i love speedrunning scandals because some guy will slap together a crappy fake and it will summon an investigation with half a dozen detectives spanning several months and five hundred supercomputer calculation hours resulting in a compiled document the length of a PHD thesis
The only thing at stake within these niche communities is pride and honor. Woe be unto the man who slights the pride and honor of another whose sank 3000+ hours into his hobby and lost a ranking to the first guy. (I say this with love, I have massive respect for speedrunners, it's just wild to me that someone in a community full of nerds who dedicate their free time to perfecting frame perfect inputs would not immediately see the issue with trying to slip a cheated run past those same nerds...one of them is gonna find you out, boy, especially with "speedrunner cheats" videos blowing up and general knowledge of how speedruns are cheated being spread.)
@@dwangoAC If you combine the essays that Groobo wrote to throw people off the track back then with the essays he wrote to whine about the investigation, they might be longer than your document.
It's so funny to me that there are obvious errors in the run that easily prove it's spliced/fake, and yet they still went through all that effort. It's like seeing someone in a retail store swap a price tag, and after seeing that the product name on the tag is wrong, still going through security footage, checking the chemical composition of the glue, recreating the entire store in virtual reality, and then running a thousand simulations to see the chance that the price tag could have magically moved itself to another product.
I guess they did want the maximum proof of all wrongdoing with the whole guiness thing and with how significant this game and category are, and at least they got an awesome defcon talk out of it
I feel like if they saw he was cheating from minute one they might as well see what else he did, since he obviously can't be trusted at this point and he's gonna be made a pretty obvious example out of as far as "don't try this or we'll know you did it."
The research done here will help scrutinize future runs as well, remember that all this information isn't just tossed out. Another fishy run comes up? Consult the research!
I get cheating 15 years ago. But 15 years later not being able to go “wow yeah, I cringe at that I was 16(or however old) and it was stupid, please take me off the speedrun I hope you guys don’t hate me for doing something so stupid” and moving on with your life and job and family is crazy. I did a lot of dumb stuff when I was younger and I wouldn’t die on a hill for any of it.
Gives Billy Mitchel…. If he ever reads this man just give it up and be honest… it’s been 15 years just say you cheated people will move on so quick especially if you are honest and say damn idk how I thought it was a good a idea I just wanted to be cool and make friends. Like I get it, but you must be in your 30s now don’t die on that hill.
It's absolutely insane. Insurmountable evidence through multiple methods, including just the opening screen, and yet they refuse to take accountability. I doubt there's even any accountability to take, they've probably mostly moved on with their life and yet they still cling to the fake record so tightly.
Cheating 16 years ago, being the most impressive thing you've ever done and wanting to hold onto that so badly, is truly the part that is saddest about all of this.
To be fair, it was a Guinness World Record. Two, in fact. Yeah, he cheated, but it makes sense that he's wanting to fight that hard. Just wish it was legit.
More like Another cheater cheats a wr in an easily detectable way. 15 years pass Community: WOW we that one record was actually cheated, I cannot believe it!!!
@@Roadrunner_KZSK And they spend what, 8 months on it? When it could have been dismissed after just looking at what nonsense happens. Insane what people spend their time on - on both sides.
abloogy making people less intelligent with every comment. Also, what he said goes to show how low IQ the speed run community is for missing such obvious fakery.
@@Roadrunner_KZSK everyone is innocent until proven guilty. Proven guilty. Submissions to the leaderboard are now rigorously checked and scrutinised. The community is awesome.
I'm so glad to see this comment here, was getting some real Todd Rodgers vibes from this whole thing, especially with "someone" posting the speedrun to Guinness.
Wow! You guys finally proved it! nice! We've mostly "known" it was cheated for years but we tentively kept our mouths shut without proof. Krzysiek - the former WR holder for the Hardcore categories (still for Warrior actually) openly called the run cheated due to the dungeon layout issue. He did a tonne of research into dungeon layouts and noticed that the likeliness of a seed with that stair rng seemed completely impossible.
a lot of these always seem like most of the hardcore community knows the run is cheated, but no one can get proof for years until some specific group comes along and decides to finally prove it. glad that finally happened here!
@sonicSnap Note: When I said "hardcore" I meant literally, not figuratively. The "Hardcore" speedrun category means permadeath, in which the term was officially adopted by the game series from Diablo 2 onwards (you can actually tick "hardcore" on character creation). For Diablo 1, this means saving/loading is not allowed and the dupe glitch is banned. The run is also single segment RTA.
Back in the day when I saw this speedrun, I knew it was cheated as well but without the means to actually prove it nothing could really be done. It is finally nice to see the proof though!
I remember watching it years ago and thinking the same thing. From years of playing Diablo of never seeing stairs that close to each other, I knew you could probably find a few levels of the stairs being right there but, not all of them.
@@mounne13 heck, Billy Mitchell would go the grave before admitting to using a standard replacement part, let alone actually cheating. You would need to water board him to get a admission of guilt, and even then he would claim it was a coerced confession later.
A lot of cheaters get caught because they got one tiny thing wrong. It seems like Groobo got caught because he got everything wrong. Like when the first screen of the game gives away that something's wrong, you know you've messed up.
The whole "we now have the whole source code to see how the game works" really reminds me of when criminals are identified for crimes they committed 40 years ago by technologies people at that time wouldn't even dream about.
Or how people try to debunk the Shroud of Turin as a forgery based on a medieval-era repair only to be BTFOed by modern forensics and 3D imaging technology.
@@chaos_ae it was possible even 15years ago, but the time and effort needed to do this..besides any1 who played diablo will tell you from just first levels that it was multiple runs and pasted together. The problem is if you're dedicated to prove it:)
Yeah I beat Super Mario 64 in 0.1 seconds all 120 stars. I just removed some (all) the frames between title screen and ending credits. but trust me that is the real thing
Something to note is that Guinness isn't the authority on world records in most of what they include. UA-camr Karl Jobst has been having to repeat this because of the various records they publish not lining up with the reality of the situation. This goes from Billy Mitchel's DK scores, to Todd "Mr. Activision" Rogers many fake records, and even the world of speed typing. It's possible the most they will do is discreetly remove the record from their next issue. In the same song but a different note, ZeroLenny beat a record (number of different controllers used to beat Dark Souls) but because he didn't want to pay for Guinness to verify it, they won't publish the new record, despite the evidence being uploaded.
Guinness may not be 'the authority', but the simple fact is that Guinness has sold 150 million copies of their books of records in a period since 1955. 'UA-cam Steves World Records' will be forgotten in five years time. Guinness records hold FAR more weight than any others, for the simple fact that, in 30 years time if your record hasnt been broken it will still be there for people who wernt even born when you set it to read. 98%+ of (to use the example in this video) Diablo speedrunning record websites are long defunct and the names of record holders lost to the void, especially those that predate wayback.
It’s okay to do an edited version of the run for a better viewing experience……. If you’re uploading it as a best moments compilation and not an official run
To be fair to everyone involved here, as stated on the SDA submissions at the time, this was allowable, as long as actual in-game footage isn't affected. The whole point of segmented runs being allowed is to be *allowed* to cut save/loads for best time scenarios, after all, so leaving in superfluous footage that doesn't count towards your time doesn't make any sense.
@@KopperNeoman it seems pretty inconvenient switching from mouse to touch just for one type of interaction, not to mention how minimal the time-save would be for a run that isn't hyper-optimised like Mario. if the theory is that it's used to save time for vendors, you'd have to find other instances that support it, which i assume don't exist, and the confirmed splice kinda debunks it anyway
One thing to note is that it'd my opinion MOST people who played D1 and did the dupe trick were probably doing it online where it was much MUCH easier because the network lag made it a lot simpler to time, but that's not relevant to this video. Mostly just your experience I'd imagine.
@@jaxm666 in my case, from the late 90's until the late 00's I would install Diablo 1 from disc for nostalgia reasons(I would uninstall it eventually, it was actually a lot of disk space even 10 years after release). I would always dupe items because that's part of the nostalgia for me. When I would attempt to dupe on single player after a fresh install I noticed it was extremely difficult. Eventually single player duping would become easier but I never figured out why duping after a fresh install was so hard until this video.
@@gamesareforfun i thought your comment was a bit funny. As of the disk space comment. So to comment on it i quickly went to see what the precise size of the installation is.... And it's almost 800mb... I don't remember it being that big. I remember diablo being such a great game while also being a game that took up almost no space. Why do i remember it as 1.2 mb or something? or am i confusing it with another great game from the 90's ? EDIT: ohb. Nope. LOL. my bad. I have hd mods installed atm. Makes a lot more sense. Whew. For a second i thought this was gonna be one of those annoying Mandela effects. 😅
That's some Todd Rogers-ass behavior. EDIT: wrong Todd. Original post said Todd Howard (CEO of Bethesda), not Todd Rogers (greasy old man best known for cheating at Dragster).
Scientists in the 60's using computers with 32 KB RAM: "Someday we'll have supercomputers and solve all the world's problems" The team: "WE FINALLY GOT YOUR ASS, GROOBO!"
"Item drop manipulation, when the seeds are generated at character creation, and monsters will always drop the same loot?" "Yes." "May I see it?" "No."
I'm no diablo 1 expert, but you mentioned in the video that the numbers would turn negative after 2038, wouldn't the game just turn those negative integers positive again?
@@Someb0dy1DayThere is no built in function to turn negative integers positive. That would be extremely inconvenient. I think you are referring to integer overflow, which is a very different sort of event.
I've played many hours of Diablo back on my day, I'm 42, and I noticed most of the things that are said on the video, the music, the weird thing with the vendor. I've played a lot on version 1.00, not much internet back on the day in my country, I knew the fountain had to be red for that quest to pop out and there were quests that couldn't be at the same time... Incredible that this thing has been there so long without anyone noticing until now. Congratulations to the team. Amazing work.
@@malachiatkinson7245 Not Diablo 1, the game did not age well. The gameplay style itself is hostile to comfortable movement IMO with the mouse movement and all, so I'd pick a more modern title like 2 remake or 3 and you'll know if you like the formula or not pretty quickly.
@@xravenx24fe garbage take, Diablo 1 is a timeless and challenging game. I'd recommend it to anyone for its atmosphere alone. Diablo 3 is loot farming slop.
@@youdonegoofed Cause recommending someone a game from the 90s who's unfamiliar with PC gaming is a sane, logically informed decision, right? Why are you gaslighting me like I'm illogical?
It's like some Shroud of Turin-type shit. Nobody who's ever tried to prove it as a hoax has been able to explain what recent forensic tech has pulled up from it.
honestly the mods of the speed run community just impress me more than the speed runners. They went through ALL of that to show this run was BS despite the fact they could/should have dismissed it the moment they noticed the discrepancy between the first screens. Kudos to them, they went above and beyond.
I, for one, am glad they thoroughly rinsed this cheater and made an example of him using a scientific process (meaning repeatable and didactic in this scenario). It proved beyond any doubt that this petty monke was full of bananas.
Part of it is that I'm not a mod, I just wanted to help make a Diablo TAS and I was dismayed when it was evident how much of a chilling effect this run had on speedrunning the game
@@dwangoAC Given that you're one of the guys who got F-Zero GX TASes playing back on real hardware via the use of a GBA as the input device - or did you get replays playing back that way? I forgor... (seriously though, any reason why more GCN/Wii TASes aren't verifiable via GBA? XD Even Smash Wii U and some Switch games could be eligible!)
the amount of detective work was crazy overkill - i was already on your side after the version stuff. But it's always telling when someone drip-feeds information "I was told I can use multiple runs" into "well I researched a lot of RNG mechanics" into "yes i used an editor to set my character but not cheating" into "weird - diablo died quicker than i expected" etc
Having source code to something not always means having knowledge and mastery to use. I run into a lot of people in my work that are confusing these two things and bringing others around them down because of it…
@@zawarudo1041 Bro, the source code literally spells out how the game drop items, loot tables, affecting variables, etc Its the whole ingredient list Or you would rather listen to 100 of hours of testing, theory crafting and spreadsheet, etc? (aka recreating from witnessing)
@@KenBladehartI am a software engineer. Having the source code to a piece of software does not automatically mean an engineer knows how it works, _especially_ if it’s buggy. It requires many hours of investigation. Some engineers don’t even understand how software they have written works. That said, I generally agree that source code _can_ produce concrete answers, especially if the engineers have put in many hours attempting to understand it as it seems they have here. Just pointing out that having source code is not an automatic leap to understanding a program.
What on earth is the point of splicing together the opening screens from different versions? Surely it's easier and more consistent capture it all from the same version, ideally the one you are supposed to be using.
Groobo submitted multiple Diablo runs to SDA, gradually improving his times, and all his previous runs appear to have been recorded using 1.09. I suspect he used already-recorded footage from a 1.09 run and lazily spliced the main menu from 1.00 merely to show that he is using a version before 1.07, to justify how he is able to use the negative health glitch. Still, I think it would have been easier to just rerecord the whole menuing segment instead of splicing so I'm not sure I'll ever truly understand.
The thing is, if he hadn't done a bad job at cheating, we wouldn't be talking about him as a cheater. I'm sure there are lots of other cheaters like him, which put more effort into their runs and we'll never know. This effect even has a name. It's called the "survivorship bias", although in this particular case it would be the opposite, because we notice the ones who didn't survive.
Damn. So this man really went in and created multiple different save files for seeds that had favorable layouts but then also used level/game manipulation for item drops. Big L
It says something about how much Diablo digs itself into your brain that, even after not playing for decades and not paying full attention to the screen, those missing frames were driving me nuts, muttering "ain't no way you get that many staircases next to each other without RNG throwing Ogden's Sign at you"
Speedrunners are hilarious people. Tryharding this hard editting the run is one thing but the guys rebuilding a whole ass dungeon and comparing it to seeds among other crazy checks for the love of the game is insane
I find the investigation into the cheated runs and the methodology for cheating a run often far far far more interesting than speedruns themselves honestly.
Glad to see Diablo being covered. The only game I played more as a teen was Morrowind. Hopefully you'll do a Morrowind video soon, that would be fascinating. I'm sure there is all kinds of cheating in runs with it.
Yeah, they are hilariously stupid. In the unpatched CD you can make a potion strong enough to be immune to physical damage, then nab the Icarian Scroll guy as he spawns. Then you can then just grab Keening without a fight, equip it a bunch of times to get unlimited speed, just look up Waz Morrowind, they broke it down to a science and beat it _below 3 minutes!_. Even doing it legit is pretty quick though because the first minute has the same beats (steal platter to sell, gobble pots, fly, you are just avoiding the speed glitch), If you think about it the majority of your time is spent in combat, talking, then in menus, _so skipping all that..._
"Abandon your foolish quest. All that awaits you is the wrath of my TASers! You are too late to save the speedrun. Now you will join it, in Hell." - Archbishop Lazarus, probably.
Really nice video, I'm glad people are documenting these herculean efforts to verify runs. This is the sort of thing that would normally never be seen by people outside of a specific community. 6:40 I want to clarify a small thing a little bit. Not saying you are wrong, but some background context. Back in the day, segmented runs weren't just a thing purely because longer videos were difficult to upload. Segmented runs were very common in the early days of speedrunning, in part because that was the standard for Doom and Quake. Normal runs were called "single segment" runs. SDA was initially a Quake-only site before it started accepting other games in 2006, and the single segment run for Quake was not considered prestigious or important-being part of a segmented run was. Basically, for any longer game, it was considered fairly obvious that there would be a segmented run. Over time, for a variety of reasons, "single segment" became the standard, and these days we don't even use the term anymore. Segmented runs still exist, but mainly for things like FPS games or racing games that have discrete level boundaries (such as Half-Life and Portal), although these are called IL/individual level runs in that case. In fact, segmented runs were kind of problematic in a way, when you think about it. There were no rules that stated exactly where the segments should start and end. So a runner who decides to cut a game into 40 segments instead of 20 would have a much easier time getting a record, and this was legal per SDA rules at the time. In practice, people would usually reuse the same segments as the previous record holder to keep things fair, and because their segment boundaries would usually make sense, but doing things single segment definitely makes things a lot less complicated. 23:10 By the way, this is just my own personal opinion, but I find it *extremely* unlikely that Mike Uyama would have given Groobo permission to use save files from totally different playthroughs for his segmented run. I know he says he doesn't remember, but there's just no way. 0% chance. Two reasons: one, it goes completely against the spirit of the rules, and runs were rejected by verifiers for doing the same, as you point out. Doesn't matter if it explicitly says so in the rules or not. Runs always needed to be consistent and contiguous across segments. Two, if he genuinely had gotten special permission, there's no way this would not have been documented for future runners to make use of the same benefit, because it totally changes the difficulty level of the run.
I remember Unreal segmented run on sda used some footage from different playthroughs, but they were deemed inconsequential enough so this was allowed, still felt like an exception but not unheard of.
The cheating is often times not an issue of skill and is more often an issue of RNG. The problem with speed runs is that there are often times multiple people with the skill necessary to pull off the speed run but the limiting factor is not skill its RNG. These runs become so heavily RNG dependent that your just repeatedly doing the same thing thousands upon thousands of times if not more just for the RNG to line up in a way that there is even a possibility of even hitting some type of record.
@@kylesmith987 With the scanning tool we build it's now possible to take out a lot of the RNG since you can locate a good seed and set your system clock so that you get that specific game.
I love how it was instantly proven from the goddamn title screen he cheated, but the investigating group decided to go "well while we know he did it what ELSE did he cheat?"
@@thehoodedteddy1335Not only that. But it's an opportunity to publish full case studies about the game providing knowledge for future investigations. It's also good for attracting attention to the scene through people like Aby and Karl
Let's face it ... the analytic skills of the MODs is INSANE. What a master piece. You could thank Groobo for cheating so the MODs are able to shine in such a way. Groobo should just confess at this point - but it is not needed anymore.
So in the end, a bunch of nerds got together and did ungodly work and calculations, proved that the Guinness Book of World Records team doesn't check enough, and outted a 16 year old kid that tried to hold on to a lie for too long. I mean when everyone is in their teens they never try to do something bigger than themselves or lie a little too much. Kid did something wrong back in the day, nerds feel good for finding it. Hope the kid has grown up to be a reasonable person.
When I was a kid in the 80s, if someone claimed they beat a game in a certain amount of time, the only tools we had were, "Really?", "You swear?". Usually that was enough and the person would confess to lying or exaggerating their feat. Now there are whole groups, tools, and access to the game's programs, but it takes people a longer amount of time to admit to cheating. I miss the simpler days. 😂
Even ignoring health regen, the odds of 19 hits in a row doing max, or very close to max damage are unfathomably low. I estimate 1 in 1978419655660313589123979 (19^19).
Given how pseudorandom RNGs (especially LCGs) work, it's probably not just improbable, but actually impossible. Diablo uses 32 bit PRNG, so it has exactly 2^32 states. It outputs every single one of those 4.2 billion numbers once in a sequence. The seed value determines where in that sequence you start. Because of this, the generator can not output every possible sequence of numbers.
@@stanimir4197 I give it 99% chance it being implemented wrong using the modulo operator (because I've been programming since the 90s and that's how we all learned it, since we didn't know any better), which doesn't result in equal distribution. If it is done with modulo, LCGs often have patters in their low order bits (which unsurprisingly are exactly the important bits when sampling with modulo) which makes it extra unlikely that a long sequence of max rolls would exist somewhere in the random sequence. Even today, if the game is made in C or C++ (and doesn't use something like Unreal Engine), I would still wager a 50/50 chance that it uses modulo. Other languages have better standard libraries which usually gives you a proper RNG API to use instead of having to roll your own, so those usually have correct ranged random function.
@@tylisirn That was my point - it's like just mod. Or even mod 100 (for simple percentages). For instance civ4 uses mod (no rejection retries)... since the source code is available.
No one ever thought this “record” was legit. The stair placements were so mathematically unlikely that it was essentially impossible. Still an interesting video though!
No one every proved it to SDA so the record stood there for 15 years, now legit runners can compete for the first place again. Will be interesting to see how it evolves :)
@@Baddaby And that's the weirdest thing about this whole thing. Obvious mistakes, obvious improbabilities, and this trick passed. I bet it was about money, which points to corruption among judges and modders.
Keep returning and re-watching your videos over and over, this is one of my favourites because it covers a game I have nostalgic ties to. It's so interesting to take apart various features, to see how they work. Playing games I never think how intriguing the processes in the background could be. Thank you for your work! Also, take a (back)shot each time Abyssoft mentions seed.
From the first time I ever saw that run (years ago), I never took it seriously, I thought it was more like a proof of concept of what could be done. It's clear that there's waaaaay too many cuts. I had no idea that people actually accepted that as a real run.
This is why I would never even consider cheating in a speedrun, if I was even talented enough to do speedruns that is, which I'm not. But I mean look at this shit, this investigation is simply amazing, I love watching stuff like this, it's so wild to me the amount of talent that's out there, crazy.
It's funny watching this speedrun again. Seeing it get torn apart is cathartic. Even back then, I knew it was complete bs. Anyone who played the game would be able to tell you how unfathomably unlikely it would be for all of the exits to spawn right next to the entrances. This run in specific lead me to doubt any seemingly impossible gaming feats, like no-hit runs in Dark Souls. Even if it was legit, it's so ridiculous that it becomes less impressive than something more believable.
Very interesting! 3 things immediately struck me when I first saw this speed run many years ago: the missing ring, adrias refresh and the fountain quest. the number of fireballs agains diablo seemed really low to me but that was only a rough gut feeling on my side. Since it got recognized as WR i didnt think much of it and/or brushed it off as not being knowledgeable enough. But now it seems like a gross oversight by the people responsible for verifying the run. When a moderately passionate diablo player like me noticed this, it surely should have been spotted by them.
To be fair it seems like a lot of people suspected this run had to be faked from how improbably lucky it was. But at the same time, getting improbably lucky is one of the core parts of speedrunning any game that involves RNG in level generation, loot drops, etc. Hunting for the perfect seed or grinding for that one perfect "one in a million" run where everything goes right is just part of how speedrunning works. So in a competition where luck is just as important as skill you can't reject someone JUST for being "too" lucky. You need real definitive hard proof they've cheated, and back in the day the communities were smaller, there wasn't nearly as good forensic tools, and the internet in general was just a more trusting place.
Considering it originates more from drunkenly boasting to your mates at the pub than it does from actually keeping legit records... definitely best not to take it too seriously.
Guiness was never founded to be an accurate authority or useful reference. It was always intended to serve as an arbitrary authority to settle disputes at a moment. The Internet has rendered the entire organization obsolete. Getting a record is not about performing a feat, it is about paying them and adding the false pretense of authority that they bring while performing the feat.
Was never into the D1 speedrunning myself - but I always assumed most runs would start with the same predetermined seed. I saw several speedrunners stream, and they mostly had close stairs for the first many levels. I will however say, that if Funkmastermp was part of the investigation, then I have 100 % confidence in all details and issues listed.
It's been a huge challenge just to find seeds that are favorable for speedruns so a Set Seed category wasn't particularly viable. If you were watching streams recently, however, then they may have been using seeds that we picked out for the TAS using our scanning tool. Funkmastermp did a few Set Seed runs on Twitch in the days leading up to the DEFCON talk.
The amount of dedication to this investigation is mind-boggling. Also it is a forensic masterpiece. This is beautiful in itself and if that is the only outcome of this sordid tale, that alone makes it worth it. I genuinely am in awe.
The absolute amount of effort that went into the forensic analysis of this run is bonkers. Massive props to the dedicated team who volunteered so much time and compute power to expose a cheated run that was more than 15 years old!
Same, I've never done a speed run, I probably play games slower than most people even, but either watching parts of real speed runs or stuff like this is pretty interesting lol...
I have never played this game, and my knowledge is limited. I have to say I absolutely love this kind of work from the community. From people who love this game. I'll never understood why people lie about records being set because that's not the purpose of why we play these beautiful video games. Abyssoft keep up the great content
What a detective story! It feels like at some point it became very clear that groobo was cheating, but "detectives" continued investigation just because it is fun.
Something about well narrated and very technical videos about subjects i almost dont care about just hits different when im supposed to be sleeping. I played Diablo once a bit on PS2 when i was a kid, i dont honestly care about speedrunning unless its about a game i really like (and even then i´ll watch one or two and be done)... but this is like an incandescent lightbulb to my tired moth brain.
So patch notes back then weren't great, and they may not have explicitly tried to remove music from the character creation screen, but another change they did doo may have caused the music to stop playing; that's what the team told me when I asked, but we know for a fact the music was removed, whether intentional or not.
@@kphoenix137 Shout outs to my favorite one from Hellfire; it adds a new weapon modifier that makes your weapon stronger, but gives it a chance to duplicate the enemy. This is not supposed to work on unique enemies, but my mom once duplicated Diablo himself.
@Abyssoft thanks for the response. I guess I'm not too surprised to hear Blizzard coding has always been somewhat rough, but yeah, that all makes sense
Jesus the analysis of this guy's cheating is DEEEEEP. Don't fuck with the speedrunning community. Don't fuck with the Diablo community. DEFINITELY don't fuck with the Diablo speedrunning community.
Something that stands out to me is the ridiculous amount of time and effort put in by multiple people to make this report. The level of dedication is absolutely wild.
RNG Manipulation is quite a common thing for any and all games which has one or more rng systems. One of the most known, is well Pokémon, and Minecraft. But you will find for example Darkest Dungeon, games like Vampire Survivor, Bounty of One, Dwarf Fortress and more. These can be manipulations on example drop rates for items, spawn rates of items (Like the invisible chest that gives you a Zodiac weapon which can be further manipulated to give you exactly the weapon you wan by 100%, when this chest has a normal 0.01% chance of spawning: Per New Game), and many, many, more things. So it is possible that he could have gotten access to the seed, edited it, and or used a program outside of the game, which manipulated code inside the game as it runs. This is a great video, I like watching these specially when they are well made and showcase how it all works.
The Team: "You claim theres a way to Manip item drops despite saying before that you knew it can't be done, mind backing that up with a source?" Groobo: "My source is I made it the fuck up."
Didn't he literally open his inventory and unequip his old staff before the new one even dropped? His entire chain of actions seemed like he already knew which enemy would drop it when or did I miss anything that mentions it in the video?
It makes sense for him to do that if he had been telling the truth, because he claimed on the original video to be able to manipulate RNG to guarantee item drops. We know now that the claim could not have been true, but at the time the claim was made, it was at least feasible
@@fafdsfr To this day no one has found a way to manip item drops. His claim was always false. He only made that claim because people (rightfully so) would point out how impossibly unlikely it is to have a game seed where all 15 stairs are perfectly next to each other AND him to get the exact 2 items needed for the health glitch in that same seed. "I found a way" was a way to dismiss them out of hand, and he never did explain the supposed method that he "discovered". Ever.
@@fafdsfr Another thing to keep in mind is that it was a segmented run, which was allowed, so if we pretend the staff drop was legit, then if he ran that segment even once before he knows that monster is going to drop that staff, so the pre-dropping of his charged bolt staff isn't suspicious by itself. What's more suspicious is the other stuff, like how that floor's seed isn't possible, and seeds that lead to the same layout don't include that monster dropping the Naj's Puzzler staff.
I wonder what would happen if you sneakily invited a buncha cheaters to speedrun some game they have in common and add a cash price on top. I would love to see the sort of shenanigans they get up to. But anyways that was unrelated. Pretty cool to see diablo covered though! that's really unexpected.
Um... What? "Acrossed" is pronounced "acrost" by everyone. No-one has said pronounced it *"acrossèd"* in centuries. The E became silent, and the D is pronounced as a D after voiced sounds, and is pronounced as a T after voiceless sounds.
Hey; I've been playing this game for close to 30 years, never got close to beating it in 3 minutes, my best time was between 12-45 minutes and that's with items donated from other characters. I initially thought the WR run was based on RNG manipulation via some kind of glitch, given the stairs were all right by each other; I didn't suspect cheating on the onset but given the sheer statistical improbability of getting 14 entrance-exit pairs in a row, we all should have been a little sus from the getgo. Furthermore, in my 30 years of playing Diablo, I have only found Naj's Puzzler TWICE. That's like 2 in 140,000 games. Getting that staff randomly on L10 from the very first enemy makes no god damned sense. Amazing work on the investigative team, and crazy TAS tools.
I'd like to express my heartfelt thanks to these legendary guys for investigating this run. This run bothered me from the first moment I saw it, especially because of the perfect alignment of dungeon starts/exits. And also a big thank to you, Abyssoft, for making this great video exposé!
22:21 They don't say they used a 32 core and 140 core system. What they're saying isn't entirely clear actually - My initial parsing was they used a total of 32 systems, each with 140 cores. But that's a pretty weird configuration too.
Ajenbo's first language isn't English, and what he meant to say was clarified to me when I was revising my script, so what I narrate iin the video is correct.
To clarify, AJenbo has a 32-core workstation. kevans provided the 140-core server. These are the systems where we did the majority of our most time-consuming scans.
@@Abyssoft I was also super tired at the time of writing :P I'm fully capable of saying nonsense in multiple languages, native or not. The initial message was meant as "systems with between 32 and 140 core".
I remember having seen that speedrun vid like a decade ago and immediately dismissing it as fake just from what I figured was the sheer impossibility of getting the staircases that conveniently placed every single floor. That was only the tip of the iceberg apparently lmao.
Click this link to get a BIG discount on your new favourite Magical Rogue! ↓↓↓
click.fan/Magicraft_Abyssoft
Damn noita looks different then I remember (and I honestly like that more friendly game decided to adopt spell building system)
Ok
i got the game a year ago bro
This game is amazing. I streamed it and had a blast. It's like Binding of Isaac with a lot of wands and spells and relics. Adds a bit more depth. ⌛
I have 155 hours in Magicraft - it's pretty fun and there are some interesting speedrunning builds, too
i love speedrunning scandals because some guy will slap together a crappy fake and it will summon an investigation with half a dozen detectives spanning several months and five hundred supercomputer calculation hours resulting in a compiled document the length of a PHD thesis
The only thing at stake within these niche communities is pride and honor.
Woe be unto the man who slights the pride and honor of another whose sank 3000+ hours into his hobby and lost a ranking to the first guy.
(I say this with love, I have massive respect for speedrunners, it's just wild to me that someone in a community full of nerds who dedicate their free time to perfecting frame perfect inputs would not immediately see the issue with trying to slip a cheated run past those same nerds...one of them is gonna find you out, boy, especially with "speedrunner cheats" videos blowing up and general knowledge of how speedruns are cheated being spread.)
In fairness, when I started assembling all the notes into the analysis document I wasn't expecting it to become thesis length
@@dwangoAC If you combine the essays that Groobo wrote to throw people off the track back then with the essays he wrote to whine about the investigation, they might be longer than your document.
"You scientifically suck at this game. Here's a 157 page document proving it"
digital forensics is fun
The very first screen?! So is that a new "Getting Caught" speedrun world record? 🤔
no, minecraft speedruns can be caught just as fast
Then you might have to give that one to Billy Mitchell too, with his "WR" that loads in the first screen exactly like an emulator
@@dknf4036 Can't say that don't want to get sued
@@maxb148 I'll do it. Not only is Billy Mitchell a cheater, he touched my willy in 1989 and got all my classmates addicted to Pall Mall cigarettes.
@@maxb148 Sued for what, having an opinion? You can say whatever the fuck you like you're not a public figure, and even belief isn't a crime.
It's so funny to me that there are obvious errors in the run that easily prove it's spliced/fake, and yet they still went through all that effort. It's like seeing someone in a retail store swap a price tag, and after seeing that the product name on the tag is wrong, still going through security footage, checking the chemical composition of the glue, recreating the entire store in virtual reality, and then running a thousand simulations to see the chance that the price tag could have magically moved itself to another product.
I guess they did want the maximum proof of all wrongdoing with the whole guiness thing and with how significant this game and category are, and at least they got an awesome defcon talk out of it
I feel like if they saw he was cheating from minute one they might as well see what else he did, since he obviously can't be trusted at this point and he's gonna be made a pretty obvious example out of as far as "don't try this or we'll know you did it."
The research done here will help scrutinize future runs as well, remember that all this information isn't just tossed out. Another fishy run comes up? Consult the research!
That's a great analogy 😆
@@DimumoutoThe research also helps the next cheat proof his cheated run against protection.
I get cheating 15 years ago. But 15 years later not being able to go “wow yeah, I cringe at that I was 16(or however old) and it was stupid, please take me off the speedrun I hope you guys don’t hate me for doing something so stupid” and moving on with your life and job and family is crazy. I did a lot of dumb stuff when I was younger and I wouldn’t die on a hill for any of it.
Gives Billy Mitchel…. If he ever reads this man just give it up and be honest… it’s been 15 years just say you cheated people will move on so quick especially if you are honest and say damn idk how I thought it was a good a idea I just wanted to be cool and make friends. Like I get it, but you must be in your 30s now don’t die on that hill.
It's absolutely insane. Insurmountable evidence through multiple methods, including just the opening screen, and yet they refuse to take accountability. I doubt there's even any accountability to take, they've probably mostly moved on with their life and yet they still cling to the fake record so tightly.
Exactly. Doing dumb shit as a kid is normal, not acknowledging it was dumb and that you regret it as an adult is just cringe.
Yea but when that’s the only thing you’ve ever done in your life
@@TheBritCollkarl jobst lawsuit is almost over and once it concludes it's gonna be juicy
Cheating 16 years ago, being the most impressive thing you've ever done and wanting to hold onto that so badly, is truly the part that is saddest about all of this.
yup
Do you know they to affirm this 😅
pathetic level: Tallarico
Billy Mitchell and Todd Togers have entered the chat.
To be fair, it was a Guinness World Record. Two, in fact. Yeah, he cheated, but it makes sense that he's wanting to fight that hard. Just wish it was legit.
Speedrunning community: you cheated
Groobo: but I haven't even started playing yet!
Speedrunning community: we know
More like
Another cheater cheats a wr in an easily detectable way.
15 years pass
Community: WOW we that one record was actually cheated, I cannot believe it!!!
@@Roadrunner_KZSK And they spend what, 8 months on it? When it could have been dismissed after just looking at what nonsense happens. Insane what people spend their time on - on both sides.
abloogy making people less intelligent with every comment.
Also, what he said goes to show how low IQ the speed run community is for missing such obvious fakery.
@@Roadrunner_KZSK everyone is innocent until proven guilty. Proven guilty. Submissions to the leaderboard are now rigorously checked and scrutinised. The community is awesome.
@@leocurious9919 where there is competition, there is always a cheater.
Can't believe he started the dungeon generation in second gear. What an amazing player.
He learned it from Todd Todgers
I'm so glad to see this comment here, was getting some real Todd Rodgers vibes from this whole thing, especially with "someone" posting the speedrun to Guinness.
It's obvious that the judges are not taking the "human element" into consideration when calling this run fake
He had a really good gaming chair.
That's the part that still blows my mind that I managed to stumble into both of these situations unwittingly
Wow! You guys finally proved it! nice!
We've mostly "known" it was cheated for years but we tentively kept our mouths shut without proof. Krzysiek - the former WR holder for the Hardcore categories (still for Warrior actually) openly called the run cheated due to the dungeon layout issue. He did a tonne of research into dungeon layouts and noticed that the likeliness of a seed with that stair rng seemed completely impossible.
a lot of these always seem like most of the hardcore community knows the run is cheated, but no one can get proof for years until some specific group comes along and decides to finally prove it. glad that finally happened here!
@sonicSnap Note: When I said "hardcore" I meant literally, not figuratively. The "Hardcore" speedrun category means permadeath, in which the term was officially adopted by the game series from Diablo 2 onwards (you can actually tick "hardcore" on character creation). For Diablo 1, this means saving/loading is not allowed and the dupe glitch is banned. The run is also single segment RTA.
Back in the day when I saw this speedrun, I knew it was cheated as well but without the means to actually prove it nothing could really be done. It is finally nice to see the proof though!
I remember watching it years ago and thinking the same thing. From years of playing Diablo of never seeing stairs that close to each other, I knew you could probably find a few levels of the stairs being right there but, not all of them.
@@DarkTenka yeah, i know! i wanted to say something where i was like not the hardcore CATEGORY but i didn't :(
Honestly if I had cheated nearly 16 years ago, I think I would just instantly come out like “oh yeah, that old thing? Yeah I cheated”.
Tommy Tallarico vibes
Billy Mitchell
@@mounne13 heck, Billy Mitchell would go the grave before admitting to using a standard replacement part, let alone actually cheating. You would need to water board him to get a admission of guilt, and even then he would claim it was a coerced confession later.
Even if it were recorded in the Guinness Book of World Records?
@@JonesCrimson yeah. I use to love those books and still think they’re pretty cool but these days, they may cross my mind once a year, if even.
A lot of cheaters get caught because they got one tiny thing wrong. It seems like Groobo got caught because he got everything wrong. Like when the first screen of the game gives away that something's wrong, you know you've messed up.
If you're going to put in a lot of effort covering up issues, expect to create ten-fold the issues that you're covering up by doing that.
Guinness as a world record organization is laughable at this point
They've always been a novelty book company.
Guinness nowdays is a scam.
Yeah they're a grift
@@xethifiednot true, now they're ALSO a PR firm for dictatorships
You know everything they do is BS, and it has never been a good source of factual information regarding records?
The whole "we now have the whole source code to see how the game works" really reminds me of when criminals are identified for crimes they committed 40 years ago by technologies people at that time wouldn't even dream about.
😊😊
😊
Yea imagine splicing multiple runs only for someone decades later to pull up when you created each character down to the second
Or how people try to debunk the Shroud of Turin as a forgery based on a medieval-era repair only to be BTFOed by modern forensics and 3D imaging technology.
@@chaos_ae it was possible even 15years ago, but the time and effort needed to do this..besides any1 who played diablo will tell you from just first levels that it was multiple runs and pasted together. The problem is if you're dedicated to prove it:)
"Cut out some frames for better viewing experience" yeah right thats normal in a speedrun lol
"Cool and normal." -TheJuiceMedia
You know, the frame dip from when he adjusts in his otherworldly good gaming chair really does ruin the viewing experience
*Flashbacks to save and load floating speedruns of Half Life 2*
Yeah I beat Super Mario 64 in 0.1 seconds all 120 stars. I just removed some (all) the frames between title screen and ending credits. but trust me that is the real thing
"Wow. I'm having such a great viewing experience at Groobo cheating."
Something to note is that Guinness isn't the authority on world records in most of what they include. UA-camr Karl Jobst has been having to repeat this because of the various records they publish not lining up with the reality of the situation. This goes from Billy Mitchel's DK scores, to Todd "Mr. Activision" Rogers many fake records, and even the world of speed typing. It's possible the most they will do is discreetly remove the record from their next issue. In the same song but a different note, ZeroLenny beat a record (number of different controllers used to beat Dark Souls) but because he didn't want to pay for Guinness to verify it, they won't publish the new record, despite the evidence being uploaded.
Guinness is one show and corruption, like the entire USA.
you gotta mention Hbomberguy's work on this as well, can't forget how proud Tommy Tallarico's mom must be
They're not an authority on anything, their records hold less weight than the paper the book is printed on
Guinness may not be 'the authority', but the simple fact is that Guinness has sold 150 million copies of their books of records in a period since 1955. 'UA-cam Steves World Records' will be forgotten in five years time.
Guinness records hold FAR more weight than any others, for the simple fact that, in 30 years time if your record hasnt been broken it will still be there for people who wernt even born when you set it to read. 98%+ of (to use the example in this video) Diablo speedrunning record websites are long defunct and the names of record holders lost to the void, especially those that predate wayback.
For people who actually earn one for something meaningful, It's now just an expensive photo opportunity
As soon as he said "oh i edited the save/load so it was a better viewing experience" its like bro...DON'T EDIT IT AT ALL!
It’s okay to do an edited version of the run for a better viewing experience……. If you’re uploading it as a best moments compilation and not an official run
To be fair to everyone involved here, as stated on the SDA submissions at the time, this was allowable, as long as actual in-game footage isn't affected. The whole point of segmented runs being allowed is to be *allowed* to cut save/loads for best time scenarios, after all, so leaving in superfluous footage that doesn't count towards your time doesn't make any sense.
That's basically a smoking gun which by itself proved the whole thing was an edited fake.
9:59 its not just the music skipping, you can see that the cursor instantly locks on the vendor, skipping frames it needed to "travel" to her.
Would a touch screen input be allowed in the category? If so, that would explain cursor teleportation.
@@KopperNeoman A touchscreen in 2008? With precision input for a mouse cursor?
Do people actually do that for games that aren't similar to Osu?
@@KopperNeoman it seems pretty inconvenient switching from mouse to touch just for one type of interaction, not to mention how minimal the time-save would be for a run that isn't hyper-optimised like Mario.
if the theory is that it's used to save time for vendors, you'd have to find other instances that support it, which i assume don't exist, and the confirmed splice kinda debunks it anyway
@@KopperNeoman No
@@AceStrife wacom tablet, like the one osu players use. they;ve been around since the late 90s
The part about different versions of the game requiring different amounts of precision to dupe solves a lifelong mystery for me.
Bonus!
One thing to note is that it'd my opinion MOST people who played D1 and did the dupe trick were probably doing it online where it was much MUCH easier because the network lag made it a lot simpler to time, but that's not relevant to this video. Mostly just your experience I'd imagine.
@@jaxm666 in my case, from the late 90's until the late 00's I would install Diablo 1 from disc for nostalgia reasons(I would uninstall it eventually, it was actually a lot of disk space even 10 years after release). I would always dupe items because that's part of the nostalgia for me. When I would attempt to dupe on single player after a fresh install I noticed it was extremely difficult. Eventually single player duping would become easier but I never figured out why duping after a fresh install was so hard until this video.
@@jaxm666 not only this, you'd use a tool for it, it's easier:)
@@gamesareforfun i thought your comment was a bit funny. As of the disk space comment. So to comment on it i quickly went to see what the precise size of the installation is.... And it's almost 800mb... I don't remember it being that big. I remember diablo being such a great game while also being a game that took up almost no space. Why do i remember it as 1.2 mb or something? or am i confusing it with another great game from the 90's ?
EDIT: ohb. Nope. LOL. my bad. I have hd mods installed atm. Makes a lot more sense. Whew. For a second i thought this was gonna be one of those annoying Mandela effects. 😅
that he edited out frames of moving his mouse and frames of animations to make the record unbeatable is just next level garbage behavior
That's some Todd Rogers-ass behavior.
EDIT: wrong Todd. Original post said Todd Howard (CEO of Bethesda), not Todd Rogers (greasy old man best known for cheating at Dragster).
@@alaeriia01 It's actually Todd Howard and Activision duopoly.
@@alaeriia01 My gosh someone is obsessed with TH! Get a girlfriend or boyfriend and some fun!
@@BBRocker75wym nobody was talking about Thailand
@@BBRocker75 found Todd Howard's alt
Scientists in the 60's using computers with 32 KB RAM: "Someday we'll have supercomputers and solve all the world's problems"
The team: "WE FINALLY GOT YOUR ASS, GROOBO!"
I'm fkin dying. 🤣
Kinda weird that one chooses to speedrun a game where you can, at best, speedwalk.
Got a chuckle out of me. You earned this like.
I remember how excited I was to run in d2 when it came out
Oh, you rascal
"Item drop manipulation, when the seeds are generated at character creation, and monsters will always drop the same loot?"
"Yes."
"May I see it?"
"No."
Gotta hand it to Groobo, it was one unforgettable Speedruncheon.
🍔🔥
Well, Groobo, you're an odd fellow, but I must say you fireball a good Diablo.
Gaming forensics are so entertaining to see in action.
Now if I could only figure out how to turn it into an actual job, heh
It's because gamers actually give a shit. It's not about tearing someone down either, it's about making sure the playing field is fair for everyone.
Not forensic is it. It's archaeology at this point lol.
My dude cheated in every way he could. Kinda surprising he wasnt caught until recently it seems.
New category for impressive cheating
board of shame
Just more proof of how speed running shouldn't be taken seriously.
Guinness isn't known for their thoroughness.
@@shinrapresident7010just like your comment
Lot of the old WRs are like this because they weren't scrutinized very hard back then.
“Diablo uses a random number generator”
My God he really is evil.
This is so DUUUUMB lmao I love it
FK I'm in stitches!
Couldn't stop laughing for 5 minutes straight!
🤣
so does that mean every Diablo game started after 2038 will have the same layout?
Yes
@@Abyssoft I feel like that could lead to an interesting category (if it isn't already a thing) of speedrunning for it.
I'm no diablo 1 expert, but you mentioned in the video that the numbers would turn negative after 2038, wouldn't the game just turn those negative integers positive again?
Does that count as Hell freezing over?
@@Someb0dy1DayThere is no built in function to turn negative integers positive. That would be extremely inconvenient. I think you are referring to integer overflow, which is a very different sort of event.
I've played many hours of Diablo back on my day, I'm 42, and I noticed most of the things that are said on the video, the music, the weird thing with the vendor. I've played a lot on version 1.00, not much internet back on the day in my country, I knew the fountain had to be red for that quest to pop out and there were quests that couldn't be at the same time... Incredible that this thing has been there so long without anyone noticing until now. Congratulations to the team. Amazing work.
I've never had access to PC gaming; if I were to ever have the chance of playing it, would you recommend Diablo to someone unfamiliar with its format?
@@malachiatkinson7245 Not Diablo 1, the game did not age well. The gameplay style itself is hostile to comfortable movement IMO with the mouse movement and all, so I'd pick a more modern title like 2 remake or 3 and you'll know if you like the formula or not pretty quickly.
@@xravenx24fe garbage take, Diablo 1 is a timeless and challenging game. I'd recommend it to anyone for its atmosphere alone. Diablo 3 is loot farming slop.
@@youdonegoofed Cause recommending someone a game from the 90s who's unfamiliar with PC gaming is a sane, logically informed decision, right? Why are you gaslighting me like I'm illogical?
@@xravenx24feeveryone in the 90s was unfamiliar with PC gaming, and yet we didn't have any problems.
Bro was not expecting nerds in the future to analyze every possible seed for dungeon layouts when he put together this run back in the day 😂😂
He said as much in our conversations, heh
Never underestimate the nerds!
@@Hoacakes forreal man they really left no stone unturned debunking this run it’s impressive
It's like some Shroud of Turin-type shit. Nobody who's ever tried to prove it as a hoax has been able to explain what recent forensic tech has pulled up from it.
Nerds with much better computers.
honestly the mods of the speed run community just impress me more than the speed runners. They went through ALL of that to show this run was BS despite the fact they could/should have dismissed it the moment they noticed the discrepancy between the first screens. Kudos to them, they went above and beyond.
I, for one, am glad they thoroughly rinsed this cheater and made an example of him using a scientific process (meaning repeatable and didactic in this scenario). It proved beyond any doubt that this petty monke was full of bananas.
Part of it is that I'm not a mod, I just wanted to help make a Diablo TAS and I was dismayed when it was evident how much of a chilling effect this run had on speedrunning the game
@@dwangoAC Given that you're one of the guys who got F-Zero GX TASes playing back on real hardware via the use of a GBA as the input device - or did you get replays playing back that way? I forgor... (seriously though, any reason why more GCN/Wii TASes aren't verifiable via GBA? XD Even Smash Wii U and some Switch games could be eligible!)
The methods they did can be used to analyze future runs.
So there's that
@@phantomsnare5925 We already used it to verify a few other runs just as a way to prove our methods work.
the amount of detective work was crazy overkill - i was already on your side after the version stuff. But it's always telling when someone drip-feeds information "I was told I can use multiple runs" into "well I researched a lot of RNG mechanics" into "yes i used an editor to set my character but not cheating" into "weird - diablo died quicker than i expected" etc
>the investigators have the source code to the game
>the guy denies their method
Bro...
Having source code to something not always means having knowledge and mastery to use. I run into a lot of people in my work that are confusing these two things and bringing others around them down because of it…
@@zawarudo1041 Bro, the source code literally spells out how the game drop items, loot tables, affecting variables, etc
Its the whole ingredient list
Or you would rather listen to 100 of hours of testing, theory crafting and spreadsheet, etc? (aka recreating from witnessing)
@@KenBladehart I mean 3 of his parts are apparently from outside the actual range, I wonder how much time dude spent to splice it together
@@KenBladehartI am a software engineer. Having the source code to a piece of software does not automatically mean an engineer knows how it works, _especially_ if it’s buggy. It requires many hours of investigation. Some engineers don’t even understand how software they have written works.
That said, I generally agree that source code _can_ produce concrete answers, especially if the engineers have put in many hours attempting to understand it as it seems they have here.
Just pointing out that having source code is not an automatic leap to understanding a program.
@@RaycrowX Youre an software engineer, not a game developer?
In this case, its more about data than programming
What on earth is the point of splicing together the opening screens from different versions? Surely it's easier and more consistent capture it all from the same version, ideally the one you are supposed to be using.
If I had to guess, probably didn't care other than to get clean footage. Assuming it was his recording and not spliced from other videos.
Groobo submitted multiple Diablo runs to SDA, gradually improving his times, and all his previous runs appear to have been recorded using 1.09. I suspect he used already-recorded footage from a 1.09 run and lazily spliced the main menu from 1.00 merely to show that he is using a version before 1.07, to justify how he is able to use the negative health glitch. Still, I think it would have been easier to just rerecord the whole menuing segment instead of splicing so I'm not sure I'll ever truly understand.
The thing is, if he hadn't done a bad job at cheating, we wouldn't be talking about him as a cheater. I'm sure there are lots of other cheaters like him, which put more effort into their runs and we'll never know.
This effect even has a name. It's called the "survivorship bias", although in this particular case it would be the opposite, because we notice the ones who didn't survive.
My dumbass brain read the title as Dildo Speedrunning.
Never clicked faster.
I believe Abyss covers that speedrun on his Onlyfans lol!
😂😂😂😂 @SevenGC89 comments like this are what the internet is for lol
Damn, how longs it been you yah at down bad 😂
My favorate game is dildo 3 : reaper of ass.
@@joshybunny893that escalated quickly haha
Damn.
So this man really went in and created multiple different save files for seeds that had favorable layouts but then also used level/game manipulation for item drops.
Big L
@@nonono9194 Step aside the Onyx is here!
It says something about how much Diablo digs itself into your brain that, even after not playing for decades and not paying full attention to the screen, those missing frames were driving me nuts, muttering "ain't no way you get that many staircases next to each other without RNG throwing Ogden's Sign at you"
By the way, check out the seed used by the TAS - they get good or at least decent stairs on almost every floor EXCEPT dlvl1
@@KaneYork My favorite part is that the TAS seed would have been created in 1994.
Speedrunners are hilarious people. Tryharding this hard editting the run is one thing but the guys rebuilding a whole ass dungeon and comparing it to seeds among other crazy checks for the love of the game is insane
You can never run from Diablo vets. They know the game better than their own families.
if they catch rabies you definitely can.
Lmao
That is pretty cool that creating a TAS helped expose cheating
it's happened a few times. TAS software can be great analytical tools
"Not even death can save you from me" is engraved in my brain forever, great way to end the vid
There should be a score board for the teams figuring out these cheaters. The work they do is impressive
I find the investigation into the cheated runs and the methodology for cheating a run often far far far more interesting than speedruns themselves honestly.
16:21
Oh boy, it's Christmas, let's cheat in Diablo!
Glad to see Diablo being covered. The only game I played more as a teen was Morrowind. Hopefully you'll do a Morrowind video soon, that would be fascinating. I'm sure there is all kinds of cheating in runs with it.
It's Bugthesda, you know the drill
*"IT'S A FEATURE!"*
There's some sub 10 minute runs in full on UA-cam.
tl;dr grinding potions is busted
@@Poglavnit_Pferdefuhrer you're kidding me?! Sub 10 minute runs?!
Yeah, they are hilariously stupid. In the unpatched CD you can make a potion strong enough to be immune to physical damage, then nab the Icarian Scroll guy as he spawns. Then you can then just grab Keening without a fight, equip it a bunch of times to get unlimited speed, just look up Waz Morrowind, they broke it down to a science and beat it _below 3 minutes!_. Even doing it legit is pretty quick though because the first minute has the same beats (steal platter to sell, gobble pots, fly, you are just avoiding the speed glitch),
If you think about it the majority of your time is spent in combat, talking, then in menus, _so skipping all that..._
@@Sapient_Pearwood yah, you can jump across the map with some things afaik.
The idea that people knew which glitches were in which versions, the background music not being in certain versions etc, is just so fascinating.
You underestimate the amount of free time one has when they graduate uni OR decide to skip uni altogether.
Nerds don't like fake nerds 😂
@@ftniceberg874 keep the nerd bloodline stronk!
"Abandon your foolish quest. All that awaits you is the wrath of my TASers! You are too late to save the speedrun. Now you will join it, in Hell."
- Archbishop Lazarus, probably.
Really nice video, I'm glad people are documenting these herculean efforts to verify runs. This is the sort of thing that would normally never be seen by people outside of a specific community.
6:40 I want to clarify a small thing a little bit. Not saying you are wrong, but some background context.
Back in the day, segmented runs weren't just a thing purely because longer videos were difficult to upload. Segmented runs were very common in the early days of speedrunning, in part because that was the standard for Doom and Quake. Normal runs were called "single segment" runs. SDA was initially a Quake-only site before it started accepting other games in 2006, and the single segment run for Quake was not considered prestigious or important-being part of a segmented run was. Basically, for any longer game, it was considered fairly obvious that there would be a segmented run.
Over time, for a variety of reasons, "single segment" became the standard, and these days we don't even use the term anymore. Segmented runs still exist, but mainly for things like FPS games or racing games that have discrete level boundaries (such as Half-Life and Portal), although these are called IL/individual level runs in that case.
In fact, segmented runs were kind of problematic in a way, when you think about it. There were no rules that stated exactly where the segments should start and end. So a runner who decides to cut a game into 40 segments instead of 20 would have a much easier time getting a record, and this was legal per SDA rules at the time. In practice, people would usually reuse the same segments as the previous record holder to keep things fair, and because their segment boundaries would usually make sense, but doing things single segment definitely makes things a lot less complicated.
23:10 By the way, this is just my own personal opinion, but I find it *extremely* unlikely that Mike Uyama would have given Groobo permission to use save files from totally different playthroughs for his segmented run. I know he says he doesn't remember, but there's just no way. 0% chance.
Two reasons: one, it goes completely against the spirit of the rules, and runs were rejected by verifiers for doing the same, as you point out. Doesn't matter if it explicitly says so in the rules or not. Runs always needed to be consistent and contiguous across segments. Two, if he genuinely had gotten special permission, there's no way this would not have been documented for future runners to make use of the same benefit, because it totally changes the difficulty level of the run.
I remember Unreal segmented run on sda used some footage from different playthroughs, but they were deemed inconsequential enough so this was allowed, still felt like an exception but not unheard of.
Its wild that oftentimes these cheaters seem to put more effort into their cheating than they could've spent actually getting good
So true
The cheating is often times not an issue of skill and is more often an issue of RNG. The problem with speed runs is that there are often times multiple people with the skill necessary to pull off the speed run but the limiting factor is not skill its RNG. These runs become so heavily RNG dependent that your just repeatedly doing the same thing thousands upon thousands of times if not more just for the RNG to line up in a way that there is even a possibility of even hitting some type of record.
@@kylesmith987 With the scanning tool we build it's now possible to take out a lot of the RNG since you can locate a good seed and set your system clock so that you get that specific game.
Often the cheaters are really good and the cheat allow them the extra edge to beat the records.
It's just pure symptom of human garbage being. You know the people who put themselves above everybody in the community.
I love how it was instantly proven from the goddamn title screen he cheated, but the investigating group decided to go "well while we know he did it what ELSE did he cheat?"
The investigating group were communist: "I trust everyone's instinct that he cheated, but let me verify."
Partly it had to do with how easy it would be to dismiss it as "oh, I just used an earlier video in the encode"
At that point I feel like half of the reason was because they found it to be a fun puzzle to solve
@@thehoodedteddy1335Not only that. But it's an opportunity to publish full case studies about the game providing knowledge for future investigations. It's also good for attracting attention to the scene through people like Aby and Karl
Let's face it ... the analytic skills of the MODs is INSANE. What a master piece.
You could thank Groobo for cheating so the MODs are able to shine in such a way. Groobo should just confess at this point - but it is not needed anymore.
So in the end, a bunch of nerds got together and did ungodly work and calculations, proved that the Guinness Book of World Records team doesn't check enough, and outted a 16 year old kid that tried to hold on to a lie for too long. I mean when everyone is in their teens they never try to do something bigger than themselves or lie a little too much. Kid did something wrong back in the day, nerds feel good for finding it. Hope the kid has grown up to be a reasonable person.
When I was a kid in the 80s, if someone claimed they beat a game in a certain amount of time, the only tools we had were, "Really?", "You swear?". Usually that was enough and the person would confess to lying or exaggerating their feat.
Now there are whole groups, tools, and access to the game's programs, but it takes people a longer amount of time to admit to cheating.
I miss the simpler days. 😂
Yeah liars and cheaters didn’t really exist before the 90s
Even ignoring health regen, the odds of 19 hits in a row doing max, or very close to max damage are unfathomably low. I estimate 1 in 1978419655660313589123979 (19^19).
Given how pseudorandom RNGs (especially LCGs) work, it's probably not just improbable, but actually impossible. Diablo uses 32 bit PRNG, so it has exactly 2^32 states. It outputs every single one of those 4.2 billion numbers once in a sequence. The seed value determines where in that sequence you start. Because of this, the generator can not output every possible sequence of numbers.
@@tylisirn That would also depend how the cut off is implemented, given the odds (for max) are not a power of two.
@@stanimir4197 I give it 99% chance it being implemented wrong using the modulo operator (because I've been programming since the 90s and that's how we all learned it, since we didn't know any better), which doesn't result in equal distribution. If it is done with modulo, LCGs often have patters in their low order bits (which unsurprisingly are exactly the important bits when sampling with modulo) which makes it extra unlikely that a long sequence of max rolls would exist somewhere in the random sequence.
Even today, if the game is made in C or C++ (and doesn't use something like Unreal Engine), I would still wager a 50/50 chance that it uses modulo. Other languages have better standard libraries which usually gives you a proper RNG API to use instead of having to roll your own, so those usually have correct ranged random function.
@@tylisirn That was my point - it's like just mod. Or even mod 100 (for simple percentages). For instance civ4 uses mod (no rejection retries)... since the source code is available.
Groobo started fireballing in 2nd gear.
He made sure he was double clutching
In his native language you skip the 1st gear.
@@thecrowcook and luckily is was gated by Diablo so he doesn't go off the track (of the splice).
No one ever thought this “record” was legit. The stair placements were so mathematically unlikely that it was essentially impossible.
Still an interesting video though!
No one every proved it to SDA so the record stood there for 15 years, now legit runners can compete for the first place again. Will be interesting to see how it evolves :)
"No one ever thought"
It's still listed on the guiness world reords website.
No one ever thought, but no one ever took action. That's the difference
@ 100%. That’s what makes this video so great!
@@Baddaby And that's the weirdest thing about this whole thing. Obvious mistakes, obvious improbabilities, and this trick passed. I bet it was about money, which points to corruption among judges and modders.
Keep returning and re-watching your videos over and over, this is one of my favourites because it covers a game I have nostalgic ties to. It's so interesting to take apart various features, to see how they work. Playing games I never think how intriguing the processes in the background could be. Thank you for your work!
Also, take a (back)shot each time Abyssoft mentions seed.
Imagine being the guy who comm'd to get a TAS faster than the WR only to inadvertently help discover a huge cheating scandal with said speedrun.
That Diablo fight is super easy to explain, you just need to gear shift into second on the first frame DUH!
From the first time I ever saw that run (years ago), I never took it seriously, I thought it was more like a proof of concept of what could be done. It's clear that there's waaaaay too many cuts.
I had no idea that people actually accepted that as a real run.
as a poc it could have been fine, but as the record on two record bodies it was a depressing smudge :(
can't believe I'm watching a 30 minute long video about 2 guys named groobo and dwango and it's not a funny cartoon
Groobo and Dwango? That isn't a collectathon?
@@thelastgogetaI sure hope not :)
I didn't expect Unix epoch and year 2038 problem being mentioned in a speedrun video, but here we go
This is why I would never even consider cheating in a speedrun, if I was even talented enough to do speedruns that is, which I'm not. But I mean look at this shit, this investigation is simply amazing, I love watching stuff like this, it's so wild to me the amount of talent that's out there, crazy.
I'm so happy to see the 2038 bug used in a cheating investigation. It's like a bunch of professional and hobby stuff of mine overlapping.
For DevilutionX we are fixing it by using unsigned integers so that it won't be a problem until 2106 :D
@@AJenbo finally I can keep playing after death!
@@AJenbo Just punt that can down the road to the next century of modders, huh?
Groobo here, thanks for finding no proof of me cheating. As CEO of Diablo, I’m going to ban everyone from speed running except myself. Thank you.
I'm Groobo, cheater!
Am gorbo
That's a little close to home, heh
No im durty dan
Gorbo is ban
It's funny watching this speedrun again. Seeing it get torn apart is cathartic. Even back then, I knew it was complete bs. Anyone who played the game would be able to tell you how unfathomably unlikely it would be for all of the exits to spawn right next to the entrances.
This run in specific lead me to doubt any seemingly impossible gaming feats, like no-hit runs in Dark Souls. Even if it was legit, it's so ridiculous that it becomes less impressive than something more believable.
we now know for sure that no game of diablo will ever have stairs this good, we can even tell when the next good level one will happen :)
So solely booting up Dark Souls is more impressive than no-hit runs?
@@Geheimnis-c2e It would be for the guy in the video, if he could get passed the title screen without any inconsistencies.
Very interesting! 3 things immediately struck me when I first saw this speed run many years ago: the missing ring, adrias refresh and the fountain quest. the number of fireballs agains diablo seemed really low to me but that was only a rough gut feeling on my side. Since it got recognized as WR i didnt think much of it and/or brushed it off as not being knowledgeable enough. But now it seems like a gross oversight by the people responsible for verifying the run. When a moderately passionate diablo player like me noticed this, it surely should have been spotted by them.
The judges where not personally familiar with the game so relied a lot on the runners explanations.
In short, who tf accepted this submission?
To be fair it seems like a lot of people suspected this run had to be faked from how improbably lucky it was. But at the same time, getting improbably lucky is one of the core parts of speedrunning any game that involves RNG in level generation, loot drops, etc. Hunting for the perfect seed or grinding for that one perfect "one in a million" run where everything goes right is just part of how speedrunning works. So in a competition where luck is just as important as skill you can't reject someone JUST for being "too" lucky. You need real definitive hard proof they've cheated, and back in the day the communities were smaller, there wasn't nearly as good forensic tools, and the internet in general was just a more trusting place.
the amount of cheats and swindlers involved with Guiness is fascinating, and, I'm sure, entirely coincidental.
Considering it originates more from drunkenly boasting to your mates at the pub than it does from actually keeping legit records... definitely best not to take it too seriously.
Guiness was never founded to be an accurate authority or useful reference.
It was always intended to serve as an arbitrary authority to settle disputes at a moment.
The Internet has rendered the entire organization obsolete. Getting a record is not about performing a feat, it is about paying them and adding the false pretense of authority that they bring while performing the feat.
It has quite literally always been something that you pay money to get into and they made an obscure record for you. (Or just lie about it)
Was never into the D1 speedrunning myself - but I always assumed most runs would start with the same predetermined seed.
I saw several speedrunners stream, and they mostly had close stairs for the first many levels.
I will however say, that if Funkmastermp was part of the investigation, then I have 100 % confidence in all details and issues listed.
It's been a huge challenge just to find seeds that are favorable for speedruns so a Set Seed category wasn't particularly viable. If you were watching streams recently, however, then they may have been using seeds that we picked out for the TAS using our scanning tool. Funkmastermp did a few Set Seed runs on Twitch in the days leading up to the DEFCON talk.
a pillar of the community, that one. Agreed fren
Great video! Impressive how far people have to go to disprove a cheater.
The amount of dedication to this investigation is mind-boggling. Also it is a forensic masterpiece. This is beautiful in itself and if that is the only outcome of this sordid tale, that alone makes it worth it. I genuinely am in awe.
The absolute amount of effort that went into the forensic analysis of this run is bonkers. Massive props to the dedicated team who volunteered so much time and compute power to expose a cheated run that was more than 15 years old!
I have no interest in speed running but yet am fascinated by its history
Same, I've never done a speed run, I probably play games slower than most people even, but either watching parts of real speed runs or stuff like this is pretty interesting lol...
I have never played this game, and my knowledge is limited. I have to say I absolutely love this kind of work from the community. From people who love this game. I'll never understood why people lie about records being set because that's not the purpose of why we play these beautiful video games. Abyssoft keep up the great content
the dedication from the team to catch one cheater is bloody admirable..
GG mates..
What a detective story! It feels like at some point it became very clear that groobo was cheating, but "detectives" continued investigation just because it is fun.
Something about well narrated and very technical videos about subjects i almost dont care about just hits different when im supposed to be sleeping.
I played Diablo once a bit on PS2 when i was a kid, i dont honestly care about speedrunning unless its about a game i really like (and even then i´ll watch one or two and be done)... but this is like an incandescent lightbulb to my tired moth brain.
Wow he was ahead of his time for a cheater
Some cheaters are made, many if not most are born.
Why did they remove the music from the character select? I don't know much about Diablo, but that seems to be an odd choice
So patch notes back then weren't great, and they may not have explicitly tried to remove music from the character creation screen, but another change they did doo may have caused the music to stop playing; that's what the team told me when I asked, but we know for a fact the music was removed, whether intentional or not.
Probably a bug. Diablo code is absolute spaghetti and the amount of bugs in the game are immeasurable
@@kphoenix137 Shout outs to my favorite one from Hellfire; it adds a new weapon modifier that makes your weapon stronger, but gives it a chance to duplicate the enemy. This is not supposed to work on unique enemies, but my mom once duplicated Diablo himself.
@@hoodedman6579 they fixed it in hellfire 1.01 :D
@Abyssoft thanks for the response. I guess I'm not too surprised to hear Blizzard coding has always been somewhat rough, but yeah, that all makes sense
Jesus the analysis of this guy's cheating is DEEEEEP. Don't fuck with the speedrunning community. Don't fuck with the Diablo community. DEFINITELY don't fuck with the Diablo speedrunning community.
Honestly i find it a massive red flag whenever someone tries to defend themselves by attempting to debunk a reliable method
Something that stands out to me is the ridiculous amount of time and effort put in by multiple people to make this report. The level of dedication is absolutely wild.
The research put to expose the cheater and leave no doubts is insane. Ggs
It's funny how all of this would've been completely prevented if moderators noticed issues in time. You don't even need to recreate run😅
9:45 "tell me if you can hear anything" ... goes to commerical... yeah well there was pretty big splice for a motorala razor ad in there.
RNG Manipulation is quite a common thing for any and all games which has one or more rng systems.
One of the most known, is well Pokémon, and Minecraft.
But you will find for example Darkest Dungeon, games like Vampire Survivor, Bounty of One, Dwarf Fortress and more.
These can be manipulations on example drop rates for items, spawn rates of items (Like the invisible chest that gives you a Zodiac weapon which can be further manipulated to give you exactly the weapon you wan by 100%, when this chest has a normal 0.01% chance of spawning: Per New Game), and many, many, more things.
So it is possible that he could have gotten access to the seed, edited it, and or used a program outside of the game, which manipulated code inside the game as it runs.
This is a great video, I like watching these specially when they are well made and showcase how it all works.
The Team: "You claim theres a way to Manip item drops despite saying before that you knew it can't be done, mind backing that up with a source?"
Groobo: "My source is I made it the fuck up."
Didn't he literally open his inventory and unequip his old staff before the new one even dropped? His entire chain of actions seemed like he already knew which enemy would drop it when or did I miss anything that mentions it in the video?
It makes sense for him to do that if he had been telling the truth, because he claimed on the original video to be able to manipulate RNG to guarantee item drops. We know now that the claim could not have been true, but at the time the claim was made, it was at least feasible
@@fafdsfr To this day no one has found a way to manip item drops. His claim was always false. He only made that claim because people (rightfully so) would point out how impossibly unlikely it is to have a game seed where all 15 stairs are perfectly next to each other AND him to get the exact 2 items needed for the health glitch in that same seed. "I found a way" was a way to dismiss them out of hand, and he never did explain the supposed method that he "discovered". Ever.
@@fafdsfr Another thing to keep in mind is that it was a segmented run, which was allowed, so if we pretend the staff drop was legit, then if he ran that segment even once before he knows that monster is going to drop that staff, so the pre-dropping of his charged bolt staff isn't suspicious by itself. What's more suspicious is the other stuff, like how that floor's seed isn't possible, and seeds that lead to the same layout don't include that monster dropping the Naj's Puzzler staff.
I wonder what would happen if you sneakily invited a buncha cheaters to speedrun some game they have in common and add a cash price on top. I would love to see the sort of shenanigans they get up to.
But anyways that was unrelated. Pretty cool to see diablo covered though! that's really unexpected.
Bro saying "acrost" @4:25 gave me whiplash
The way he over pronounces every single word but randomly mispronounces a word really quickly is incredibly jarring
Um... What? "Acrossed" is pronounced "acrost" by everyone. No-one has said pronounced it *"acrossèd"* in centuries. The E became silent, and the D is pronounced as a D after voiced sounds, and is pronounced as a T after voiceless sounds.
@@TiesThatBind513That's called an accent, not "mispronouncing".
Hey; I've been playing this game for close to 30 years, never got close to beating it in 3 minutes, my best time was between 12-45 minutes and that's with items donated from other characters. I initially thought the WR run was based on RNG manipulation via some kind of glitch, given the stairs were all right by each other; I didn't suspect cheating on the onset but given the sheer statistical improbability of getting 14 entrance-exit pairs in a row, we all should have been a little sus from the getgo.
Furthermore, in my 30 years of playing Diablo, I have only found Naj's Puzzler TWICE. That's like 2 in 140,000 games. Getting that staff randomly on L10 from the very first enemy makes no god damned sense.
Amazing work on the investigative team, and crazy TAS tools.
12 minutes is a really good time, it's about what the WR is for single run of an unknown dungeon
I'd like to express my heartfelt thanks to these legendary guys for investigating this run. This run bothered me from the first moment I saw it, especially because of the perfect alignment of dungeon starts/exits. And also a big thank to you, Abyssoft, for making this great video exposé!
Oh, glorious algorithm, give your blessings now
22:21 They don't say they used a 32 core and 140 core system. What they're saying isn't entirely clear actually - My initial parsing was they used a total of 32 systems, each with 140 cores. But that's a pretty weird configuration too.
Ajenbo's first language isn't English, and what he meant to say was clarified to me when I was revising my script, so what I narrate iin the video is correct.
To clarify, AJenbo has a 32-core workstation. kevans provided the 140-core server. These are the systems where we did the majority of our most time-consuming scans.
@@Abyssoft I was also super tired at the time of writing :P I'm fully capable of saying nonsense in multiple languages, native or not. The initial message was meant as "systems with between 32 and 140 core".
@@staphen-dev Alright, so the final clarification is thus, a 32 core work station and a 140 core server.
@@staphen-devIs the seed finder program (& source) available for download?
Look out summoning salt and Karl Jobst, there's a new man rising quickly! Great video!
I was disappointed to not be called an absolute legend.
I'm glad this matter is put to rest, it has haunted me for 15 years now, and i can finally die in peace now the miscreant has been brought to justice.
I remember having seen that speedrun vid like a decade ago and immediately dismissing it as fake just from what I figured was the sheer impossibility of getting the staircases that conveniently placed every single floor. That was only the tip of the iceberg apparently lmao.
Thank you for saying “raises the question”. I’m so tired of hearing people say “begs the question” incorrectly when they mean “raises the question”
"we need to talk about this videos sponsor-"
naah we dont.
vidoe starts at 2:01
MVP
Thank you. The intro was annoying lol
@@xXLunatikxXlul They typically are, and im super tired of them.
@@DarthSinistris agreed. I hope more UA-camrs just get right into the topic at hand, instead of all the miscellaneous BS they have at the beginning
I also have a fish brain
This is forensic level investigation. Great job guys.