ccuracy 100 Is only possible if all moves are the best moves. And many moves in this game are excellent but not the best. Watch your own signs on the pieces.
@@scottwarren4998 Game review is by chess.com and accuracy is measured on lichess as lichess is more accurate and doesn't run in my browser. Also chess.com is paid and so then I usually don't save the accuracy numbers for when I am editing so I use lichess for the accuracy.
100% accuracy for both sides in every game, doesn't that imply, that even after they eventually develop a 4500 rated engine, it can't beat either of these 2?
@@sam2725 no, because a 4053 elo engine is rating the accuracy. Also I use chess.com game review for the games but lichess computer analysis for the accuracy because it is more powerful
@@enpassantcheckmate Do you really think your free to use lichess analysis which takes 20 seconds, is as strong as 2 cutting edge engines running on supercomputers? Anyway it's not my point. My point is that 100% acc implies it's a perfect game, but it's not because these engines aren't perfect. Regardless it's a useless static and it's weird why you even put it in the video.
@@sam2725 I get your point, but it is 100% accuracy when reviewed by the current strongest engine it as and some people do want to see it. Also the games aren't played on supercomputers
It's difficult to assign an elo rating to the current strongest computers due to them being a lot more advanced than any humans, making it difficult to compare them and implement them in our elo system. An accurate elo rating can only be assigned when a player or computer has played a ton of games against opponent's who's strengths and elo ratings are known. The elo system is basically flawed so we cannot determine the real elo. Their elo isn't infinite because it's definitely possible and very probable that stronger chess engines will be developed who'll be able to beat the current strongest chess engine with ease
how can i pair two computers against each other and give them a specific time of thinking? for example: i want to pair leela vs stockfish, can you explain how to do it?
Well, we could set the time for them to calculate each move to basically anything and make them calculate forever, but changing the depth doesn't change their top engine move that often really. Their strength is determined by their algorithm and how they evaluate certain positions, and which "branches" they overlook. Stockfish 16 vs Stockfish 16 at a depth of 100 moves of 20 moves will still end in a draw. But to make stockfish calculate at a depth of 100 moves, a relatively good pc would take like 2 days to calculate that far ahead, but wouldn't increase stockfish's strength all that much and could possibly make its strength worse. Which isn't practical, that's why people tend to pick 5 or 3 seconds for each move to be calculated, because that's the perfect amount of time for them to look far enough into the future etc
I mean, the power of the computer doesn't play that big of a role if you set the engine to calculate to a certain depth. Having a slower computer will just make stockfish calculate less quickly, leading to it looking less in the future. But if you make both engines calculate to depth 20 for example, it'll only take slightly longer to do so per move for the slower computer compared to the more powerful computer, even though they play at the same strength. Depth also doesn't change their strengths that much either. The deciding factor is always going to be the engine's algorithm.
@Stockfisher I see. Thank you for the clarification. "Perfect" chess is kind of a cool topic to learn about and your channel and others like it have helped me immensely.
Can anyone explain why white just didn't capture a pawn in the end of the second game? I see, they have an opportunities to do so, but they played as if they were both trying to lead everything to draw. Of course, where is a chance, that the black king can find a position for a stalemate in the opposition on the edge of the board, but as for me, it's better to try with a chance of winning or drawing, than just toss kings for 50 moves to reduce everything to a guaranteed draw.
@@enpassantcheckmate It's a draw in both positions, so why not take a pawn? Okay, I don't understand the logic of these programs, probably there is none.
@@RedHorseArcher The pawn was on the edge of the board and the promotion square wasn't the same colour as the bishop, meaning it was a draw (it just leads to unavoidable stalemate)
Torch v2 is currently unavailable to the public, you can also refer to (CCRL index) for more info on which engines are open source, free, commercial or private.
White couldn't promote it, if white captured the black pawn, they'd lose their pawn aswell or their pawn wouldn't be able to promote due to opposing kings. Many endgame theory videos are on RUclips that explain these principles.
That is a great move because it denies an immediate draw from the 50 move rule in chess, giving him more chances of possibly winning the game if his opponent were to make a mistake. Sacrificing the bishop isn't a problem or doesn't worsen black's position due to white stil lbeing unable to advance and promote their pawn. The king can always keep the opposition..
That wouldn't have changed the outcome of the game. If white did capture the black pawn, white would never be able to promote his own pawn due to the black king keeping the opposition
Join my discord! discord.gg/63xdTCZDPM
Join my channel! youtube.com/@enpassantcheckmate/join
ccuracy 100 Is only possible if all moves are the best moves. And many moves in this game are excellent but not the best. Watch your own signs on the pieces.
@@scottwarren4998 Game review is by chess.com and accuracy is measured on lichess as lichess is more accurate and doesn't run in my browser. Also chess.com is paid and so then I usually don't save the accuracy numbers for when I am editing so I use lichess for the accuracy.
Bro that was theory for 13 MOVES!!!
Nice video... Now its time to watch it
the fact that the pieces move in sync to the music
nah it's not synced
100% accuracy for both sides in every game, doesn't that imply, that even after they eventually develop a 4500 rated engine, it can't beat either of these 2?
@@sam2725 no, because a 4053 elo engine is rating the accuracy. Also I use chess.com game review for the games but lichess computer analysis for the accuracy because it is more powerful
@@enpassantcheckmate Do you really think your free to use lichess analysis which takes 20 seconds, is as strong as 2 cutting edge engines running on supercomputers?
Anyway it's not my point. My point is that 100% acc implies it's a perfect game, but it's not because these engines aren't perfect. Regardless it's a useless static and it's weird why you even put it in the video.
@@sam2725 I get your point, but it is 100% accuracy when reviewed by the current strongest engine it as and some people do want to see it. Also the games aren't played on supercomputers
He means that when better engines come out, it will be used for analysis to see better moves
Bro stopped uploading monthly!
oh boy two top level engines facing each other, i sure hope they dont draw with another berlin!
Why did torch v2 just simply gave up a bishop at 5:00 even though itd a draw?
Maybe it wanted to prolong the 50-move rule by sacrificing a piece for capture to reset the move limit.
Stockfish**
yeah its a draw no matter what so it just starts playing any move that is still a draw because each of those moves are tied for the best
@@-Yttrium correct
Torch has sense of humor
Nice video. I hope stockfish will win this match.
how do you do this bot vs bot?
Why not try to take the black pawn in the endgames for the second game
Finally i found the music
N'to trauma
WAIT THEY ARE 4000 NOW????
They didn't actually change in strength... he just changed the numbers lol
It's difficult to assign an elo rating to the current strongest computers due to them being a lot more advanced than any humans, making it difficult to compare them and implement them in our elo system. An accurate elo rating can only be assigned when a player or computer has played a ton of games against opponent's who's strengths and elo ratings are known. The elo system is basically flawed so we cannot determine the real elo. Their elo isn't infinite because it's definitely possible and very probable that stronger chess engines will be developed who'll be able to beat the current strongest chess engine with ease
how can i pair two computers against each other and give them a specific time of thinking? for example: i want to pair leela vs stockfish, can you explain how to do it?
@@Y_0508_ You have app playing leela and another with stockfish, then you mirror the moves. Or you can use a chess GUI to play them for you
Stockfish is simply unbeatable.
Stockfish 16.1 NNUE vs Stockfish 16.1 please ❤
What is the maximum time we can set to make a move?
Since in this game both were given a time of 5secs
@@purvesh1565 I would say 30 seconds
@@enpassantcheckmate how can we set time?
Well, we could set the time for them to calculate each move to basically anything and make them calculate forever, but changing the depth doesn't change their top engine move that often really. Their strength is determined by their algorithm and how they evaluate certain positions, and which "branches" they overlook. Stockfish 16 vs Stockfish 16 at a depth of 100 moves of 20 moves will still end in a draw. But to make stockfish calculate at a depth of 100 moves, a relatively good pc would take like 2 days to calculate that far ahead, but wouldn't increase stockfish's strength all that much and could possibly make its strength worse. Which isn't practical, that's why people tend to pick 5 or 3 seconds for each move to be calculated, because that's the perfect amount of time for them to look far enough into the future etc
Can anyone explain how we can make engines that play better? Or is it all just depending on the power of the computer?
Its about algorithm, have u heard about computer science?
U must learn algorithm to understand it
ELI5:
they keep learning by themselves but do it a million times faster than any human can.
I mean, the power of the computer doesn't play that big of a role if you set the engine to calculate to a certain depth. Having a slower computer will just make stockfish calculate less quickly, leading to it looking less in the future. But if you make both engines calculate to depth 20 for example, it'll only take slightly longer to do so per move for the slower computer compared to the more powerful computer, even though they play at the same strength. Depth also doesn't change their strengths that much either. The deciding factor is always going to be the engine's algorithm.
@Stockfisher I see. Thank you for the clarification. "Perfect" chess is kind of a cool topic to learn about and your channel and others like it have helped me immensely.
@@AGrasso-ut1ss Yeah, I find AI fascinating aswell.. Can't wait until a new chess engine takes over too, even if it's another stockfish
I missed this awww
@@yifansun2596 No need to worry :)
Can anyone explain why white just didn't capture a pawn in the end of the second game? I see, they have an opportunities to do so, but they played as if they were both trying to lead everything to draw. Of course, where is a chance, that the black king can find a position for a stalemate in the opposition on the edge of the board, but as for me, it's better to try with a chance of winning or drawing, than just toss kings for 50 moves to reduce everything to a guaranteed draw.
If they would capture the pawn, it would still be a draw but I am not a chess engine so I have no idea why it didn't capture the pawn.
@@enpassantcheckmate It's a draw in both positions, so why not take a pawn? Okay, I don't understand the logic of these programs, probably there is none.
@@RedHorseArcher Maybe because the engine prolongs the game in case they accidentally make a mistake they can take advantage of
@@RedHorseArcher The pawn was on the edge of the board and the promotion square wasn't the same colour as the bishop, meaning it was a draw (it just leads to unavoidable stalemate)
@@fblazt9434 So, 50 moves is not a draw?
How much time are they given to think?
@@locrianphantom3547 both about 5 seconds per move
@@enpassantcheckmate Thanks!
Why isn’t stockfish rating its own moves as best
@@Trash_can_Smash the ai that analyzed the game is weaker than these 2
How do you play against torch v2? Is there a link to download syzygy or? Im new :)
Torch v2 is currently unavailable to the public, you can also refer to (CCRL index) for more info on which engines are open source, free, commercial or private.
@@rooksreview torch 2 is available on chess.com analysis
How was the second game a draw, white should have promoted the pawn
@@betagamer398 The black pawn is in front of it
White couldn't promote it, if white captured the black pawn, they'd lose their pawn aswell or their pawn wouldn't be able to promote due to opposing kings. Many endgame theory videos are on RUclips that explain these principles.
Are we getting closer to a full solution to chess?
@@AyushTH We already have 7-man syzygy tablebases but we still have a long way to go
Closer yes, close no
Tourch, pq não foi até o peão negro pega-lo e promover uma dama? Isso é 4000 de rating? 😂😂
Woe
4:59 Wtf is that move by blacks?
Is Stockfish really 4k ELO?))))
It isn't a bad move, that move is probably beyond our understanding.
stockfish is a memer
@@FoxfernXD is a sucker :/
That is a great move because it denies an immediate draw from the 50 move rule in chess, giving him more chances of possibly winning the game if his opponent were to make a mistake. Sacrificing the bishop isn't a problem or doesn't worsen black's position due to white stil lbeing unable to advance and promote their pawn. The king can always keep the opposition..
+5000 meme points
Wtf Torch v2 is very stupid.
They drew?
here is a game played by magnus bot and hikaru bot (magnus was white): 1.d4 d6 2. e4 g6 3. Nf3 Bg7 4. Bd3 Nc6 5. O-O Bxd4 6. Nxd4 Nxd4 7. Na3 Ne6 8.
f4 Nf6 9. b4 d5 10. exd5 Nxd5 11. f5 gxf5 12. Bxf5 Rg8 13. Bxh7 Rg7 14. Be4 Nc3
15. Qxd8+ Kxd8 16. Bd3 a5 17. Bh6 Rg4 18. Nb5 Nxb5 19. Bxb5 Bd7 20. Rad1 c6 21.
Be2 Rh4 22. Be3 axb4 23. Rxf7 Rxa2 24. g3 Rxc2 25. gxh4 Rxe2 26. Bb6+ Ke8 27.
Rh7 Nf8 28. Rh8 Re5 29. Bd4 Rf5 30. Bg7 Bc8 31. Rf1 Rxf1+ 32. Kxf1 Bh3+ 33. Ke2
Kf7 34. Bxf8 Bf5 35. Bh6 b3 36. Bc1 e5 37. Bb2 c5 38. Rh6 Bg4+ 39. Kd3 b5 40.
Bxe5 Be6 41. Rf6+ Ke7 42. h5 b4 43. Rf4 c4+ 44. Rxc4 Bxc4+ 45. Kxc4 Kf8 46. Kxb3
Kg8 47. Kxb4 Kh7 48. Bf4 Kh8 49. Kc5 Kg8 50. Kd4 Kf7 51. Ke5 Kf8 52. Kf6 Ke8 53.
h6 Kd7 54. h7 Kc6 55. h8=Q Kd5 56. Qc8 Ke4 57. Qf5+ Kf3 58. Bc1+ Ke2 59. h4 Ke1
60. h5 Kd1 61. Bg5 Ke1 62. h6 Ke2 63. h7 Ke1 64. h8=Q Kd1 65. Qd8+ Ke2 66. Qd2#
1-0
Thank you, but next time may you please enter it into the submit-games channel in my discord? Thank you for the game
@@enpassantcheckmate ok
Torch V3 will beat fish
Fish 17 will beat torch
@@nhan1710 torch 98 will beat fish 45
You will beat your meat
@@samukinhav0836 And Levy still isn't a GM
Почему он пешку не ест??? 6:07
That wouldn't have changed the outcome of the game. If white did capture the black pawn, white would never be able to promote his own pawn due to the black king keeping the opposition