I'd love to know the win rate by time control. It's a lot different in bullet than it is in rapid, for instance. Shorter time controls favor dubious openings by players that know all its lines, where the same can't be said for longer time controls.
Finally a haloween gambit video, i play this always, i played it in a otb tourny against a 1900 and had the advantage for majority of the game (i ended up losing because of a silly mistake in the endgame) but it make me happy to see it is finally get more recognition.
I think, there might be some misconception regarding Hallowing gambits. It’s not particularly a “if you are at 1600 and you play Halloween Gambit, you will have a 55% winrate.” But more so that if you play Halloween gambit at a 1600 level, you are damn well versed in this opening, and also have an element of surprise
I think the winrate goes down in those middle rating ranges because more people plays Ng6 instead of Nc6 as the rating goes up. A theory, they see that Nc6 loses a lot of tempos
I first learned the Halloween gambit a couple of months ago from one of your older videos. 4.Nxe5 is the only thing I've been doing from the four horsemen and I'm winning 57% and losing only 40%
I think it's realy necessary to see the gamemode as well. I'm 1100 and the halloween gambit really works for me in rapid, but I doubt it works as fine for a 2000 elo. My guess is that as you go up the ratings, the propotion of faster ganemodes increase and it affexts the winrate
It would have helped if you included the win rate for white with all openings for white at the different levels, and possibly for all lines after a 4 nights opening.
Just like the other common gambits (except for Queen & King), Offering a gambit is a mistake before knowing how to utilize the advantageous position from the exchange. As for the opponent accepting the gambit, it is an inaccuracy from that point onwards. But as the video concludes, at higher levels offering the gambit becomes a risk again just as in low levels.
on lichess at all levels except 400 it's at least a few percent better, white win rate for starting pos/e4 tends to be the same and goes from 50% to 48% as rating increases and draws become more likely
You know that at top level this is a LOSING move - if you check the engine at top level it's actually negative (about minus 1.4 cp) for white!. So it's not unexpected that when 2000-2200+ play it (and they rarely would) they are playing other tough players who know how to take advantage of that "blunder/inaccuracy". While anyone < 2000 would have an opponent that isn't so strong to take advantage of that. On the other hand, those < 1000 don't count, they play so badly and randomly it's almost like tossing coins.
1350 here i think you saying at this level we are starting to get an idea of how to attack is partly correct, i'd say, at my level, people start to know a couple openings as white, but not quite as black and 4.Nxe5 is definitely not what we see everyday, that probably throw black out of their comfort zone. (i think in the 4knight opening, i see 4. Bb5 and 4.Bc4 are a lot more)
Only thing I wonder is what time controls is this played most at to give these numbers? I always hear of this as a great bullet or blitz opening, but not as good in classical.
Have you considered comparing the win rates of playing the Halloween gambit with playing other moves in the same position? That might be useful in figuring out whether or not a move is good in a position.
@D1rtyD3bra only against a computer. According to the computer the Halloween gambit is not good but as you can see here against a person if you know what you are doing it is good. Remember people aren’t computers and you have to play people not theory.
@@delayedhoe9714 Well sure but stockfish will still tell you the best moves in a bad position. But there may be moves that are good practically that it won’t show you because it sees some obscure refutation
Lots unsaid here. What are the ratings of the opponents? What are the baseline win rates of starting with white? What are the win rates when forgoing the gambit? I’ve heard the Chinese are good at chess-maybe I should move to China?
I'm more interested in counters to gambits, traps and other tricks. One idea I had was to counter sacrifice a knight for a pawn, I think you should be able to win a second one. You would face being behind in development though...
Given that various comparisons of win rates for white since 1851 have placed the base first-move advantage anywhere from 52-56%, I'm not sure this gambit is helping you that much. Your base win rate never exceeds 55%. What this does seem to show is that if you're black and this is played on you, it's significantly better the retreat the knight to G6 than to C6. Moving the KN all the way back to G8 looks silly, but the combination of the other knight at G6 and later moving out the KN to H8 are probably providing a pretty decent king-side defense.
Interesting video! Would be good to know the sample sizes. The win rates of 53-55% in the first column might be statistically virtually identical depending on how many games are included, similarly for 59-61% in the second column and 52-54% in the last column. And is the average game length (time control) correlated with the rating at all? Would be interesting to see split by time of game.
יכול להיות שזה עולה בגלל שאנשים לא משחקים את זה ברמה 1200 ומעלה ולכן שנתקלים בזה זה מבלבל ומפסידים חלק אבל אדם עם טיפטפה מכיר את הפתיחה יקטול אותך בנוסף לזה בקושי משחקים את זה ולכן אולי אין הרבה על מה להתבסס צריך הרבה מאוד משחקים כדי לעשות סטטיסטיקה מדויקת ויעילה
Thanks for the info, very useful as usual! I'm 991 Elo and I always play this gambit when in the four knights position. I really thought it woild get less success starting from ~1400 elo, since people would have learnt to counter this, but the results you showed completely proved me wrong^^
Ah statistics, don't you just love them? Thanks for your research on this. As an unrated player I think I could have some fun with it against my friends.
Why do you say a 50% win rate is "pretty good" (2200+ Move 4 WR)? Isn't the win rate for White, before the game even begins, 52% or more? Maybe I'm missing something, but doesn't that suggest that an opening like that is a sub-par opening? Thanks.
Nelson! I just tried this for the first time.. I hammered a player 200 point above me to smithereens in a very fun and interesting game.. thank you! Also I got tired from playing the Queens gambit all the time :)
If you play this gambit against highest level Stockfish (from the gambits starting position), according to Stockfish you're in a -1 position, but with perfect play (both sides playing best moves) I managed to beat Stockfish from that -1 position, shows how strong those centralized pawns are.
You know what else you can do against it? Mainline Halloween gambit after 5d4, black can just answer with Bd6 and has instantly, fully equalized the position, and it’s easy to play Is it the best move? Of course not, but it fully neutralizes the gambit, and white literally gets absolutely nothing from the opening 🤷🏼♂️
@@ifbfmto9338 Interesting, that directly transposes into the Nxe4 line of the 4 knights italian with reversed colors. What I like to play as black is 5.d4 Nc6 6.d5 Ne5 7.f4 Ng6 8.e5 Bc5, white gets the piece back, but has overextended pawns and thus a weak king, black is significantly better.
2:52 The Irish Gambit is 1. e4 e5 2. Nf3 Nc6 3. Nxe5 Nxe5. The guy who was asked on his deathbed why he took the pawn, and he replied, "Because I didn't see that it was protected," and then died. I'm not so sure that the Win Rate itself is a good statistic; it's been tried many times over the past century or so. If White blunders a queen or checkmate later in the game, that really didn't have anything to do with the opening. And longer games ought to be thrown out as well, after the position gets back to equality, because the opening doesn't really affect things, either. Also, a century or so ago, someone pointed out that if someone won a game 9 times with a certain opening, and the 10th time, a refutation was found, then no one would ever play it again, yet it would have a 90% win rate.
I'd love to know the win rate by time control. It's a lot different in bullet than it is in rapid, for instance. Shorter time controls favor dubious openings by players that know all its lines, where the same can't be said for longer time controls.
Finally a haloween gambit video, i play this always, i played it in a otb tourny against a 1900 and had the advantage for majority of the game (i ended up losing because of a silly mistake in the endgame) but it make me happy to see it is finally get more recognition.
I think, there might be some misconception regarding Hallowing gambits.
It’s not particularly a “if you are at 1600 and you play Halloween Gambit, you will have a 55% winrate.”
But more so that if you play Halloween gambit at a 1600 level, you are damn well versed in this opening, and also have an element of surprise
Nelson does 'Spooky entanglement ' better than a particle physicist.
Faster than the speed of light
Your opponent gets tricked and you get a treat
this is an awesome video idea. would love to see more of this analysis for different openings.
Practice what you preach. Peter Patzer now has to play this until 2200 🎉
I think the winrate goes down in those middle rating ranges because more people plays Ng6 instead of Nc6 as the rating goes up.
A theory, they see that Nc6 loses a lot of tempos
I first learned the Halloween gambit a couple of months ago from one of your older videos. 4.Nxe5 is the only thing I've been doing from the four horsemen and I'm winning 57% and losing only 40%
Up until the 9th minute, I was confident that I would have a 59% win rate against Magnus Carlsen using the Halloween Gambit.
I play this gambit i always play this gambit when i get the chance. Im ~1800 elo now and from 1400 to 1800 climbing i’ve had around 60% wins
You could say that this opening is a, *KNIGHTMARE*
It's not an opening that I would have risked before watching. I may have to give it a go for experimentation sake 😊😊😊❤😊😊😊
I think it's realy necessary to see the gamemode as well. I'm 1100 and the halloween gambit really works for me in rapid, but I doubt it works as fine for a 2000 elo. My guess is that as you go up the ratings, the propotion of faster ganemodes increase and it affexts the winrate
It would have helped if you included the win rate for white with all openings for white at the different levels, and possibly for all lines after a 4 nights opening.
Just like the other common gambits (except for Queen & King),
Offering a gambit is a mistake before knowing how to utilize the advantageous position from the exchange.
As for the opponent accepting the gambit, it is an inaccuracy from that point onwards.
But as the video concludes, at higher levels offering the gambit becomes a risk again just as in low levels.
How does this compare the general win percentage when playing with the white pieces?
on lichess at all levels except 400 it's at least a few percent better, white win rate for starting pos/e4 tends to be the same and goes from 50% to 48% as rating increases and draws become more likely
Obviously Kg4 after pawn attack is not an option even for low-rated players, but that's exactly where I would move my knight to 😁
K is for King, we use N for Knight because the king already uses up the K ❤
@@riley_028 Sure, thanks!
You know that at top level this is a LOSING move - if you check the engine at top level it's actually negative (about minus 1.4 cp) for white!. So it's not unexpected that when 2000-2200+ play it (and they rarely would) they are playing other tough players who know how to take advantage of that "blunder/inaccuracy". While anyone < 2000 would have an opponent that isn't so strong to take advantage of that. On the other hand, those < 1000 don't count, they play so badly and randomly it's almost like tossing coins.
I wish that you are gonna win a game with this on the rating climb😉
1350 here
i think you saying at this level we are starting to get an idea of how to attack is partly correct,
i'd say, at my level, people start to know a couple openings as white, but not quite as black
and 4.Nxe5 is definitely not what we see everyday, that probably throw black out of their comfort zone.
(i think in the 4knight opening, i see 4. Bb5 and 4.Bc4 are a lot more)
Only thing I wonder is what time controls is this played most at to give these numbers? I always hear of this as a great bullet or blitz opening, but not as good in classical.
You should do this for more gambits
Yes. It appears better win rates are coming from just being better to me.
Chess vibes could you please make a video about the Danish Gambit, mainly the traps? Thank you if you do
Ive never seen a video posted 1 minute ago
This is a very good gambit, very fun and enjoyable.
Have you considered comparing the win rates of playing the Halloween gambit with playing other moves in the same position? That might be useful in figuring out whether or not a move is good in a position.
Well if you wanted to know if a move was good wouldn’t just checking the engine lines probably suffice?
@@DaltonDaDiggaNot necessarily imo. The halloween is losing by Stockfish's standards, but it has high win rates
@D1rtyD3bra only against a computer. According to the computer the Halloween gambit is not good but as you can see here against a person if you know what you are doing it is good. Remember people aren’t computers and you have to play people not theory.
@@delayedhoe9714 Well sure but stockfish will still tell you the best moves in a bad position. But there may be moves that are good practically that it won’t show you because it sees some obscure refutation
To be fair though all the other openings in the 4 knights are trash
Yay almost first! Imma play only Halloweem gambit. Tis the season 🎃🎃🎃
Lots unsaid here. What are the ratings of the opponents? What are the baseline win rates of starting with white? What are the win rates when forgoing the gambit? I’ve heard the Chinese are good at chess-maybe I should move to China?
I'm more interested in counters to gambits, traps and other tricks.
One idea I had was to counter sacrifice a knight for a pawn, I think you should be able to win a second one. You would face being behind in development though...
Given that various comparisons of win rates for white since 1851 have placed the base first-move advantage anywhere from 52-56%, I'm not sure this gambit is helping you that much. Your base win rate never exceeds 55%.
What this does seem to show is that if you're black and this is played on you, it's significantly better the retreat the knight to G6 than to C6. Moving the KN all the way back to G8 looks silly, but the combination of the other knight at G6 and later moving out the KN to H8 are probably providing a pretty decent king-side defense.
win rate for white on lichess is 50%, black is 47%
It's dubious with perfect play, especially if your opponent knows the main line, after taking a free piece.
Very interesting i wish you show to us how to play this in your level now in rapid 10 minutes games.Thank you for this video!
Very informative, thanks. Could you do please all other openings as well, so I could pick a new one matching my rating? :)
1.e4 e5 2.Nf3 Nc6 3. Nc3 Nf6 4.Nxe5 Nxe5 5.d4 Nc6 6.d5 Ne7....i get this a lot from the black pieces.
the best gambit i've seen so far
like 5 minutes in and all i can think is knight G4 after first pawn attack.
Thank you Teacher Nelson
If the win rate always drops in your chart from Nc6 box to the Ng6 box doesnt this suggest there are better moves for white after Nc6?
what are those rating more precisely? Lichess, Chess or something else? Bullet, blitz or rapid?
Interesting video! Would be good to know the sample sizes. The win rates of 53-55% in the first column might be statistically virtually identical depending on how many games are included, similarly for 59-61% in the second column and 52-54% in the last column.
And is the average game length (time control) correlated with the rating at all? Would be interesting to see split by time of game.
יכול להיות שזה עולה בגלל שאנשים לא משחקים את זה ברמה 1200 ומעלה ולכן שנתקלים בזה זה מבלבל ומפסידים חלק אבל אדם עם טיפטפה מכיר את הפתיחה יקטול אותך בנוסף לזה בקושי משחקים את זה ולכן אולי אין הרבה על מה להתבסס צריך הרבה מאוד משחקים כדי לעשות סטטיסטיקה מדויקת ויעילה
Love these type of videos!!!
Fitting you make this video the day after a chance you got to play Halloween gambit during rating climb
Gonna try this out, thanks!
Thanks for the info, very useful as usual! I'm 991 Elo and I always play this gambit when in the four knights position. I really thought it woild get less success starting from ~1400 elo, since people would have learnt to counter this, but the results you showed completely proved me wrong^^
How do these numbers compare to the success rate of simply starting as white for each rating?
win rate for white on lichess is 50%, black is 47%
Ah statistics, don't you just love them? Thanks for your research on this. As an unrated player I think I could have some fun with it against my friends.
There are lies, damned lies, and statistics. -- Mark Twain
How livestream going?
Nelson having some serious murderous eyes, peed a bit in my pants
Can you do the same for the Ponziani?
halloween gambit is crazy ideas
I had no idea it's that successful! Wow.
The real lesson is for black not playing the main line chasing of your knight.
Why do you say a 50% win rate is "pretty good" (2200+ Move 4 WR)? Isn't the win rate for White, before the game even begins, 52% or more? Maybe I'm missing something, but doesn't that suggest that an opening like that is a sub-par opening? Thanks.
No because draws most "good" openings only have winrates of like 48-49% or so
Halloween Gambit has to be one of the most sound unsound sacrifices in Chess.
Amazing video :)
Nelson! I just tried this for the first time.. I hammered a player 200 point above me to smithereens in a very fun and interesting game.. thank you! Also I got tired from playing the Queens gambit all the time :)
Who is here after the live?
Statistically, most of us.
I'm 😊
Meetoo
Me
Me
Time control? Draw rates?
So all I need is to reach 1600 by end of this month?
What time control? Blitz or rapid?
And then the Q f3. Mate (400)?
Who is here after Halloween 🤔
If you play this gambit against highest level Stockfish (from the gambits starting position), according to Stockfish you're in a -1 position, but with perfect play (both sides playing best moves) I managed to beat Stockfish from that -1 position, shows how strong those centralized pawns are.
You know what else you can do against it? Mainline Halloween gambit after 5d4, black can just answer with Bd6 and has instantly, fully equalized the position, and it’s easy to play
Is it the best move? Of course not, but it fully neutralizes the gambit, and white literally gets absolutely nothing from the opening 🤷🏼♂️
Stockfish depth and moveset please
@@ifbfmto9338 Interesting, that directly transposes into the Nxe4 line of the 4 knights italian with reversed colors.
What I like to play as black is 5.d4 Nc6 6.d5 Ne5 7.f4 Ng6 8.e5 Bc5, white gets the piece back, but has overextended pawns and thus a weak king, black is significantly better.
So you beat max level stockfish? that sounds legit
Wow youre so cool you beat max level stokfish from a -1 opening yea bro I definenetly believe that
I think I’ve only lost 1 game with the Halloween gambit
Make a video playing Chess 960
I have been playing halloween gambit nonstop since Nelson posted that first video and I have been super succesful with it for quite a while now. haha
Interesting. But what do you do after those four moves? Lol 🤷♂️
He has a vid dedicated to your question in his links under the title.
What to do next after all these moves????? How to win??? We still behind in material, still looks like a waste of a knight.
You have greater center control and a huge lead in development, should be really helpful in the opening
@@elithemitzu pointless if no extra knight
It's Halloween!
Why is this called the Halloween gambit?
cuz its scary to play
Hey who was it the stream
Missed first🥇 by 23 seconds :(
2:52 The Irish Gambit is 1. e4 e5 2. Nf3 Nc6 3. Nxe5 Nxe5. The guy who was asked on his deathbed why he took the pawn, and he replied, "Because I didn't see that it was protected," and then died.
I'm not so sure that the Win Rate itself is a good statistic; it's been tried many times over the past century or so.
If White blunders a queen or checkmate later in the game, that really didn't have anything to do with the opening. And longer games ought to be thrown out as well, after the position gets back to equality, because the opening doesn't really affect things, either.
Also, a century or so ago, someone pointed out that if someone won a game 9 times with a certain opening, and the 10th time, a refutation was found, then no one would ever play it again, yet it would have a 90% win rate.
I am 1300 😊player.
A guy
check!!
We want Kristen
NICE
.
1st
lets go
Come from the stream :D
First comment
I’m first viewer
First