they were together, while Einstein only ever wanted to talk about chess.
@@michaellisinski2822 Wilhelm Steinitz paid a visit to Paul Morphy down in New Orleans at least one time that I know of, a few years after Morphy had quit playing chess competitively.. However, Morphy agreed to meet with Steinitz only on the condition that they did not talk about chess or play chess. I don't know much more about that incident but it struck me as rather funny. Steinitz was perhaps the strongest player in the world at the time with the possible exception of Paul Morphy.
Even Carlsen would get destroyed in a match vs a strong engine if Carlsen had only pawn odds. For the match to be close Carlsen would need a minor piece odds.
I want to mention that there was this iconic series where Kasparov did best Alphago when chess engines where just about getting to the point where it would be impossible for players to beat but not yet to the point there was no hope at all. There are a lot of video essays on it. It was last stand of humanity vibes.
FYI, we play tic tac toe wrong. It was originally 3 mans morris (from Rome I believe). Where you each only get 3 pieces and once played you move them around trying to get 3 in a row. The game later grew to 6 and 9 man morris which is the currant popular game. Though I have heard of 12 and 15 though they sound impossible to play.
The argument about how he is young and has not lived in a world without chess bots is actually such a good point! Also the match in the end is hilarious😂
In tic tac toe you can be doomed already from the second piece. First guy places middle, If the second one places in one of the middle squares(not the corners) you have a forced win for the first player... This is the only "strategy" I ever found in tic tac toe, after a lot of searching through bored days in my youth. I did not explore fully other starting positions, as they were obviously inferior, but I believe I fully explored starting center. But it is kind of fun being able do declare your victory on the opponents first placement!
@@silphv the corner is maybe more likely to trick a careless opponent (or a child) but I don't know if you can can call it better, since either move is a trivial forced draw as long as both sides play sensibly
@@dm9910 That's just how the game is. If both players play optimally it's a draw. Which includes starting in the corner. I dunno, look it up, it's not really an open question.
@@silphv I'm not disagreeing with the claim that going in the corner is *an* optimal move. You claimed that the corner is a better first move than the middle. I'm disagreeing, on the basis that they are both optimal moves.
First move advantage is absolutely a thing. It's like the Gawain and the green knight where they agree to take turns chopping of each other's heads. Having the first swing is a big deal: 1. Mathematically, you've had more moves than your opponent, unless you make a mistake and lose a tempo. This means the first-move advantage would decrease the longer the game goes on, except 2. The 1st move stops certain moves from your opponent. You took the best move, which in many cases means they can't take the move they wanted to play. So you can maintain disproportionate control through to the midgame. 3. Moving first means you have more control over the opening you play. If I know you're a Sicilian player, I can play a queen pawn game that you're not as comfortable with. Or at lower levels, the first move advantage has more room to take the other player into a weird opening they never played before. I still think chess will be a draw when it gets solved, but there's definitely 1st move advantage. Even in tic-tac-toe, there's very clear 1st move advantage. If the game is played imperfectly and doesn't end in a draw, it's almost always the 1st player who wins.
There's actually a very important factor you seemingly missed in your research. The main way that people think Hans cheated who actually know what they're talking about isn't engine help but a prep leak - he seems to have studied a very specific, off-beat line the night before, which Magnus happened to play into. It's astronomically unlikely unless someone had leaked to him that Magnus was planning to do that.
This is interesting! Do you think that counts as cheating? I don’t know if I do. If Magnus is discussing his strategy pregame and Hans studies it, I mean.
@@acollierastro It's definitely cheating. Prep is not casual discussion, it's intensive, computer-aided study of specific lines, often with anonymous partners using burner accounts. If you know what lines your opponent is preparing and they don't know anyone knows, you've essentially violated the only information inequality that can exist in an open-information game. The point is, you had to have done something untoward to even *get* that information.
I think this is what he SAID he did. If you don't believe that he had prepared this very specific line, then you believe he is lying about it. Why would he lie... (did he get the best move from a computer instead of from doing his homework the night before)?
@@Pablo360able I disagree - prep spying is not against the rules of the tournament so it isn't cheating. I mean, it's definitely scummy behavior, but it's not the same as using a computer during the game. And yeah, if you had to hack a website or something to get it you might be breaking some rules / laws. But if it's just that you get one of Magnus' team to talk to you, that's not cheating. That said, prep leak is pretty unlikely, Magnus hasn't stayed on top of the world in chess without having an incredibly airtight team. And really deep prep is maaaybe 20 moves, and the game went for 57, with as others have mentioned, Niemann outplaying Carlsen's notoriously strong endgame. Carlsen doesn't have some magic prep for endgames that you could get from a leak, he just calculates really accurately.
that is really fascinating. first video of your i saw, was about crackpots in physics, and the second - the one when you've totally put yourself into their shoes (without aggression part)!))
as far as I understand it the first move advantage essentially means that white gets to decide a whole lot about the opening of the game because black is always catching up. also another sign of cheating you didn't mention (or maybe I wasn't paying attention in that case my bad) is that cheaters will take an unusual amount of time (both unusually long or unusually short) to make their move. In online chess cheaters will most often take the time it takes to input their opponent's move into an engine and get the best move out to play their move, no matter what move. Whereas sometimes a move is obvious and you would expect someone (at a certain level of course) to play it instantly, or sometimes it's hard to spot and you would expect them to think about it for a while. So when they take roughly the same amount of time to play every move this is also a possible sign of cheating
What people mean when they say that Hans played like a bot is not quite how you understood it. Basically there are times when you can have a move that puts you at a short-term disadvantage but eventually pays off 10 15 20 years in the future. The thing is, outside of very forcing sequences there's a cognitive limit to how much even the best chess players can see ahead. Stockfish might be willing to sacrifice a rook because it can see that 20 moves later it can win a queen. That kind of move is one that a human could never play because they would have to not only see the idea that remember takes 20 moves to materialized, they would also have to see the response every adjustment possible by the other player, and see that there's no way for the other player to get out of it, and be incredibly confident in their assessment because if they're wrong then they basically just lose. So if a human plays that move and is able to follow up on it, then that means they're cheating because they've seen something that it is literally cognitively impossible for any human to see. The best players in the world can only see 4 or 5 moves ahead outside of special circumstances.
I was at the first tournament where a computer beat an active grandmaster. It was in Long Beach, in the late 80s -- Deep Thought vs Bent Larsen. Bent tried to play into complications against the machine, and it just out-calculated him. Mikhail Tal was at that tournament, but he signed the "NC" (no-computer) list, meaning he didn't want to play any machines. We all understood that history had just been made, and it was interesting (and in some ways a bit depressing) because we knew where the trend was going -- we just didn't know when we'd get there. All that said, I don't think modern computers have "solved" chess as far as humans are concerned. They have become excellent coaches and training partners, and that will continue. But humans still have to put in the work. I made it briefly over Elo 2200 both in USCF and FIDE in 1999, and then life got in the way and those numbers came down. Finally made a return to competitive chess last year after being out for 20 years... and found myself in a sea of young people (sometimes -really- young) who could freaking PLAY. They were 2-3 categories better than their ratings, all because they had been studying and playing online during covid, and now they were showing up and very hungry. It's an interesting (if painful) time to get back into the swing of things.
besides the history of cheating i think the other reason ppl believe niemann cheated even though there isnt solid proof of it is just that hes so unlikable lol. dude is just annoying
Also, Magnus' PR. Starting with the name. Shame his parents weren't more creative. Calling him Optimus Prime or Megatron would have made it all so much more entertaining.
From what I understand magnus played a specific variation that he only played like one or two times before and somehow hans was prepared for that specific one, which according to some high level chess players(Hikaru and Gotham chess who are content creators so there's some possible alternative motive there)i s extremely unlikely to have happened. Also there's this whole controversy with han's mentor also being some sort of cheater but i'm not sure about that one.
When I was studying computer programming, one of my first programs was a Tic-Tac-Toe game. I was able to write an algorithm such that the computer would either win or draw, but never lose. This was not a brute force algorithm as Angela seems to be suggesting, but rather an abstracted procedure which guaranteed a non-losing outcome.
On the "My son is so smart, he's just bored. That's why he's failing maths" note. This is a flag for either undiagnosed neurodivergence or shitty parenting. Usually both. Felt personally attacked, sorry.
Sacrificing pieces in a crazy way can be surprisingly effective even when you don't know what you're doing. I'm terrible at chess, I have ADHD I can barely think half a move ahead and always miss obvious stuff; I tend to get confused by all the pieces so I will often deliberately (or more often accidentally) sacrifice a piece just to get fewer pieces on the board to worry about. And 20 years ago I played an older teacher who was decent at chess in an informal school tournament; I lost obviously; but because I kept sacrificing pieces including a queen for a queen and a bunch of other pieces; he was really thrown off because none of his usual strategies worked since he'd unexpectedly lost important pieces (because a normal player would at least at the time never pointlessly sacrifice a queen that way) early on and he could not rely on me being rational for any move, the game went on and on for much longer than reasonable when playing such a noob as me.
Just to further prove how terrible I'm at chess.. These days I barely even remember the rules for how the pieces move; but 20-ish years I at least knew that. And once I played a woman at an after party (so to be fair we were both drunk) who had to learn how the pieces move before playing, so I logically should have had an advantage since I had played the game tens of times before while she had never played. After a long while I suddenly noticed she was moving a piece (probably a pawn) such that she she was putting her own king in check, which obviously is not allowed; so I protested and she moved the piece back. And that's when I noticed she was still in check; with at least two pieces.... and not only that but my king was also in check with at least one of her pieces; and we had both no idea for how many turns we'd both been in check. So we decided to give up on the game... Is that a draw.... or is there another word for this?
See, you frame that as a success story, but you did in fact lose. All you did, at the very most, was draw out your loss. Pointless sacrifice can never actually give you an advantage unless you make your opponent run out the clock.
I was that child who refused to do homework. I did the work in class, and aced the exams, but when I got home, I just didn't want to think about it any more. This wasn't because I thought I was so much smarter than anyone else, but because my school was so slow and behind in the curriculum. I had already learned everything they were teaching from my older brother, years prior. There were so many other things that I could spend my time doing rather than something I had already done many times before.
Back in the late 80's, I had a Chess program for the Commodore 64. The program was actually printed in the pages of a magazine (in machine code), and I sat for hours typing it, and saving it to a tape (as in audio cassette) drive. The principle was that each piece had a 'value', based on its importance. A pawn would have a value of 1, a bishop might have a 5, a queen 10, and the king a 1000 (an 'impossibly high' value, because the whole game hinges on not losing him). The game had 10 levels. On level 1, the program would look at the current positions on the board and consider the score after every possible move, as well as every possible response. This took about 2 seconds. At level 5, it considered every possible move, through 5 iterations of moves and responses, trying to maximize its attack score, and minimize the potential losses. Because of the limited processor speed, you could literally prepare and eat a sandwich, between each move, above about level 7, once the pieces were developed to the point where more than one or two of them had attack/defend options. As things progressed, chess engines were loaded with databases containing every tournament chess game ever recorded, along with the calculated win/loss statistics for every move, for practically every configuration of the pieces on the board. One basic advantage is that, the computer should never ever blunder (lose a piece needlessly, because it didn't consider the opponent's very next move, or next few moves). So, the program knows the best moves, by the best players, in every game that has ever been recorded... and, because of today's processing power and speed, it can weigh every possible move and response for 10, 20, 30 moves into the future. If the meat-based player gets creative and does something wholly unusual (as related to the database), it can recalculate and 'preview' the new scenario, in milliseconds. I'm sure there are layers of even more clever concepts baked in, as it's been a while since I've thought very deeply about any of this.
My dad told me a story about when he had the opportunity to play against some chess grand-master (I forget the name). My dad was an amateur, and used some very unconventional moves in the early game. According to his story, my dad's non-standard opening baffled this grand-master, who then stood up and threw the board across the room, shouting "Why must I let this idiot beat me". I'm not sure how much of that story was true, but my dad was a sailor. So . . . grain of salt and all that. But the point is that someone who is so well trained in the "proper way" to play a given game, just might be thrown off balance by someone who has no idea what they're doing.
Seems unlikely. If their levels were closer, perhaps, but a GM isn't going to be rattled by sub-optimal moves in the opening from someone well below their level. Instead, it's actually the other way around - the GM is going to know exactly how to punish the suboptimal play and cruise to victory. Now, two GMs against each other and one of them can surprise the other with computer preparation - in that case, it's usually one deviation from the established main and side lines, and the advantage is largely in time gained as the opponent is unfamiliar with it and has to think longer while the other can blitz out the move. This is why, in these cases, the GM "defending" will often play moves that complicate the position and make it unclear, to get the other GM out of prep so they are back on even terms and thinking on their own.
Not even a little bit plausible. A novice playing a weird opening might throw off someone with more a little experience than the novice, but a GM would not have any trouble destroying them. The only way I could see if is if it were like a blindfold simul and the GM was playing 20 boards and just lost track of your dad's. Even then, the whole throwing the board thing sounds made up.
Did you see Wargames? The Matthew Broderick movie from the '80s? They actually forced the computer to play tic tac toe against itself to illustrate that some games are unwinnable and should not be played, or at least are of this type and you know that going in. You get to see the computer learn this. Obviously this is a movie and it's all fake but it was still pretty neat, especially in an era when people teaching computers had learned on punch cards
I think typically when someone says they play like a computer instead of a human it means more like they played a move that makes the position super complicated. Like people like positions that are simple and the moves being intuitive. Sometimes computers find a line where you have to keep making the best move over and over again which is super hard for humans to spot. I don't really think Hans did that this game, but whatever. Also I heard another story of people cheating and it was so clever, but they got caught. So basically they had one guy watching the game and they had a system where depending on where their friend sat in the venue they would know what move to play. I think they got caught because they figured out that he kept glancing away at the board.
There's a reason why a subject matter specialist's opinion on average is more valuable than that of an outside observer. To make an analogy, imagine if Nature magazine was sending Physics papers to famous violinists for peer review. Some important details would certainly be missed in such exchange, wouldn't they?
Wow what you said about math being scary when it start getting hard is so real that stopped me before but you’re really helping me get into this stuff as an adult
The drama was quite interesting. It sparkle quite intensive disscusion, what should be considered proof of cheating, which methods are reliable, a lot of analysis (even fake one)...so yes, the drama was fun. Anyway I think he didn't cheat in that one specific game. magnus either had just a bad day or he already suspect hans of cheating and it affect his gameplay, because he just played poorly (for his standart).
In the 1970s, there was an article in Scientific American (I think it was Martin Gardner's column) on making a simple computer using matchboxes and M&Ms to play tic-tac-toe.
In a human game part of the first move advantage is you can practise from after the first move while the opponent has to either predict your first move for practice from before the first move.
Magnus just played poorly against Hans. It is easy to cheat at chess online but very difficult to do it in person at a FIDE sancioned event. Poor sportsmanship by Magnus to accuse his opponent of cheating with no evidence
Possibly....but Hans has cheated, admitted to cheating, and been shown to have cheated more than he admitted. So maybe Magnus is a poor sport...but Hans is a terrible person. He should not be allowed to compete at the highest level anymore since he has no ability to demonstrate he "isn't" cheating....and we all know, you can't prove negatives.
@trepidati0n533 The point is, no matter how well respected a player is. They themselves are not the best impartial judge on whether they were cheated against or not. This applies to Garrett Adelstein and Magnus Carlsen
also i dont think its so much that chess players thought only a human could win at chess bc it required some kind of human passion that a computer couldn't emulate, just that no one could have imagined what computers would become. like you said chess computers used to suck, programming chess used to be this really big challenge so it prob seemed unlikely to a lot of people that it could be done well. it is pretty surreal how far weve come and how rapidly technology has improved lol
Not so sure about that, I think even if someone understood conceptually computers playing chess and always getting better, they might still entertain a bit of superstition about the "human element". I guess I was reminded of the AlphaGo documentary (about the first AI to beat a Go champion) where there was a lot of that sentiment going around. Of course Go is a very different (and even significantly more complex) game than chess, where trying to think X moves ahead like a decision tree just isn't really how you play well, because there are far too many possible moves. So in that case it is actually true that it requires some human element-the AI did win but it was trained on a massive data set of human-played games. Still, even with chess I think it is a really tempting (hubristic) thought.
@@silphv that's less that a human element is required, and more that it's much easier to program that way; in theory, it would absolutely be possible to train an AI on AI only to always play perfectly, we just don't have the processing power yet to do that in a reasonable amount of time. Hell, if it's a matrioshka brain running for a million years it might even actually solve it.
@@nerdyspinosaurid Actually no you're completely right, I read more about it and, while the AlphaGo that was featured in the documentary was trained on human-played games at first (before playing a bunch of games against itself to improve), they've since made multiple newer versions including one that taught itself without ever seeing a human-played game and which was many times stronger of a player than the previous models. As of several years ago even it's entirely untouchable by human players, so I think they just stopped there.
The thing that's always confused me and rubbed me the wrong way about this scandal is the fact that Magnus's evidence for Hans cheating was that... he lost the game. I don't mean to draw a direct comparison in all accounts when I say this, but try to think to the last time you've seen this reaction from someone after they lost in a multiplayer game. Say, for instance, maybe someone who takes Smash Bros. a little too seriously got salty and accused you of cheating reflexively. Something like that. Now think back to the last 10 times you've seen that accusation made. 50. 100. Every time you've ever seen or heard anyone lose at a multiplayer game and accuse their opponent(s) of cheating. Regardless of whether or not they're right, has there ever been a time in your life you've ever seen this accusation being made where it *hasn't* been transparently petty and a symptom of being a whiny entitled baby? Because I, personally, have never seen someone lose a game and accuse the winner of cheating that wasn't an obvious attempt to handwave their own performance by claiming an unfair disadvantage. I have no reason to believe Magnus's accusation was the sole exception to this observation. But that's all just preamble, here's the actual problem I have with this: why even play the game, then? Essentially Magnus is saying that because Hans was not expected to win, the fact that he did is therefore indicative of cheating. Really think about that for a second, because if that's the case, then it's not possible for Hans to have won. And if you're thinking "yeah well he just wasn't as good so of course he didn't," then... so it was pre-decided for him that he should lose? If the only acceptable outcome is for Hans to lose, then why is he even playing it? We know he doesn't like playing games when he knows he won't win, so what gives? So now the inverse has to be true, if the only way for Hans to win is if he cheats, then the only way he would even decide to PLAY the game is if he started cheating, right? So now it's a foregone conclusion that as soon as Hans sits down, he's not only already lost, but he's already cheating. There is no point in going through with the game anymore, just give Magnus his trophy, deliver it straight to his home and save him the trouble of having to play the game or even show up at all. And since Magnus is so great at chess and always wins forever all the time... well, just keep doing that. If any other player in a tournament with a lower power level is up against him, that player's only options are to forfeit or cheat, and since cheating isn't, y'know, allowed, the coordinators of the tournament should just call the match before it happens, not even allow it to happen, and send Magnus his trophy in the mail. Meanwhile, Hans will now realize that they can never join another tournament so long as a single other opponent has more than 5 million points. But then when he does so, the tournament will just send HIM the trophy in the mail, right? No games need to be played if he just stomps out the little leagues. But then, who would ever join a tournament if they already knew solely based on experience points who the winner is going to be? Surely they would all pull out the second Hans, or especially Magnus, joins, right? So... why... why have tournaments at all anymore? How do you convince someone to ever play chess professionally ever again if every single match can be determined simply by who has the highest level? Is chess an actual game or is it just a pissing contest? Obviously that's a slippery slope, and I... well I assume that most professional tournaments have match ups between people with near enough battle ranks that it's more open, and I also assume that the level of disparity between Magnus and Hans was unusually large for a matchup at this level of play. But even so, that doesn't make the "I lost so you must be cheating" accusation any less childish nor does it make the paradigm of "the only way you can beat a level 99 chessmancer is to cheat" sensible or tenable in any capacity when it comes to the infrastructure that the idea of a tournament is based on. And I say this regardless of whether or not he cheated, it doesn't matter if he did, as long as your standard of evidence is "well I'm so good that if you beat me it's only because you're using hax" then like... fuck you? No? That's not how games work? That's not how anything works? p.s. Kids don't like doing homework. I hated it. I struggled really hard in grade school and it wasn't until I got into college that I actually started having "good grades" as it were. I've also seen two children both piss and moan about doing homework when they'd rather be playing video games, one aced their stuff when I finally nailed them down to do it, the other was pretty clearly not understanding the material, but they had an exactly equal enthusiasm for putting up with the mindless busywork hellscape that is grade school. The only appreciable difference between my experience as a child and my experience with children is that I wanted to play Primal Rage and they wanted to play Minecraft.
The thing about first move advantage is that chess isn't solved. So, most people agree that if chess was solved it would most likely be a draw ("a chess game, when played to perfection, ends in a draw" is a real saying, not just a meme). But, as long as it isn't, it's been observed that most non-draw games end in white winning. Even for engine games, which are thousands of time better than the best human players.
I think it's interesting that you can program an algorithm to play various levels of less than perfect. Like the program knows the "right" move, but chooses to do a "wrong" move every so often depending on how low you set the "level". I wonder how they validate that a "level 6" delivers just the right amount of "bad' moves at the right frequency.
It's not fascinating at all Every algo has the candidate moves rank list. It simply chooses to play something other than #1 move. If say Stockfish decides to play #2 candidate move throughout the entire game, it would still beat the sht out of Carlsen. And it would proly do it without losing a single game out of 100. OTOH if the algo goes for #8, it would lose against me LOL I promise you that. Fascinating part would be, as you correctly wonder, rating the series of such handicapped moves with correct ELO. Which in fact they DONT. There is an attempt of doing just that with Leela version called Maia, but it's not all that.
It depends on the algorithm, but most of them are built around searching to a certain "depth" (read: number of moves ahead) to find the best move. So instead of purposefully picking a bad move, they can just do a less extensive search which would result in more natural bad moves similar to what a human would do.
I agree. And a surprisingly hard programing problem is making them bad in a human-like way. There's things that human players are more likely to miss, that programmed-down chess computers don't accurately mimic. For example, less-good humans have board-blindness, a tendency to focus on a small area of the board and miss things going on further away in space and time. This makes humans more likely to, for example, miss a capture a bishop can make all the way across the board when that bishop was last moved a long time ago than they are to miss an exchange capture a pawn can make (like you just took my piece, my piece was defended, even as bad player I'm likely to notice I can take back). Chess computers don't tend to act this way, so they will miss obvious moves a human player below the computer's intended level will catch. And in the other direction, because they are evaluating moves by how relatively bad they are, they will play moves that are better than the intended level to avoid really bad outcomes like mate. Low level chess computers almost never get caught by a surprise-mate, something that happens to low level players far more often.
In response to the "Aiden rant" at 3:05, I do want say that the description you give could genuinely apply to someone with ADHD/ASD such as myself. I also know a decent few kids in middle/high school who fit the description you gave, but I for example in 8th grade would refuse to do assigned algebra work because I was bored with it, instead doing calculus problems on my own. I certainly wouldn't say it's the fault of the teacher to assume that a kid who doesn't demonstate proficiency doesn't have it, but I would say there is nonetheless a fault in the Western educational framework in which proof of knowledge is valued so much more than the actual learning itself. I know that there has to be some metric by which to judge student aptitude for practical purposes, but I think the present heirarchical model of learning is at minimum not well suited in particular to autistic people in ways that can lead us to disengage in ways similar to what you described here. I'm not detracting from the point being made because the people you describe do exist, but I would like to add that an additional category of person exists for whom the first half of the description may fit but not the other.
Magnus is arguably the BEST player ever, it's at the point where even the GMs are like "yeah, he's the best". You can almost count on one hand how many times Magnus has lost playing white, in classical games. Niemann couldn't even analyses his game AFTER having just played it. He said that the winning move was one he saw in another game... that game never happened...
You don't play chess, do you? The debate about the non-existent game was about the opening line, not the winning move. And there turned out to be a game where Carlsen had reached the opening line via transposition, so this is not conclusive. It would be a lucky guess for Niemann, as he said himself, to prepare for that line with a different move order, but possible given the previous game.
@@eljanrimsa5843 Oh, hi. We making personal assumptions about each other for no reason? Cool. "You don't wipe your butt after you poop". Cool. Great way to start this off. It WASN'T about the opening line, it was the mid game where there was a move made that was an improvement... but sure, Niemann somehow looked up that specific game that morning, but also didn't remember when or where that game was played... sure... Hey, was it fun watching someone rated 300 points lower than Niemann out analyse him after the game? Basically everything Niemann said was wrong, like OBJECTIVELY wrong. But, hey, cool.
@@aliince9372 but you don't "see" a winning move in another game. that's not how chess preparation works, that's why I thought (and think) you may not be familiar with the process
@@aliince9372 I believe that their point was that it doesn't seem like you know as much about this topic as you seem to think you know, and that make be affecting your judgement on the matter. I'm not all that experienced with chess, so I'm not sure if that's actually true or not, but generally speaking things _tend_ not to be anywhere near as absolute as people on the internet make it seem. I'd wager that probably applies here.
3:40 - as someone who was thought to be "bored" in high school, I can answer this one. I was neither bored nor struggling to understand. I just didn't care. When I got to college and was paying money out of my pocket, I did very well because I was motivated to care.
I dropped out of high school because I was miserably bored in every class to the point where I could barely focus on anything and felt like every assignment was busywork. At least in my district, they seemed to repeat the same material year after year, and I just couldn't handle it. I ended up going to community college at like 17-18 and blowing through everything with a 3.8 GPA before moving on to an actual university. I wasn't dumb -- which is what my teachers thought and legit tried to put me in special education before I dropped out -- I just couldn't focus on anything due to the lack of novelty. My mother was the mother in your story, but y'know, sometimes it is the school's fault. My folks argued for advanced placement or skipping grades or anything else they could think of, but the faculty at both my middle and high schools thought I was a moron because I skipped class and had terrible grades.
@@ronald3836 Did you just start immediately drooling and smashing your keyboard without finishing the paragraph? I'm a dermatology PA. Right now I make a little over 100k and only work four days a week. It's pretty good. If high school had been more accommodating, more reactive to my needs -- who knows? Maybe I'd have been prepared and comfortable enough to go into a STEM field like I dreamed of when I was a kid, but in reality I scrambled for a degree that was relatively affordable, had reliable job prospects, and paid well. I don't think -- and, again, this is in the context of what I experienced in my school district -- that being able to put your nose to the grindstone and monotonously complete very similar assignments perfectly year after year is the same thing as being intelligent. The kids who got scholarships at my school were all socially inept, had the personalities of bricks. They could barely hold a conversation and ten years on, from what I've seen through social networks, didn't do much of anything with their free ride through an Ivy league.
The thing that always bugged me, personality wise, is that first-turn surrender. What's the rationalization behind that? What could a person possibly see on the first turn that makes them change their mind about the game? Or, more likely, what makes a person decide they're going to go to this game with the full intention of forfeiting immediately? Very strange behavior. I can't put myself in that mindset.
I think it was to make clear that Magnus was refusing to play on purpose. If he just doesn't show, there could be all kinds of speculation, maybe he's sick, whatever. But playing the first move demonstrates "I am fully present and able to play this chess game, but I choose not to." There may also be a tournament-rules issue. Forfeiting by no-showing is recorded different from resigning. I don't know if that has some implication for future games / tournaments Magnus might have cared about.
Someone told me that if you resign or don't move on the first move, it doesn't register as a proper resignation or something but rather the match was aborted somehow? So Magnus playing the first move was to make sure that it didnt happen and that he was for sure resigning intentionally due to his opponent.
It was interesting to me that humans being unable to beat computers at chess was completely unremarkable, and it is today. I still can't help cling to the perspective from my youth when it was unclear if computers could beat humans at chess. It's one of those things were I feel like the world lost a little of its wonder when it was no longer even competitive.
That struck me, as well. Granted, I don't have my finger on the pulse of the chess community. But I remember the excitement back in the 1990s about a computer that might be able to beat a chess champion (for the first time). Apparently now, the computers are so good that no human can ever beat them. Crazy.
I appreciate your very reasonable take. There were thousands of people who were ready and willing to burn Niemann at the stake because he talked funny in an interview. I'm being serious. They thought how he talked, his accent, was proof that he was cheating. People are f*cking stupid. It's nice to hear someone not jump to conclusions for once. I'll defend the fact that we don't know he cheated in this game. Even though he seems like a piece of sh*t otherwise
Your point at the end I think is the reality. Niemann and his coach played a lot against the bots and with the bots that they've started to develop "bot-style" play intuitively. That kind of play could definitely throw off the other players that have trend paths worn in by humans. An engine could suggest a move that has no obvious stength for 5 moves, but if you've steeped yourself with that instinct it just "feels right".
Unfortunately, there isn't any bot-style play for a new generation of humans to "grow up with". It's pure logic and mathematics, which lets bots calculate best moves. And they are playing against the best move that you could make next, as well, so that when you don't make that best possible move, you are automatically at a disadvantage. And so, unless you can "keep up" with the bot, it gets better and better as you get worse and worse. No one can defeat this by intuition, instinct or feelings. And even grand masters do rely on these intangibles, as you can hear when they comment their games, saying "something about this move seemed wrong", even though they are calculating as best they can. You can't reach a 3500 elo by just growing up using and playing with bots. Their moves are often not understandable until the game is over simply because human brains can't think that way or that far ahead. It's not a talent that can be developed somehow, it's pure mathematical brainpower. The bots have taught chess players a lot about what works in openings and endgames, true, but those points in the game have the fewest variables and so are at least minimally comprehensible to humans. But it will take a cyborg or another leap in evolution of the human brain before we can compete with these bots or play like them.
All you can say about lichess level 3 is, it doesn't violate the rules of the game, but it violates all the rules beginners learn in the first lessons, like not letting pawns take your pieces.
what I dislike with lichess lvl3 is that it sometimes plays very well then just hangs their queen when you attack it with a pawn. It's still stronger than me
That was uber entertaining - thanks. Perfect chess by both sides ends in a Draw [not a stalemate]. Chess ratings are Gaussian - so going from say 2,600 Elo to 2,800 Elo is not a linear calculation. Lastly, in the first 10 moves of chess, there are 169,518,829,100,544,000,000,000,000,000 variations - so you can imagine the number for a 30 move game.
"Perfect chess by both sides ends in a Draw [not a stalemate]." Like she said in the video, I feel like we would've heard about it if they solved chess
@@Ken.- dude - if you watch smart yt vids like this then you must realize that saying a stalemate is a type of draw does not entail every draw is a stalemate.
@@JamesJoyce12 Dude - if you watch smart yt vids like this then you must realize that saying "perfect chess ends in a draw [not stalemate]" is not even close to being proven true
18:00 ish -- one non-null tic-tac-toe result that can arise is when you have a stroke mid-game. while there are a few downsides under this situation, strokin out mid-game has many advantages. the most reliable of these is that you no longer have to play tic-tac-toe of course but there is another related one that can occur if you are lucky which is that you dont ever have to play the game, or really any game, ever again
I was maybe a little too young to understand the forced moves when my grandpa was teaching me, so he just started taking the pieces when I didn't do the forced moves. I still didn't learn, I just accepted that my pieces would randomly disappear.
That's interesting, because as a kid I was never taught about being forced to take pieces, or at least, it wasn't emphasized. Of course, no one as a kid would give up a capture, nor would they likely be able to see how forcing an opponent to make one might be a good strategy. And certainly once I learned chess I had no interest in checkers.
The Grandmaster (I think) UA-camr Gotham Chess has some interesting move-by-move commentaries on games played by Stockfish, currently one of the top chess engines. He points out a variety of “a human would never do this” moves.
I'm so triggered by idea that 'first move advantage' would require the game to be fully solved to have an impact lol. I know this isn't a game design channel, while game design kinda my main 'thing', and this channel has a whole, interesting video about how it can feel to hear a smart person talking about something you're an expert in that I'm trying to keep in mind. Still made me twitch though lol.
I have 2 thoughts: 1. I am instinctively disturbed by the idea that someone could be falsely accused of cheating bc they just happen to play a really good game, or because they are just really really good/develop a new way of playing. 2. If this guy has a history of cheating and admits to it, why was he allowed in this tournament in the first place? Were the games he cheated in not tournament games so they dont count?
To make sure it's clear to people watching this video Carlsen was not on a 53 game win streak. He was on a 53 game no loss streak. The record at high level chess for consecutive wins is 20 or 21 unless someone broke that record.
Yah cuz winning 53 games of chess at the profesional level in a row would be insane lmao
I'm on a professional streak of never losing, not even once.
@@culwin that’s solid work
@@xGotDemFragzJRx Yeah almost impossible for any of the best players in history.
@@dannygjkHe is probably joking, and probably played like 2 games and won/tie
That idea that he sacrifices the queen like a bot might do reminds me of a time I sat in with the chess club in high school, and there was this puzzle about how to achieve checkmate from some seemingly hopeless position, which did happen in some big game. Everyone was throwing out ideas and the instructor kept shooting them down, and eventually it went quiet. I just said, "sacrifice the queen." The guy next to me was like, "NOOOOO" and the instructor was like, "yes", and walked us through how this sacrifice offset one other piece by a single square, opening up some diagonal move allowing mate to happen after a bunch of other moves. I sat there looking like this galaxy-brained genius who'd figured out this thing nobody else could, but I only said it because it was the only idea left on the table after everyone else had already said their thing, and it seemed like a taboo. I had no idea how to actually do it.
I played against the guy who yelled "NOOOOO" one time. He checkmated me in like a couple of moves then said I was a good player. Thanks, man.
I've coded a few games, decades ago. You do see computers making these "unusual" moves. When it happens with games you've coded, you can think it's messed up, but then things suddenly turn around many moves later (exactly like one of the games in the AlphaGo tournament). When I've analysed them I've found that the only reason that these moves make sense is if you can do a deep enough analysis to determine that there is a guaranteed potential technical advantage, which means they cannot be successfully replicated by humans without the deep search capability of a computer. I've also noticed that you often never observe from subsequent play just what advantage the computer originally saw when it made the move. The reason is that computers don't always make the compensating payoff move as soon as they can, they just move to _keep that option open._ They often keep rolling forward that option without exercising it, until the human makes a mistake that creates a better option for the computer (or the computer just sees a better option), and it then moves to roll that option forward, and so on. That makes it harder to learn just by playing the computer, because there may be a lot of moves between sacrifice and realised payoff, and the observed payoff may not be the one the computer originally saw when it made the sacrifice. You can make the computer show its current best line of play, and I assume that is what these players do when they use the engines to practice, but humans are always up against their limited search depth compared with computers if they try to replicate that except in the exact same position.
That DOES compute!....terrific answer from a suspected Trolling chess 'bot!...😁....trying to gain a psychological advantage by convincing future human opponents... Remember ..you're beaten BEFOREHAND.... "Nice try carbon based skinsuit".....DIABOLICAL chessbot!👺
As a beginner, there have definitely been times where I played what was evidently a mistake and the analysis tool was like "wow, yes, that's what I would've done too! You're so smart :)" because it saw a sequence of winning moves that I couldn't have predicted - and indeed, because I _didn't_ predict it, my follow up move is guaranteed to get a "...aaaand you're losing now" from the engine. Always makes me laugh when that happens 😅
I’m curious how you can cheat at chess with a vibrating butt plug.
@@SloverOfTeuth in modern chess computers are a huge component of preparation. A2/H2 are super common from computers and having a chess brain you can see these moves then begin to see situations when these games are made despite not being able to see so far into the future. Also, the randomness it injects into the game hurts prepared players. Magnus himself says he plays a young player strategy where he plays suboptimal lines to throw opponents out of their preparation.
I like her nerdy sass. "If your child was so smart then why doesnt he just spend 5 mins to do his homework and then go sleep." I feel like she been tempted say this to people directly. lol
I fucking love her.
I feel for the child with parents who live vicariously through them to make up for their own shortcomings in life.
Lol
lol
@@horscategorieyeap
Note: The "first move advantage" thing is not purely psychological, because when chess engines play chess engines, white wins more often than black (of the 0.01% of games that *aren't* draws, which, fair enough.)
She acknowledged that in the video
To be clear, a lot more than 0.01% of games are wins for one computer or the other. You're off by several orders of magnitude in fact, in the 2020 edition of the Computer Chess Championship for instance there were 26 decisive games out of 200.
The number of draws depends on how powerful the hardware is and how much time is allowed to "think" per move. The longer the time between moves, the more likely a draw is. So if you want "interesting" games between computers that don't end in a draw all you have to do is reduce the time available until they stop drawing their games. I couldn't find the rules for the Computer Chess Championship but I'm pretty sure it's set in such a way so as to minimize the number of draws instead of setting it up so as to get the "best" possible games.
@@oscarprieto9013 The Computer Chess Championships are a bit of an outlier for our purposes because in those, the engines are intentionally given specific openings to play, which artificially reduces the number of draws far below what happens when chess engines are freestyling.
Anyway, the goal here isn't to find interesting games. The goal here is to determine whether white having a higher winrate than black is caused by human psychology vs. by white being objectively (although by only a tiny amount, to be fair) stronger than black positionally.
@@tudornaconecinii3609 The advantage of playing first is actually very substantial relative to the small differences between top players.
If you look up common openings like the Ruy Lopez or Queen's Gambit on Wikipedia the diagrams should make plain why this is to anyone who knows the moves: basically the early initiative the player of the white pieces possesses can be converted to enduring structural advantages. Win rates for the first mover are much greater at high levels and significantly greater even at much more modest skill levels.
That anecdote about playing checkers in the office was infuriating, I lived through so many similar situations myself. Physics students can be quite something sometimes.
sometimes... lol
@@acmhfmggru The difference is that in tic tac toe you can just know by heart what to do in each situation. In checkers you can't. You would still need a computer. For humans it will still be an interesting game. It's like it if said in the news that someone finally solved chess and has an AI that will always win (or hold a draw or whatever). That wouldn't mean that humans playing chess with other humans suddenly becomes pointless.
@@luxshokk Some of the top chess players often play checkers casually in between games. If the top chess players with their extremely good calculation abilities still find checkers difficult enough to play and not trivial to draw, then I doubt random "I don't play checkers" guy (and "she's just bitter" over here) can easily win.
@@luxshokk Finding a game unapealing =/= being a condescending prick. Elon comment was spot on, I prefered turn based and real time strategy games over chess since forever for many reasons.
@@Derzull2468 you couldn't have said it better. You can find a game unappealing, but you don't have to be a condescending prick about it
You not mentioning the "Chess has more positions than there are atoms in the observable universe" trivia was a real subversion of expectations
Tbh you can probably say that about most games.
@@RunstarHomer yeah but chess seems (at first glance) much simpler than most games
Thats only if you count illegal boards. legal boards is a much more reasonable 10^40. Go on the other hand....
I dream of a day where people stop having this exact conversation:
1) "Chess has more positions than ___."
2) "Well that's only if you count wrong."
3) "This other game has more."
Like, come on. Please stop doing this.
How many possible positions are there for players on a football field, or for a bowling ball in a lane?
My understanding of moves "looking computer-y" is that there's a difference in risk/reward calculation between players and computers. Computers will regularly make moves that humans, even very good ones, would probably consider either too dangerous or too slow. A strong human player can see a "risky" move which could (say) sacrifice material to generate a strong attack, but they might be unable to confidently calculate forward enough to know for sure that it will work. A computer would be able to see the same move and know its not risky at all, and go for it. Conversely computers will often do things that just... *look* like a pointless waste of time, but actually aren't? A subtle slight repositioning of a piece that doesnt obviously generate attacking potential or neutralize danger, but has some strong importance far down the road. Something a human, even a very strong one, would probably consider a waste of time, failing to see the long term value that move could provide. At least this is what ive learned from watching chess streamers -- im not very good at chess myself.
Actually, sacrificing material for an advqntage can be a very human thing to do. _Taking_ material and then defending perfectly for 20 moves in a passive position until they can untangle is the kind of thing that humans try to avoid and computers love to do.
@@greenUserman that, too. Computers can see the light at the end of the tunnel in locked positional games where humans feel crushed
It's true and a lot of it lies in the justification for the moves. See, if a player was to completely change what they're doing after every single move made on the board, then they're going to find themselves trying to do an insane number of calculations. Meaning that they're very likely to make mistakes or run out of time, so you're better off playing with (or forming during the game) a solid plan on how to win and trying to make it work as best you can (all the while, trying to figure out what your opponent is plotting to do against you). This is why even grandmasters constantly study and practice various positions, so that they already know what the plan is should they find themselves in that very situation during a live match. (When you see them taking 30 minutes to make a move, that's not a great sign; you know their plan has gone awry and they're now having to improvise).
When spectating games (and especially when watching them in hindsight) you can often find the reasoning behind their moves when you consider how these moves effect those plans of theirs, as they're going to be presented with a lot of questions from their opponents during the game. "Do I want to trade these pieces?", "he's attacking my knight, how am I going to react?" etc. You'll find time and time again that players will often try to make the moves that will best aid their plans, whilst avoiding things that are counterproductive to them.
Chess bots, of course, do not work like that at all. They will happily recalculate the position after every single move that they and their opponent make, completely changing what they're doing on the fly. They'll make the best move, simply because it's the best move. They don't care if a move is gonna trash their plans. They don't even have plans. What they're doing is making a lot of calculations to figure out every possible combination of moves, 20 steps ahead or more, so that it can simply make whichever one brings them closer to winning whilst causing the least damage to their position. This lack of adherent to any sort of plan leads to some very bizarre-looking behaviour.
As an example, a real player typically wouldn't set themselves up for a flawless attack and then as everything is going to plan, scrap the whole thing at the last minute, and push a random pawn for what is (to us) seemingly no reason, because a completely different way to win in 10 moves, rather than the 11 moves your plan would have taken, has presented itself. Real humans would just go ahead with their already calculated plan of attack, it doesn't matter if it's not the absolute optimal, 100% efficient way to win, so long as you're comfortable that your plan is going to work. You're not even going to waste the time trying to figure that out. In the human's mind: The clock is ticking, and I'm winning, so let's go!
One of the key factors in Niemann's cheating was just how unreasonably perfect his play was. From recollection he made less than a third of the mistakes of the world's best players, yet kept that ridiculously high standard through every single game no matter how exhausted. It was particularly on show when it mattered, and particularly obvious in unusual or difficult situations. People just can't play every move absolutely perfectly like that, computers very much can. And that's a very key factor in play looking computery. As well as the more typical (especially to lower players) awkward moves. They will switch rapidly between playstyles and do moves no player would ever dream of because the learned experience over thousands of games says "don't do x play". But computers looking at billions of potential moves can say when that move is optimal, and doesn't have a specific playstyle beyond just doing the best move.
Well then sometimes you get rookie players beating more experienced ones because the rookies have no clue what they’re doing, which throws off the experienced guy. Suppose an experienced player had the idea to replicate this, throw in a couple of bold and risky movies that just don’t seem like the optimal choice to a human but that computers might do, that would probably throw off their opponents
3:05 I run a math tutoring business and my job these days consists mostly of talking with parents who have no idea what their child is even learning, and I cannot possibly describe how cathartic this mini-rant about Aiden's mom was.
Today was the day, that youtube actually recommended an interesting Video from a small Creator :)
Came to say same.
79th like
@@the_gammaman1st like
For me it was the String Theory video. Instantly subscribe after that and after I saw titles and thumbnails for the other videos she had.
If we like and comment we can help her grow.
As a chess lover but hopefully not a snob (I haven't solved checkers) who recently started binging your videos, I just wanted to address a couple of your kinda open questions/statements:
1. Regarding making a surprise move when someone has been studying your game: this is totally a thing! It's just called a "variation" or, if specifically prepared, well...a "prepared variation." It's less exciting though because *usually* this happens like 20 moves in, between top players.
(In a weird way chess sort of is solved, for like, the first move. And then kinda less solved for the 2nd move. And much less solved for the 3rd move. And so on. It's all theory and "book" [memorized] moves for crazy good GMs until a variation happens).
2. The first move advantage is definitely a real thing: Playing as white is sort of like being...the home team? You are "expected" to go for the win, get maximum points for your tournament game, while when playing as black, players often intentionally try to play for the draw. Here is why:
There's a concept called "tempo" - if playing with tempo, it just means, the other player needs to "respond" to your moves. You are in control and "asking questions of your opponent." You are like hey I am attacking this piece, what will you do now? Getting this "tempo" concept down is really key - I have tutored kids at chess quite a bit and I introduce the idea pretty early on.
Well, with mediocre players like me, this doesn't make a huge difference. But at the very top level of the game, the basic idea is, white starts with tempo, and if everyone keeps playing super accurately, they can maintain their tempo. Even though playing 1. e4 looks innocuous enough, you are actually "attacking" both the empty c5 and d5 squares. You are saying to black, "these are mine now, what are you going to do about it?" So a very good, very aggressive player, can maintain this tempo throughout the opening and into the later stages of the game, which is why white does have a first move advantage.
Anyways, longer than I intended. Kinda rambly. Hope it made sense. I really like your channel, keep up the great content! =)
Even for mediocre players (I say this as one) white can be a big advantage because you're going to play an opening you know, whereas f you're black you might need a defense against something you've never seen and just have to reason through it.
One of Magnus's positions that he said in an interview is that the level of competition is so stiff, and the difference between winning and losing is so close, that all a player might need is someone to have a binary signal that "something about this turn has a good move available" and that would be enough to turn the game on its head. Like, Niemann didn't need to have a full computer engine available to him, but rather he'd just need someone to cough in the audience on turn 55 to know that something needs a closer look.
Resigning because you suspect somebody cheated with a single cough and no concrete evidence is still pretty extreme. I feel like a normal response would be to play it out and examine all the video footage later. In a best case scenario he would be able to flex that he won despite the cheating, in the worst case he just loses. It sounds like he made a mistake and was mentally compromised, and subconsciously wanted something to blame, and the easy scapegoat is his opponent who has been caught in the past.
@@gogokowai Yeah, he definitely got tilted, the threat of someone cheating got to him I think, to some degree. But I think the general consensus in chess is that at the highest level when you are playing black, you are going for the draw at best, not for an outright win. So this was a very surprising result.
It almost sounds as if he was excessively paranoid...I guess being at the top gets to you sometimes
@@gogokowai some important context:
- Carlsen didn't resign that game or the next game, he played this game out and decided to pull out of the tournament after it. Then in a different tournament he had to play against Niemann again and decided to just not play that one game, resigning after 1 move because he couldn't just forfeit that single game otherwise.
- The decision of Carlsen to pull out of the tournament came only the next day, after having reviewed the full game with his team, and after having contacted the tournament organization about his suspicion and after being told by them that nothing could be done about it and there would be no investigation, but they would start doing the absolute basics to prevent cheating for the rest of the tournament (metal detectors, radio wave detectors)
- The "single cough" thing is not a suggestion of how Niemann would have cheated, it is just to show how little a top player needs to get a significant advantage in a game. At the top level, if there is a winning move, and you tell a superGM there is a winning move, they will find it. So it's not "I think he cheated because someone coughed on move 50", it's "I think he cheated, I don't know how, but even a cough at the right time is enough for a big advantage so he's probably found a way". If you know you'll be alerted when there's a good move, that even gives you a double benefit as a top player: 1) you will find the good move and 2) you don't need to spend time trying to figure out how good the position is (which is a large part of top level chess, and is something a lot of their allotted time is spent on)
- Finally: Cheating has been an open secret in top level chess for a while. It's just extremely hard to prove someone is cheating unless you catch them literally with their pants down (Like with Igor Rausis).
@@nolifeorname5731Are you implying that a content creator would leave out details for the sake of a more cohesive narrative and that I assumed the worst and started spreading misinformation in the comment section? That would never happen.
Except in tournaments it is “touch move”. Once you touch a piece you have to move it. So if your partner wants to tell you not to move the piece he can’t give you a signal when you touch it. I never thought of this but that’s probably to circumvent cheating as you described. I realize it could still be done by say gesturing towards the piece but it wouldn’t be easy.
I'm not sure if you covered it, but iirc the deep reinforcement learning-based chess engines like AlphaZero have a distinct flavor to their moves vs. the ones with handcrafted evaluation engines like Stockfish. both will beat a human every time, but AlphaZero is known for being ruthlessly aggressive, playing dramatic sacrifices and doing a lot with pawn position. You can even tweak the engines so that they toy with their opponent, or play defensively, etc. So I think maybe there's so much room above human play for the machines that the concept of "top engine move" isn't so well defined, in terms of catching cheaters. Not that I think Niemann cheated, like you said there's no tangible evidence, just a fun fact on chess software.
Oh so we can make the computer mean or we can have it play like the terminator lol thats awesome
You might find it interesting to know that Stockfish nowadays also uses a neural network in its evaluation function, i.e. it is no longer handcrafted. It is not a deep network like that of AlphaZero and it does not need a GPU to evaluate (its uses the CPU's SSE.AVX/AVX2/AVX512 instructions).
So we worship the engine now as if it's omniscient and always producing the best moves, but we also did that for engines 10 years ago. Engines now crush the engines of yesteryear. Over the years I would expect the same to happen to the current engines. We won't know if the recommended moves are best unless chess is solved
The same with evaluation, there is no gradual evaluation of a position: Either it is winning or losing or a draw for best play. All numbers in between mean that the engine is too weak or shallow to determine.
So a game to guess the evaluation of a position would make no sense, as the engines of tomorrow could have a different evaluation.
I went through a chess phase in my 20s which lasted a couple of months. Turns out, I liked the idea of chess way more than I liked playing chess. Also, as somebody who used to give piano lessons, parents and grandparents are often terrible judges of how inherently talented their kid is at something.
Lol same. I figured out a way that worked for me to improve really quickly and then promptly lost motivation. Still find high level chess really fascinating and I like how you don't have to be very good to appreciate calculations even at the Grand Master level (unlike a lot of sports/games which often require a level of proficiency/understanding to grasp how good people in the top level really are).
@@Isaac_L.. I got way more interested in the algorithms computers programs use to play chess than in learning chess strategy myself.
@@Isaac_L..Mind sharing the way to quick improvement you found? No pressure lol I’m mostly just curious :) I barely even understand chess myself
@@민정-q3m I heard Hikaru talking about how he could calculate multiple lines at once then file them away and come back to them like tabs on a computer. So I tried replicating exactly that. I played multiple games at once (on between separate tabs) against bots that were at or slightly above my elo. I'd start with like 3 games at once, then as I got better I quickly went to like 6 or 7, then I upped the bot elo and went back to like 4 games, and so on and so on. So yeah, that's my secret: play a bunch of games at once. Then when you go back to playing one game it feels like you have all this extra memory space and processing power compared to before. I also did zero studying on openings and theory, I'm sure that if I coupled the multiple tabs/games practice with studying some actual theory I could break 2000 elo in well under a year (I started at under half that).
Maybe it was just a strategy that happened to work really well for me and other strategies tend to work better for other people, but it was some of the fastest I can remember ever improving at a skill.
I just don't have interest in playing chess any more. I've gone through phases of interest in other games but this is the only one I've felt like I found a cheat code. (In poker for instance I'm convinced to get to the professional level you need to brute force learn how to do the statistics on a dime coupled with some decision theory study). And I kinda have other hobbies I'm obsessed with which take priority over mastering games that I'm already better than the average person at (but nowhere near the level of a more serious enthusiast).
I'm glad you mentioned "The Turk" , that's a really weird piece of chess history. The guy who built that thing was a real inventor, he just built it for fun to show off at a party. Much later though, some royal wanted him to bring it back out to show his friends at another party. He was busy working on steam engines, he didn't actually want to bother with the fraudulent thing, but they were SO enamored with his "invention' that he was practically commanded to fix it up and bring it.
Chess? I'm still trying to master cheese!
I never have cheese. It's been solved.
Just cheese it, bro.
The vibrating anal beads story was the most hilarious thing of the year to me, I could picture Niemann moving slightly on his chair to better "feel the move"
It resulted in a 19 year old being singled out and having a metal detector waved over his ass in front of a world stage. It was an awful thing to suggest and it had a serious impact on public perception of him, and that's regardless of whether or not he was cheating.
@@HighFlyActionGuy Did the metal detector speak for itself tho?
I love how it started on chess streams and Reddit and then was picked up by a far right German tabloid and then ballooned from there
@@lazydroidproductions1087 "ballooned" - I see what you did here
@@ExecutionSommaire I didn’t mean nothing by it! You’ve read far too far into this!
Have you considered doing stand-up? Your delivery/demeanor is excellent and the fact that you are fucking hilarious. 🇦🇺♥
Ha I appreciate this comment because most people hate my jokes!
@@acollierastro OMG, they have no idea you're taking the piss in between jokes do they? Lmfao, the sarcastic conversational narrative is priceless. Be appreciated, consider emigration.
@@acollierastro No problem, R mean rook, K mean King, + mean check, and # mean checkmate (I actually responded in the wrong comment but heh)
I agree. Dr. Collier is very entertaining. Taking a course with her would be a blast.
@@robertvarner9519 I was thinking this when listening to her talk about doing adjunct teaching. I wish I could have taken my astro class with her as the lecturer.
"The chess speaks for itself," from Hans after his match with Magnus will always be hilarious to me.
That will always be a legendary line, no matter which way you felt about the whole situation.
Then he proceeded to lose every set after iirc. what a meme.
Yeah it’s like “I dunno I played the game and won, that’s how chess do.”
I don’t think I could come up with a more analytic answer than that if I were asked the same question. Granted I’m no chess player but still
@@WaluigiisthekingASmithHe had a great tournament in Spain. Of course the immediate aftermath of something life changing would be game changing
Thank you for defending checkers! I had an argument about how "chess was superior" and me saying that I enjoy chess and checkers was somehow unacceptable.
People who think Checkers is beneath them are losers.
Magnus was mad because he lost in an endgame as black. Practically never happens to him. It was a big deal because he's the greatest endgame player of all time. So for him to lose to someone rated below 2700 in an endgame as black sent the chess community over the edge.
Also... Hans gave some of the worst commentary ever on his chess game. So it made a lot of people suspicious...
There's no significance to him having the black pieces in this scenario.
@@karans6762 You dont know anything about chess if you believe that.
@@thechesssavage6400 What if he was playing purple vs mauve? What's your belief on that? How about if he was playing banana green vs crystal pink?
How about if chess pieces were donuts with chess symbols on them and he was playing glazed pink with sprinkles vs chocolate filled with powdered sugar?
Hint: In all those cases it is up to the player, not the paint or type of wood or pigment in the plastic. Or the glaze.
@@d3nza482 If purple went before mauve it definitely makes a difference. The opening repertoire is so different between black and white because there is a functional difference between the sides
@@d3nza482 So, you don't really know anything about what people are talking about right now, and I'm gonna clue you in. When you are talking about chess "white" and "black" are also ways of saying "the player that goes first" and "the player that goes second". Pigment is irrelevant, because no one is talking about pigment of the chess pieces when they're talking about chess.
There was also postgame interviews with Niemann where he struggled to explain his moves against Carlsen and againstbAli Reza - i.e. wasn't able to tell the interviewer the lines he saw that would have justified the moves he made
One thing that complicated the image of niemann was that in a later interview, following another game, he was not able to explain some of His best moves.
This reminds me of the cheater sleuths analyzing speedruns in gaming, where they count frames and calculate odds of certain patterns occurring. It's so much fun to learn how they do it.
Something dream something something
I wonder if the Chess Federation ever offered Niemann amnesty in return for disclosing his method. I know that happens occasionally with speedrunning as well as in motorsport like Formula One.
Magnus based his assessment not on the moves, but on Niemann looking unconcentrated during the game. Also his post game interview (where the player and a commentator analyse the game) contained some rather egregious errors. If someone plays like they're possessed by stockfish and then in the interview say they could also have done this other thing (confidently blundering a piece and the game) it makes people wonder.
but none of this matters
When I played chess competitively as a teen I remember us all agreeing that the only people who think chess players are smart are non-chess players (i.e. people don't play in a club or in tournaments). Stereotypes suggest that there could be a cultural difference between Sweden and the US on such a matter, I guess.
It's not orange Pepsi, it's orange juice in a Pepsi bottle. Magnus used to always drink OJ during games until some sport physiologist told him it was a bad idea.
An interesting fact about the OJ - I wonder is that due to sugar contant🤔
I have no clue what you are saying. And what does Sweden have to do with anything?
@@jonathanbush6197 Sweden is the country I and the other chess players mentioned am from.
I think most very good chess players in the states have such weird huge egos they think they're smart as well, especially since a lot of them have never really had to do anything else, and the public generally thinks they're smart, but people who play a little or encounter them peripherally are typically very aware that not all competence is generalizable.
You can't possibly know how spot on you are on the "amount of cheating". Cheating like Nieman has been found to cheat is so obvious that it is like the guys is screaming out loud "I am a cheater, get me". But actually, any elite player knows that they just the slight hint of " you are better in this position" twice or three times in a gamewould make them almost unbeatable. This is what makes cheating so scary at chess; smart cheating is virtually undetectable
"Is Aiden so smart, or have you been calling him so smart for his entire life, and for the first time, he's confronted with the fact that he can't get stuff right away. And maybe math is a little harder than he thought and it doesn't just come easily, and he's too scared to fail. Too scared to look at the worksheet and not understand, so he just doesn't do it."
Brutal truth. Love it.
Either brutal truth or prejudice against ADHD kids, depending on Aiden's conditions.
I knew kids who were plenty smart, but lazy. They didn't have anything wrong with them, they were just slackers who didn't want to be bothered living up to their mother's dreams for them. I know this because I myself am one. Gifted classed all through school. But I never had to study. I would read the text book during the first week of class and then go off memory the rest of the semester. I'd do homework on the bus to school. Then when it got hard and I actually needed some study habits, I didn't have them, and I didn't care enough to start trying at that point.
Her comment hit home with me because it summed up my own experience quite well. I just don't have any ambition. I like to write, and have written 3 novels, but never bothered to try and query them with an agent or even self-pub them. I get a few beta-readers, and if they like it, that's enough satisfaction for me. But I know I'm lazy and I also know her words perfectly describe people like me.@@falquicao8331
Aiden is Bakugo
@@falquicao8331 yeah i had this experience. i was put in 3rd grade math during second grade an voila! my grades were better. I wasn't a struggling perfectionist. I would just rather play with markers and crafts if you aren't teaching me anything.
it is also possible the 2nd grade teacher just didn't like me/had a weird bias against me/had a weird teaching style, and the 3rd grade teachers were grading me accurately. who knows. I was a very bullied neurodivergent little kid; i dont think i realized what school was for. it was just the place where we did fun puzzles but the people were mean and loud.
With so few comments, maybe I will be the first commenter who notes that an early version of a chess computer immediately sacrificed its queen for no reason because it had been trained on GM games where that sacrifice was frequently a winning line.
That's hilarious to me.
24:00 that's not why they think he cheated. some very good chess players have said his moves seemed unnatural (the type a computer would find but a human wouldn't), and on top of that niemann showed he couldn't explain his thought process in the post game interviews
Well they say it but apparently an algorithm specifically designed to catch cheaters doesn't think so. I believe they believe he cheated. But just because a guy looks guilty doesn't mean he is. I think the whole 'scandal' was pretty disgusting. Even if eventually it turns out he cheated they certainly accused him of doing it without actually knowing it which demonstrates some really poor judgement.
My brother was one of the students sho was too stifled by the school environment to do well...
He dropped out, immediately got his ged at like 15 then taught himself how to program before joining the airforce a few years later. Hes now by far the most financially and imo socially successful person in ny family.
Yeah that whole too bored/smart for school is a real thing.
Around 9:00, when you're discussing the magnitude of cheating, I think it's important to remember that the people that are bad at cheating are the ones that get caught.
The Loeb reference went crazy
I laughed out loud lol
The idea of him using a vibrating buttplug to cheat is so funny I choose to believe it.
It does explain his videos.
Super controversial I know, but I'll put this out there as someone diagnosed long before it was a dime a dozen diagnosis... But A.D.D. is a thing. Your comment about "Aiden is so smart, he's just bored". I 100% get your point, and some kids are just brats, but that's something I've always heard and dealt with my entire life. It's not necessarily boredom it's like an impassible brick wall that'll ruin your life unless you find your own ways to circumvent your brain. No amount of "Just knock it out in 5 mins and take a nap", It's sitting there for hours, trying your hardest to just get it done, and your brain kicking and screaming and doing it's damndest to push you in another direction. I've always heard the same thing, even now in my professional life "Oh, you're so smart, if you just applied yourself!". Oh sure, I'd love to, let me jump right into that and just knock out everything with my perfect motivation and focus. It's nuanced and complicated. Anyways, Rant over. Just throwing it out there.
As someone with ADHD, I had the exact same thought during that section. I really wish I had been diagnosed and treated for it as a kid, because it's exactly this sort of "if you're so smart why don't you just sit down and do it?" attitude from my parents and teachers that made school absolute hell for me :(
(And not to mention the anxiety and depression which is often comorbid with ADHD, and gets triggered by struggling with completing work...)
The brick wall is real.
@@miss-magic-maya I was diagnosed at 13. To put that into perspective of how long ago THAT was... I was in the last phase clinical trials of Adderall. As in, It was just barely approved for use by physicians and I had to have regular checkups and bloodwork done to make sure it wasn't screwing something up.
Anyways, To my point. Even diagnosed and on meds it doesn't make it go away. It just takes the edge off. I mean, Sure, the meds help you focus but it's up to you to focus on THE RIGHT THINGS. I take a different med now but there are millions of times I took my adderall dose and then hyperfocused on a video game. But I have done some amazingly creative stuff when my mind behaves too.
tl;dr Diagnoses and medication don't change the core issue
Same. With adhd. I would test really well but never had care nor motivation to do daily work as it didn't interest me. It always manifested as a lack of motivation for that which isnt fun rather than being slow or dumb. Sucks fr. Luckily I always enjoyed tests elsewise id have never pass High school. FWIW I had excellent grades in AP classes and did extremely poorly in my middle range classes, because AP classes were interesting and so i chose them. Wish I stayed on my meds
I've had this problem my whole life. Never went to a doctor about it unfortunately.
This problem still halts me from time to time. My undergrad years were very good but it took me 18 days to apply to one grad school because I just couldn't get myself to do it.
"There's no such thing as a bad buzz." -- Hans Niemann, probably
i love that you put your game at the end. that is the most gangster display of humility i’ve ever seen. you are a treasure lollll
Even if perfectly play by both sides would result in a draw (which, I agree, seems most likely), it's still possible for white to have a real, non-psychological advantage due to having the first move. Note, for instance, that opening theory is highly asymmetrical between white and black - there are many openings in which white is able to make a strong attack and seek an early checkmate, whereas the strategic considerations for black usually focus on equalizing first and then setting up a counterattack. I suspect that if one looked at the set of games featuring only moves that a reasonably strong chess player would be able to calculate as "good moves", more of those games end up as a win for white than as a win for black.
Sorry, immediately after writing that I realized it might read rather like "man explains things to women". I didn't mean it that way; I was just musing to myself about whether the first move advantage is real or psychological, and if the former, in what way it could be real.
@@stevenklinden As I understand it, the modern chess computers learn by "AI". The computer just plays itself countless times, and slowly improves. If that's the case, I would assume you could look at the results, and see if white has a real advantage. I assume it does, but very small.
If perfect play from the opening position is a draw, then technically there are only psychological advantages.
@@ronald3836 Only on a very broad understanding of the word "psychological". Since the players aren't perfect calculating machines, there can be, for instance, achieving a position in which it is - objectively - more difficult for your opponent to calculate optimal play is advantageous, and it seems a bit misleading to me to call that merely a psychological advantage.
@@stevenklinden I agree that psychology only really comes into play when you are close to perfect play. E.g. a computer with a 32-men endgame database would have to be taught to distinguish moves that easy for moves from moves that are difficult. Otherwise the computer might play a series of drawing moves that even an amateur chess player might be able to draw against.
re: the first move advantage thing.. general rule of thumb is that in the opening of a chess game, white is playing for an advantage while black is playing for equality. you can see this reflected in the way that a lot of theoretically established openings for white have the word "attack" in their names (the keres attack, the english attack) but never "defense," while for black it's exactly the opposite (sicilian defense, king's indian defense etc, but no "attacks"). white's advantage is very important and real on some level, but it exists in an interesting sort of dissonance with your 100% correct hunch that chess is basically definitely a draw.
“Math was a little harder than he thought, and he’s just too scared to look at the worksheet and not understand so he just doesn’t do it, and so he’s failing.”
I’ll admit it, that cut super deep.
This was my entire high school calculus experience
I think a nuance that might be missed here is that when chess players are this good, any sort of indication is an amazing, amazing help to them. Sometimes, there are these incredibly hard to spot errors that the other person does. Since it's so much about pattern recognizing, it means that a blunder that one player at the top misses, the other might miss to, because it's happening somewhere obscured to them. If Niemann's plug vibrated once when it was a crucial move, that would likely be enough for any of these players to find the right move. The knowledge that "there's only one good move, and it's winning" just does something ephemeral at that level, to where they'll just 99% of the time find the right move if they have that knowledge, as compared to maybe 20% if not.
One thing important to mention in how a chess engine works is evaluation functions and search. Right, when the board is in some position, the computer will give it a score, like "this position is +0.837," or "that position is -1.936."
These evaluation functions do not have to be very good. In fact, the best computer engines often use fairly cheap and easy to calculate functions and then they search the game tree in an order that roughly looks like "best first." For complicated reasons of memory efficiency it's actually a form of branch and bound search in most cases. It plays for both sides until some depth is reached or it looks really bad for one side. If it does, that branch is abandoned because obviously neither side wants to play bad moves if they can play good moves instead. Effectively it propogates the acceptable limits of cutoffs down the tree and keeps evaluating positions for every possible move.
So the computer doesn't actually know better than you about a position. It just thinks about millions of promising future positions faster than you could and lets that information decide what move it should make now.
There is however another kind of engine that has recently become prominent. Neural network engines. This is where the engine uses some giant mess of linear algebra to learn how it plays when it's thinking really deeply, and uses it to train it to play like that in one move. Then this can be used to train an even bigger mess of linear algebra to play like that one did at really high search depth but in one move.
These neural networks are far more complex and slow to compute than the handcrafted, relatively simple evaluation functions of traditional engines, and it's harder to know if they actually "understand" chess or not, but the benefit is that their evaluation function without even thinking multiple moves deep is already master level, and they can still search down the tree to improve on that.
Finally, it's off topic, but there are LLMs. Basically, ChatGPT can help you cheat at chess, but it's more like the level of a good human player than a godlike incomprehensible superbeing the way the others are.
It's also worth noting that when it comes to Chat GPT, as with a lot of things, it will also occasionally (and confidently) produce something that's an outright fabrication. An entirely invalid move, like moving a pawn three spaces. Chat GPT doesn't seem to have the strongest grasp on the actual rules of the game and will occasionally attempt to break them unless you plug it into software that locks it into viable moves.
@@JaMaAuWright I think "strongest grasp" is an overstatement. Chat GPT doesn't know the rules of chess at all. It doesn't even know what a rule is. What it does is things that look similar to chess moves.
Chat GPT wouldn't help you cheat at chess, it's considerably worse at it than even basic computers from 20 years ago.
honestly, reaching the endgame without blundering means you’re actually p good at chess
Strong chess players (GMs , IMs, National Masters like myself) regard physicists in such high regard but some professionals that I know hold excellent chess players in very high regard too. A funny anecdote is that when I had a major (medical) operation back in 2007, the doctor was flabbergasted that he was performing an operation on a chess master. I've been using chess engines a long time, since the mid 1980s when they were much weaker. It took awhile before desktop computers were supplied with multiple microprocessors. Once that happened, the combination of fast hardware with good software made chess engines much stronger. I have not been able to handle them since around 2005 or so, when old Fritz 6 and Fritz 7 were state of the art, rated maybe FIDE 2700 or so. Nowadays a recent version of stockfish running on some kind of a monstrous piece of hardware is a little over 3600 strength. It could spot Magnus Carlsen a pawn and a match played at such odds would be fairly close.
Here's a funny story about a chess player and a physicist. Albert Einstein and Emmanuel Lasker (the second ever world chess champion) were good friends and they would visit each other to talk all the time. The problem was that Lasker only ever wanted to talk about physics while they were together, while Einstein only ever wanted to talk about chess.
@@michaellisinski2822 Wilhelm Steinitz paid a visit to Paul Morphy down in New Orleans at least one time that I know of, a few years after Morphy had quit playing chess competitively.. However, Morphy agreed to meet with Steinitz only on the condition that they did not talk about chess or play chess. I don't know much more about that incident but it struck me as rather funny. Steinitz was perhaps the strongest player in the world at the time with the possible exception of Paul Morphy.
Even Carlsen would get destroyed in a match vs a strong engine if Carlsen had only pawn odds. For the match to be close Carlsen would need a minor piece odds.
Super interesting your long trains of thought that seem unstructured at first, but they are the opposite. You've gotten yourself a new subscriber!
"Tic Tac Toe is a solved game so I don't want to play it."
"Well if you're so good at tic tac toe why wouldn't you just want to play it and win?"
...
you can't win tho. unless you play with a monkey
Checkers is different, juat because its solved doesnt mean you can memorise the algorithm
Actually the best way to cheat is to hide a mirror on the ceiling so that you can peak at your opponents pieces
I want to mention that there was this iconic series where Kasparov did best Alphago when chess engines where just about getting to the point where it would be impossible for players to beat but not yet to the point there was no hope at all.
There are a lot of video essays on it. It was last stand of humanity vibes.
Finally a video by Angela where I can go "hmmm. I wouldn't have said it like that..."
FYI, we play tic tac toe wrong. It was originally 3 mans morris (from Rome I believe). Where you each only get 3 pieces and once played you move them around trying to get 3 in a row.
The game later grew to 6 and 9 man morris which is the currant popular game. Though I have heard of 12 and 15 though they sound impossible to play.
This is becoming one of my favorite channels, always awesome stuff!
The argument about how he is young and has not lived in a world without chess bots is actually such a good point! Also the match in the end is hilarious😂
In tic tac toe you can be doomed already from the second piece. First guy places middle, If the second one places in one of the middle squares(not the corners) you have a forced win for the first player... This is the only "strategy" I ever found in tic tac toe, after a lot of searching through bored days in my youth. I did not explore fully other starting positions, as they were obviously inferior, but I believe I fully explored starting center. But it is kind of fun being able do declare your victory on the opponents first placement!
The best first move actually isn't in the middle, it's a corner. Oddly enough.
@@silphv I think this is what I remember.. I solved 3x3 tic-tac-toe sitting in the car when I was 10.. Now I have to do it again.
@@silphv the corner is maybe more likely to trick a careless opponent (or a child) but I don't know if you can can call it better, since either move is a trivial forced draw as long as both sides play sensibly
@@dm9910 That's just how the game is. If both players play optimally it's a draw. Which includes starting in the corner. I dunno, look it up, it's not really an open question.
@@silphv I'm not disagreeing with the claim that going in the corner is *an* optimal move. You claimed that the corner is a better first move than the middle. I'm disagreeing, on the basis that they are both optimal moves.
First move advantage is absolutely a thing. It's like the Gawain and the green knight where they agree to take turns chopping of each other's heads. Having the first swing is a big deal:
1. Mathematically, you've had more moves than your opponent, unless you make a mistake and lose a tempo. This means the first-move advantage would decrease the longer the game goes on, except
2. The 1st move stops certain moves from your opponent. You took the best move, which in many cases means they can't take the move they wanted to play. So you can maintain disproportionate control through to the midgame.
3. Moving first means you have more control over the opening you play. If I know you're a Sicilian player, I can play a queen pawn game that you're not as comfortable with. Or at lower levels, the first move advantage has more room to take the other player into a weird opening they never played before.
I still think chess will be a draw when it gets solved, but there's definitely 1st move advantage. Even in tic-tac-toe, there's very clear 1st move advantage. If the game is played imperfectly and doesn't end in a draw, it's almost always the 1st player who wins.
There's actually a very important factor you seemingly missed in your research. The main way that people think Hans cheated who actually know what they're talking about isn't engine help but a prep leak - he seems to have studied a very specific, off-beat line the night before, which Magnus happened to play into. It's astronomically unlikely unless someone had leaked to him that Magnus was planning to do that.
This is interesting! Do you think that counts as cheating? I don’t know if I do. If Magnus is discussing his strategy pregame and Hans studies it, I mean.
@@acollierastro It's definitely cheating. Prep is not casual discussion, it's intensive, computer-aided study of specific lines, often with anonymous partners using burner accounts. If you know what lines your opponent is preparing and they don't know anyone knows, you've essentially violated the only information inequality that can exist in an open-information game. The point is, you had to have done something untoward to even *get* that information.
I think this is what he SAID he did. If you don't believe that he had prepared this very specific line, then you believe he is lying about it. Why would he lie... (did he get the best move from a computer instead of from doing his homework the night before)?
@@Pablo360able I disagree - prep spying is not against the rules of the tournament so it isn't cheating. I mean, it's definitely scummy behavior, but it's not the same as using a computer during the game.
And yeah, if you had to hack a website or something to get it you might be breaking some rules / laws. But if it's just that you get one of Magnus' team to talk to you, that's not cheating.
That said, prep leak is pretty unlikely, Magnus hasn't stayed on top of the world in chess without having an incredibly airtight team. And really deep prep is maaaybe 20 moves, and the game went for 57, with as others have mentioned, Niemann outplaying Carlsen's notoriously strong endgame. Carlsen doesn't have some magic prep for endgames that you could get from a leak, he just calculates really accurately.
@@Pablo360ablethat alone would not be cheating. it’s a players responsibility to keep their prep private
that is really fascinating. first video of your i saw, was about crackpots in physics, and the second - the one when you've totally put yourself into their shoes (without aggression part)!))
as far as I understand it the first move advantage essentially means that white gets to decide a whole lot about the opening of the game because black is always catching up. also another sign of cheating you didn't mention (or maybe I wasn't paying attention in that case my bad) is that cheaters will take an unusual amount of time (both unusually long or unusually short) to make their move. In online chess cheaters will most often take the time it takes to input their opponent's move into an engine and get the best move out to play their move, no matter what move. Whereas sometimes a move is obvious and you would expect someone (at a certain level of course) to play it instantly, or sometimes it's hard to spot and you would expect them to think about it for a while. So when they take roughly the same amount of time to play every move this is also a possible sign of cheating
Uhm I seem to have typed out a whole ass essay in your comments. Sorry
What people mean when they say that Hans played like a bot is not quite how you understood it. Basically there are times when you can have a move that puts you at a short-term disadvantage but eventually pays off 10 15 20 years in the future. The thing is, outside of very forcing sequences there's a cognitive limit to how much even the best chess players can see ahead. Stockfish might be willing to sacrifice a rook because it can see that 20 moves later it can win a queen. That kind of move is one that a human could never play because they would have to not only see the idea that remember takes 20 moves to materialized, they would also have to see the response every adjustment possible by the other player, and see that there's no way for the other player to get out of it, and be incredibly confident in their assessment because if they're wrong then they basically just lose. So if a human plays that move and is able to follow up on it, then that means they're cheating because they've seen something that it is literally cognitively impossible for any human to see. The best players in the world can only see 4 or 5 moves ahead outside of special circumstances.
I was at the first tournament where a computer beat an active grandmaster. It was in Long Beach, in the late 80s -- Deep Thought vs Bent Larsen. Bent tried to play into complications against the machine, and it just out-calculated him. Mikhail Tal was at that tournament, but he signed the "NC" (no-computer) list, meaning he didn't want to play any machines.
We all understood that history had just been made, and it was interesting (and in some ways a bit depressing) because we knew where the trend was going -- we just didn't know when we'd get there.
All that said, I don't think modern computers have "solved" chess as far as humans are concerned. They have become excellent coaches and training partners, and that will continue. But humans still have to put in the work.
I made it briefly over Elo 2200 both in USCF and FIDE in 1999, and then life got in the way and those numbers came down. Finally made a return to competitive chess last year after being out for 20 years... and found myself in a sea of young people (sometimes -really- young) who could freaking PLAY. They were 2-3 categories better than their ratings, all because they had been studying and playing online during covid, and now they were showing up and very hungry. It's an interesting (if painful) time to get back into the swing of things.
besides the history of cheating i think the other reason ppl believe niemann cheated even though there isnt solid proof of it is just that hes so unlikable lol. dude is just annoying
Also, Magnus' PR. Starting with the name. Shame his parents weren't more creative. Calling him Optimus Prime or Megatron would have made it all so much more entertaining.
From what I understand magnus played a specific variation that he only played like one or two times before and somehow hans was prepared for that specific one, which according to some high level chess players(Hikaru and Gotham chess who are content creators so there's some possible alternative motive there)i s extremely unlikely to have happened. Also there's this whole controversy with han's mentor also being some sort of cheater but i'm not sure about that one.
When I was studying computer programming, one of my first programs was a Tic-Tac-Toe game. I was able to write an algorithm such that the computer would either win or draw, but never lose. This was not a brute force algorithm as Angela seems to be suggesting, but rather an abstracted procedure which guaranteed a non-losing outcome.
This is extremely common. Basically ever computer programming degree will have this excercise or similar.
On the "My son is so smart, he's just bored. That's why he's failing maths" note. This is a flag for either undiagnosed neurodivergence or shitty parenting. Usually both.
Felt personally attacked, sorry.
Sacrificing pieces in a crazy way can be surprisingly effective even when you don't know what you're doing. I'm terrible at chess, I have ADHD I can barely think half a move ahead and always miss obvious stuff; I tend to get confused by all the pieces so I will often deliberately (or more often accidentally) sacrifice a piece just to get fewer pieces on the board to worry about. And 20 years ago I played an older teacher who was decent at chess in an informal school tournament; I lost obviously; but because I kept sacrificing pieces including a queen for a queen and a bunch of other pieces; he was really thrown off because none of his usual strategies worked since he'd unexpectedly lost important pieces (because a normal player would at least at the time never pointlessly sacrifice a queen that way) early on and he could not rely on me being rational for any move, the game went on and on for much longer than reasonable when playing such a noob as me.
Just to further prove how terrible I'm at chess.. These days I barely even remember the rules for how the pieces move; but 20-ish years I at least knew that. And once I played a woman at an after party (so to be fair we were both drunk) who had to learn how the pieces move before playing, so I logically should have had an advantage since I had played the game tens of times before while she had never played. After a long while I suddenly noticed she was moving a piece (probably a pawn) such that she she was putting her own king in check, which obviously is not allowed; so I protested and she moved the piece back. And that's when I noticed she was still in check; with at least two pieces.... and not only that but my king was also in check with at least one of her pieces; and we had both no idea for how many turns we'd both been in check. So we decided to give up on the game... Is that a draw.... or is there another word for this?
@@SteinGauslaaStrindhaug It's probably more of a drunk than a draw.
I would like to quote Imaqtpie here: "If you don't know what the fuck you are doing, how is the enemy supposed to know?"
See, you frame that as a success story, but you did in fact lose. All you did, at the very most, was draw out your loss. Pointless sacrifice can never actually give you an advantage unless you make your opponent run out the clock.
I was that child who refused to do homework. I did the work in class, and aced the exams, but when I got home, I just didn't want to think about it any more. This wasn't because I thought I was so much smarter than anyone else, but because my school was so slow and behind in the curriculum. I had already learned everything they were teaching from my older brother, years prior.
There were so many other things that I could spend my time doing rather than something I had already done many times before.
Back in the late 80's, I had a Chess program for the Commodore 64. The program was actually printed in the pages of a magazine (in machine code), and I sat for hours typing it, and saving it to a tape (as in audio cassette) drive.
The principle was that each piece had a 'value', based on its importance. A pawn would have a value of 1, a bishop might have a 5, a queen 10, and the king a 1000 (an 'impossibly high' value, because the whole game hinges on not losing him). The game had 10 levels. On level 1, the program would look at the current positions on the board and consider the score after every possible move, as well as every possible response. This took about 2 seconds. At level 5, it considered every possible move, through 5 iterations of moves and responses, trying to maximize its attack score, and minimize the potential losses. Because of the limited processor speed, you could literally prepare and eat a sandwich, between each move, above about level 7, once the pieces were developed to the point where more than one or two of them had attack/defend options.
As things progressed, chess engines were loaded with databases containing every tournament chess game ever recorded, along with the calculated win/loss statistics for every move, for practically every configuration of the pieces on the board. One basic advantage is that, the computer should never ever blunder (lose a piece needlessly, because it didn't consider the opponent's very next move, or next few moves). So, the program knows the best moves, by the best players, in every game that has ever been recorded... and, because of today's processing power and speed, it can weigh every possible move and response for 10, 20, 30 moves into the future. If the meat-based player gets creative and does something wholly unusual (as related to the database), it can recalculate and 'preview' the new scenario, in milliseconds. I'm sure there are layers of even more clever concepts baked in, as it's been a while since I've thought very deeply about any of this.
My dad told me a story about when he had the opportunity to play against some chess grand-master (I forget the name). My dad was an amateur, and used some very unconventional moves in the early game. According to his story, my dad's non-standard opening baffled this grand-master, who then stood up and threw the board across the room, shouting "Why must I let this idiot beat me".
I'm not sure how much of that story was true, but my dad was a sailor. So . . . grain of salt and all that. But the point is that someone who is so well trained in the "proper way" to play a given game, just might be thrown off balance by someone who has no idea what they're doing.
Seems unlikely.
If their levels were closer, perhaps, but a GM isn't going to be rattled by sub-optimal moves in the opening from someone well below their level.
Instead, it's actually the other way around - the GM is going to know exactly how to punish the suboptimal play and cruise to victory.
Now, two GMs against each other and one of them can surprise the other with computer preparation - in that case, it's usually one deviation from the established main and side lines, and the advantage is largely in time gained as the opponent is unfamiliar with it and has to think longer while the other can blitz out the move.
This is why, in these cases, the GM "defending" will often play moves that complicate the position and make it unclear, to get the other GM out of prep so they are back on even terms and thinking on their own.
your dad is a liar lol
Not true. It is a story attributed to Nimzowitch
Not even a little bit plausible. A novice playing a weird opening might throw off someone with more a little experience than the novice, but a GM would not have any trouble destroying them. The only way I could see if is if it were like a blindfold simul and the GM was playing 20 boards and just lost track of your dad's. Even then, the whole throwing the board thing sounds made up.
Did you see Wargames? The Matthew Broderick movie from the '80s? They actually forced the computer to play tic tac toe against itself to illustrate that some games are unwinnable and should not be played, or at least are of this type and you know that going in. You get to see the computer learn this. Obviously this is a movie and it's all fake but it was still pretty neat, especially in an era when people teaching computers had learned on punch cards
I think typically when someone says they play like a computer instead of a human it means more like they played a move that makes the position super complicated. Like people like positions that are simple and the moves being intuitive. Sometimes computers find a line where you have to keep making the best move over and over again which is super hard for humans to spot. I don't really think Hans did that this game, but whatever. Also I heard another story of people cheating and it was so clever, but they got caught. So basically they had one guy watching the game and they had a system where depending on where their friend sat in the venue they would know what move to play. I think they got caught because they figured out that he kept glancing away at the board.
There's a reason why a subject matter specialist's opinion on average is more valuable than that of an outside observer. To make an analogy, imagine if Nature magazine was sending Physics papers to famous violinists for peer review. Some important details would certainly be missed in such exchange, wouldn't they?
Going to the bathroom and having a phone in there has actually been done by a grandmaster in a tournament. I think more than once.
Wow what you said about math being scary when it start getting hard is so real that stopped me before but you’re really helping me get into this stuff as an adult
The drama was quite interesting. It sparkle quite intensive disscusion, what should be considered proof of cheating, which methods are reliable, a lot of analysis (even fake one)...so yes, the drama was fun. Anyway I think he didn't cheat in that one specific game. magnus either had just a bad day or he already suspect hans of cheating and it affect his gameplay, because he just played poorly (for his standart).
In the 1970s, there was an article in Scientific American (I think it was Martin Gardner's column) on making a simple computer using matchboxes and M&Ms to play tic-tac-toe.
Hey, lay off Aiden ok, he just has ADHD.
In a human game part of the first move advantage is you can practise from after the first move while the opponent has to either predict your first move for practice from before the first move.
Magnus just played poorly against Hans. It is easy to cheat at chess online but very difficult to do it in person at a FIDE sancioned event. Poor sportsmanship by Magnus to accuse his opponent of cheating with no evidence
Magnus is a memory machine.
Possibly....but Hans has cheated, admitted to cheating, and been shown to have cheated more than he admitted. So maybe Magnus is a poor sport...but Hans is a terrible person. He should not be allowed to compete at the highest level anymore since he has no ability to demonstrate he "isn't" cheating....and we all know, you can't prove negatives.
@@trepidati0n533 has he cheated outside of online games?
@trepidati0n533 The point is, no matter how well respected a player is. They themselves are not the best impartial judge on whether they were cheated against or not. This applies to Garrett Adelstein and Magnus Carlsen
I'd really like to see you do a similar video about the KataGo AI and the researchers who effectively broke it.
also i dont think its so much that chess players thought only a human could win at chess bc it required some kind of human passion that a computer couldn't emulate, just that no one could have imagined what computers would become. like you said chess computers used to suck, programming chess used to be this really big challenge so it prob seemed unlikely to a lot of people that it could be done well. it is pretty surreal how far weve come and how rapidly technology has improved lol
Not so sure about that, I think even if someone understood conceptually computers playing chess and always getting better, they might still entertain a bit of superstition about the "human element". I guess I was reminded of the AlphaGo documentary (about the first AI to beat a Go champion) where there was a lot of that sentiment going around. Of course Go is a very different (and even significantly more complex) game than chess, where trying to think X moves ahead like a decision tree just isn't really how you play well, because there are far too many possible moves. So in that case it is actually true that it requires some human element-the AI did win but it was trained on a massive data set of human-played games. Still, even with chess I think it is a really tempting (hubristic) thought.
@@silphv that's less that a human element is required, and more that it's much easier to program that way; in theory, it would absolutely be possible to train an AI on AI only to always play perfectly, we just don't have the processing power yet to do that in a reasonable amount of time. Hell, if it's a matrioshka brain running for a million years it might even actually solve it.
@@nerdyspinosaurid Actually no you're completely right, I read more about it and, while the AlphaGo that was featured in the documentary was trained on human-played games at first (before playing a bunch of games against itself to improve), they've since made multiple newer versions including one that taught itself without ever seeing a human-played game and which was many times stronger of a player than the previous models. As of several years ago even it's entirely untouchable by human players, so I think they just stopped there.
The game of Nim is very simple and is solved, but its the _second_ player who can always win.
The thing that's always confused me and rubbed me the wrong way about this scandal is the fact that Magnus's evidence for Hans cheating was that... he lost the game. I don't mean to draw a direct comparison in all accounts when I say this, but try to think to the last time you've seen this reaction from someone after they lost in a multiplayer game. Say, for instance, maybe someone who takes Smash Bros. a little too seriously got salty and accused you of cheating reflexively. Something like that. Now think back to the last 10 times you've seen that accusation made. 50. 100. Every time you've ever seen or heard anyone lose at a multiplayer game and accuse their opponent(s) of cheating. Regardless of whether or not they're right, has there ever been a time in your life you've ever seen this accusation being made where it *hasn't* been transparently petty and a symptom of being a whiny entitled baby? Because I, personally, have never seen someone lose a game and accuse the winner of cheating that wasn't an obvious attempt to handwave their own performance by claiming an unfair disadvantage. I have no reason to believe Magnus's accusation was the sole exception to this observation.
But that's all just preamble, here's the actual problem I have with this: why even play the game, then? Essentially Magnus is saying that because Hans was not expected to win, the fact that he did is therefore indicative of cheating. Really think about that for a second, because if that's the case, then it's not possible for Hans to have won. And if you're thinking "yeah well he just wasn't as good so of course he didn't," then... so it was pre-decided for him that he should lose? If the only acceptable outcome is for Hans to lose, then why is he even playing it? We know he doesn't like playing games when he knows he won't win, so what gives? So now the inverse has to be true, if the only way for Hans to win is if he cheats, then the only way he would even decide to PLAY the game is if he started cheating, right? So now it's a foregone conclusion that as soon as Hans sits down, he's not only already lost, but he's already cheating. There is no point in going through with the game anymore, just give Magnus his trophy, deliver it straight to his home and save him the trouble of having to play the game or even show up at all.
And since Magnus is so great at chess and always wins forever all the time... well, just keep doing that. If any other player in a tournament with a lower power level is up against him, that player's only options are to forfeit or cheat, and since cheating isn't, y'know, allowed, the coordinators of the tournament should just call the match before it happens, not even allow it to happen, and send Magnus his trophy in the mail. Meanwhile, Hans will now realize that they can never join another tournament so long as a single other opponent has more than 5 million points. But then when he does so, the tournament will just send HIM the trophy in the mail, right? No games need to be played if he just stomps out the little leagues. But then, who would ever join a tournament if they already knew solely based on experience points who the winner is going to be? Surely they would all pull out the second Hans, or especially Magnus, joins, right? So... why... why have tournaments at all anymore? How do you convince someone to ever play chess professionally ever again if every single match can be determined simply by who has the highest level? Is chess an actual game or is it just a pissing contest?
Obviously that's a slippery slope, and I... well I assume that most professional tournaments have match ups between people with near enough battle ranks that it's more open, and I also assume that the level of disparity between Magnus and Hans was unusually large for a matchup at this level of play. But even so, that doesn't make the "I lost so you must be cheating" accusation any less childish nor does it make the paradigm of "the only way you can beat a level 99 chessmancer is to cheat" sensible or tenable in any capacity when it comes to the infrastructure that the idea of a tournament is based on. And I say this regardless of whether or not he cheated, it doesn't matter if he did, as long as your standard of evidence is "well I'm so good that if you beat me it's only because you're using hax" then like... fuck you? No? That's not how games work? That's not how anything works?
p.s. Kids don't like doing homework. I hated it. I struggled really hard in grade school and it wasn't until I got into college that I actually started having "good grades" as it were. I've also seen two children both piss and moan about doing homework when they'd rather be playing video games, one aced their stuff when I finally nailed them down to do it, the other was pretty clearly not understanding the material, but they had an exactly equal enthusiasm for putting up with the mindless busywork hellscape that is grade school. The only appreciable difference between my experience as a child and my experience with children is that I wanted to play Primal Rage and they wanted to play Minecraft.
The thing about first move advantage is that chess isn't solved. So, most people agree that if chess was solved it would most likely be a draw ("a chess game, when played to perfection, ends in a draw" is a real saying, not just a meme). But, as long as it isn't, it's been observed that most non-draw games end in white winning. Even for engine games, which are thousands of time better than the best human players.
I think it's interesting that you can program an algorithm to play various levels of less than perfect. Like the program knows the "right" move, but chooses to do a "wrong" move every so often depending on how low you set the "level". I wonder how they validate that a "level 6" delivers just the right amount of "bad' moves at the right frequency.
The low lvl ones I've played just make a few blunders in the beginning then play better as time goes on.
It's not fascinating at all
Every algo has the candidate moves rank list. It simply chooses to play something other than #1 move.
If say Stockfish decides to play #2 candidate move throughout the entire game, it would still beat the sht out of Carlsen. And it would proly do it without losing a single game out of 100.
OTOH if the algo goes for #8, it would lose against me LOL I promise you that.
Fascinating part would be, as you correctly wonder, rating the series of such handicapped moves with correct ELO. Which in fact they DONT.
There is an attempt of doing just that with Leela version called Maia, but it's not all that.
It depends on the algorithm, but most of them are built around searching to a certain "depth" (read: number of moves ahead) to find the best move. So instead of purposefully picking a bad move, they can just do a less extensive search which would result in more natural bad moves similar to what a human would do.
I agree. And a surprisingly hard programing problem is making them bad in a human-like way.
There's things that human players are more likely to miss, that programmed-down chess computers don't accurately mimic. For example, less-good humans have board-blindness, a tendency to focus on a small area of the board and miss things going on further away in space and time. This makes humans more likely to, for example, miss a capture a bishop can make all the way across the board when that bishop was last moved a long time ago than they are to miss an exchange capture a pawn can make (like you just took my piece, my piece was defended, even as bad player I'm likely to notice I can take back). Chess computers don't tend to act this way, so they will miss obvious moves a human player below the computer's intended level will catch. And in the other direction, because they are evaluating moves by how relatively bad they are, they will play moves that are better than the intended level to avoid really bad outcomes like mate. Low level chess computers almost never get caught by a surprise-mate, something that happens to low level players far more often.
@@camipco yes, precisely. I really like your examples
In response to the "Aiden rant" at 3:05, I do want say that the description you give could genuinely apply to someone with ADHD/ASD such as myself. I also know a decent few kids in middle/high school who fit the description you gave, but I for example in 8th grade would refuse to do assigned algebra work because I was bored with it, instead doing calculus problems on my own. I certainly wouldn't say it's the fault of the teacher to assume that a kid who doesn't demonstate proficiency doesn't have it, but I would say there is nonetheless a fault in the Western educational framework in which proof of knowledge is valued so much more than the actual learning itself. I know that there has to be some metric by which to judge student aptitude for practical purposes, but I think the present heirarchical model of learning is at minimum not well suited in particular to autistic people in ways that can lead us to disengage in ways similar to what you described here. I'm not detracting from the point being made because the people you describe do exist, but I would like to add that an additional category of person exists for whom the first half of the description may fit but not the other.
Magnus is arguably the BEST player ever, it's at the point where even the GMs are like "yeah, he's the best". You can almost count on one hand how many times Magnus has lost playing white, in classical games. Niemann couldn't even analyses his game AFTER having just played it. He said that the winning move was one he saw in another game... that game never happened...
You don't play chess, do you? The debate about the non-existent game was about the opening line, not the winning move. And there turned out to be a game where Carlsen had reached the opening line via transposition, so this is not conclusive. It would be a lucky guess for Niemann, as he said himself, to prepare for that line with a different move order, but possible given the previous game.
@@eljanrimsa5843 Oh, hi. We making personal assumptions about each other for no reason? Cool. "You don't wipe your butt after you poop". Cool. Great way to start this off.
It WASN'T about the opening line, it was the mid game where there was a move made that was an improvement... but sure, Niemann somehow looked up that specific game that morning, but also didn't remember when or where that game was played... sure...
Hey, was it fun watching someone rated 300 points lower than Niemann out analyse him after the game? Basically everything Niemann said was wrong, like OBJECTIVELY wrong. But, hey, cool.
@@aliince9372 but you don't "see" a winning move in another game. that's not how chess preparation works, that's why I thought (and think) you may not be familiar with the process
@@eljanrimsa5843 Yeah, that's fine. The game could/would have been analysed after. ...what's your point exactly?
@@aliince9372 I believe that their point was that it doesn't seem like you know as much about this topic as you seem to think you know, and that make be affecting your judgement on the matter. I'm not all that experienced with chess, so I'm not sure if that's actually true or not, but generally speaking things _tend_ not to be anywhere near as absolute as people on the internet make it seem. I'd wager that probably applies here.
3:40 - as someone who was thought to be "bored" in high school, I can answer this one. I was neither bored nor struggling to understand. I just didn't care. When I got to college and was paying money out of my pocket, I did very well because I was motivated to care.
I dropped out of high school because I was miserably bored in every class to the point where I could barely focus on anything and felt like every assignment was busywork. At least in my district, they seemed to repeat the same material year after year, and I just couldn't handle it. I ended up going to community college at like 17-18 and blowing through everything with a 3.8 GPA before moving on to an actual university. I wasn't dumb -- which is what my teachers thought and legit tried to put me in special education before I dropped out -- I just couldn't focus on anything due to the lack of novelty.
My mother was the mother in your story, but y'know, sometimes it is the school's fault. My folks argued for advanced placement or skipping grades or anything else they could think of, but the faculty at both my middle and high schools thought I was a moron because I skipped class and had terrible grades.
Yeah but that's survivor bias
Are you now a successful academic or professional, or are you a pot smoker?
@@ronald3836 Did you just start immediately drooling and smashing your keyboard without finishing the paragraph? I'm a dermatology PA. Right now I make a little over 100k and only work four days a week. It's pretty good.
If high school had been more accommodating, more reactive to my needs -- who knows? Maybe I'd have been prepared and comfortable enough to go into a STEM field like I dreamed of when I was a kid, but in reality I scrambled for a degree that was relatively affordable, had reliable job prospects, and paid well.
I don't think -- and, again, this is in the context of what I experienced in my school district -- that being able to put your nose to the grindstone and monotonously complete very similar assignments perfectly year after year is the same thing as being intelligent. The kids who got scholarships at my school were all socially inept, had the personalities of bricks. They could barely hold a conversation and ten years on, from what I've seen through social networks, didn't do much of anything with their free ride through an Ivy league.
@@garr123 no, my paragraph was finished! I was just wondering about this one little piece of missing information.
Well done 👍 👍 👍
The thing that always bugged me, personality wise, is that first-turn surrender. What's the rationalization behind that? What could a person possibly see on the first turn that makes them change their mind about the game? Or, more likely, what makes a person decide they're going to go to this game with the full intention of forfeiting immediately? Very strange behavior. I can't put myself in that mindset.
I think it was to make clear that Magnus was refusing to play on purpose. If he just doesn't show, there could be all kinds of speculation, maybe he's sick, whatever. But playing the first move demonstrates "I am fully present and able to play this chess game, but I choose not to."
There may also be a tournament-rules issue. Forfeiting by no-showing is recorded different from resigning. I don't know if that has some implication for future games / tournaments Magnus might have cared about.
Someone told me that if you resign or don't move on the first move, it doesn't register as a proper resignation or something but rather the match was aborted somehow? So Magnus playing the first move was to make sure that it didnt happen and that he was for sure resigning intentionally due to his opponent.
It was interesting to me that humans being unable to beat computers at chess was completely unremarkable, and it is today. I still can't help cling to the perspective from my youth when it was unclear if computers could beat humans at chess. It's one of those things were I feel like the world lost a little of its wonder when it was no longer even competitive.
That struck me, as well. Granted, I don't have my finger on the pulse of the chess community. But I remember the excitement back in the 1990s about a computer that might be able to beat a chess champion (for the first time). Apparently now, the computers are so good that no human can ever beat them. Crazy.
I appreciate your very reasonable take. There were thousands of people who were ready and willing to burn Niemann at the stake because he talked funny in an interview. I'm being serious. They thought how he talked, his accent, was proof that he was cheating. People are f*cking stupid. It's nice to hear someone not jump to conclusions for once. I'll defend the fact that we don't know he cheated in this game. Even though he seems like a piece of sh*t otherwise
Your point at the end I think is the reality. Niemann and his coach played a lot against the bots and with the bots that they've started to develop "bot-style" play intuitively. That kind of play could definitely throw off the other players that have trend paths worn in by humans. An engine could suggest a move that has no obvious stength for 5 moves, but if you've steeped yourself with that instinct it just "feels right".
every top player uses engines during training, though. it's nothing unique
Unfortunately, there isn't any bot-style play for a new generation of humans to "grow up with". It's pure logic and mathematics, which lets bots calculate best moves. And they are playing against the best move that you could make next, as well, so that when you don't make that best possible move, you are automatically at a disadvantage. And so, unless you can "keep up" with the bot, it gets better and better as you get worse and worse. No one can defeat this by intuition, instinct or feelings. And even grand masters do rely on these intangibles, as you can hear when they comment their games, saying "something about this move seemed wrong", even though they are calculating as best they can. You can't reach a 3500 elo by just growing up using and playing with bots. Their moves are often not understandable until the game is over simply because human brains can't think that way or that far ahead. It's not a talent that can be developed somehow, it's pure mathematical brainpower. The bots have taught chess players a lot about what works in openings and endgames, true, but those points in the game have the fewest variables and so are at least minimally comprehensible to humans. But it will take a cyborg or another leap in evolution of the human brain before we can compete with these bots or play like them.
All you can say about lichess level 3 is, it doesn't violate the rules of the game, but it violates all the rules beginners learn in the first lessons, like not letting pawns take your pieces.
what I dislike with lichess lvl3 is that it sometimes plays very well then just hangs their queen when you attack it with a pawn. It's still stronger than me
That was uber entertaining - thanks. Perfect chess by both sides ends in a Draw [not a stalemate]. Chess ratings are Gaussian - so going from say 2,600 Elo to 2,800 Elo is not a linear calculation. Lastly, in the first 10 moves of chess, there are 169,518,829,100,544,000,000,000,000,000 variations - so you can imagine the number for a 30 move game.
"Perfect chess by both sides ends in a Draw [not a stalemate]."
Like she said in the video, I feel like we would've heard about it if they solved chess
A stalemate is a draw. In some positions that would be the best move for a player to prevent losing.
@@Ken.- dude - if you watch smart yt vids like this then you must realize that saying a stalemate is a type of draw does not entail every draw is a stalemate.
@@JamesJoyce12 Dude - if you watch smart yt vids like this then you must realize that saying "perfect chess ends in a draw [not stalemate]" is not even close to being proven true
18:00 ish -- one non-null tic-tac-toe result that can arise is when you have a stroke mid-game. while there are a few downsides under this situation, strokin out mid-game has many advantages. the most reliable of these is that you no longer have to play tic-tac-toe of course but there is another related one that can occur if you are lucky which is that you dont ever have to play the game, or really any game, ever again
tbh I kind of agree with the insufferable chess nerd in your anecdote: even as a kid I thought checkers was kind of boring bc of all the forced moves
in the interest of full disclosure I should say that it has been so long since I played checkers that I do not remember the rules
I was maybe a little too young to understand the forced moves when my grandpa was teaching me, so he just started taking the pieces when I didn't do the forced moves. I still didn't learn, I just accepted that my pieces would randomly disappear.
That's interesting, because as a kid I was never taught about being forced to take pieces, or at least, it wasn't emphasized. Of course, no one as a kid would give up a capture, nor would they likely be able to see how forcing an opponent to make one might be a good strategy. And certainly once I learned chess I had no interest in checkers.
4:40 lmao "a new childish prodigy"
The Grandmaster (I think) UA-camr Gotham Chess has some interesting move-by-move commentaries on games played by Stockfish, currently one of the top chess engines. He points out a variety of “a human would never do this” moves.
He's just an International Master fyi
I'm so triggered by idea that 'first move advantage' would require the game to be fully solved to have an impact lol. I know this isn't a game design channel, while game design kinda my main 'thing', and this channel has a whole, interesting video about how it can feel to hear a smart person talking about something you're an expert in that I'm trying to keep in mind. Still made me twitch though lol.
"And this is actually my favorite part because there's more math"
Don't threaten me with a good time 😂
3:19
ADHD: "Allow me to introduce myself."
I have 2 thoughts: 1. I am instinctively disturbed by the idea that someone could be falsely accused of cheating bc they just happen to play a really good game, or because they are just really really good/develop a new way of playing. 2. If this guy has a history of cheating and admits to it, why was he allowed in this tournament in the first place? Were the games he cheated in not tournament games so they dont count?