Hey guys I use cheat engine with smartphone fist 5 move engine then random move try to exchange major pieces bring opponent to endgame and try to help with engine opponent become mad. I never get caught I have lots of id . I use smartphone for cheating . This is so much fun hahahahaha
i think i know why lc0 plays kf7. table base is available when 7 pieces left on the board. if rxh6 by lc0 you have 7 pieces and white would guaranteed a win. i think lc0 logic is; rxh6 is not the best move simply because it leads to losing tablebase, even if dragon can reduce it to 7 pieces on next few moves, lc0 just hardcoded to NOT do that on its own volition.
The sky is the limit 😂 it's finished not necessary to make crook straight by means of hook or crook..just kidding but yeah enough can't catch up with engines yet better enjoy a bit than sacrifice the king😂
I watched your solo commentary for the first time. I watched your duo commentaries before and could easily notice missing some deep lines. I know you aren’t gm or super gm but your solo one was much better than many super gms commentary. Well prepared, nicely explained
Sometimes in endgames the engines will desperately try to keep pieces on the board so any losing 6 piece position is to them a dead loss no matter how many complex moves are left.
@ 01:39, it is +1.3 for white in stockfish, so why don't you say I don't, it's just an artificial intelligence thing instead of "it's hanging without being hanging" explanation. Thanks for the great commentation.
"when you are down a piece the rule is sac another one said nobody" 🙂 . James commentary makes these sacrifices even more enjoyable . Edit :- Kf7 was a weird move though . Don't know why Leela played it in the end even though the game was anyway lost .
Tal said: „If there are two pieces under attack, offer a third one!“ (exept a fork) That was his magic. Because one of these pieces has an extra ride and goes devastating! 🤪
@@bendeguire6965 When the bar says +75 for white, basically any move for black is losing. We as human might say this move would make more sense but in reality for the engine even those moves are bad for black. I see your point to play always for win or draw but computers do not make mistakes. Once the game is a one-sided situation the black pieces are in a forced checkmate condition and as far as I followed these games there is no concept of lose with dely vs lose immediately.
The short answer to why leela played kf7 is that it simply played a move that wasnt "losing in the simplest way". while it may seem easier for a human to just simplify position, for an engine that uses database and calculation, "concepts" arent really a thing for them. This can be seen in the early 2000s when engines like komodo could be easily exploited into playing a "worst" move when facing pawn fortress situations.
The game reminded me of the series where Leela blasted apart Stockfish, often locking down development of one side of the board (typically queen side) despite being -4 - -6 score wise from pieces alone, the active pieces were just too much. I might have wrecked blacks defence to get 3 past, connected and central pawns, that to me feels quite logical, but the rest of the game and especially that willing knight and pawn sacrifice is amazing. I assume the whole purpose of that is just to slow down blacks development further and get the queen away, but it takes some vision, to put it mildly. The chess engines really have become a wonder to watch.
At 11:38, i think leela has evolved that if it knows it will lose endgames in less than 20 moves, it will just throw harder so the lose will be quicker
no they are coded to win. they don't know how to lose, fast or slow. i think i know why lc0 plays kf7. table base is available when 7 pieces left on the board. if rxh6 by lc0 you have 7 pieces and white would guaranteed a win. i think lc0 logic is; rxh6 is not the best move simply because it leads to losing tablebase, even if dragon can reduce it to 7 pieces on next few moves, lc0 just hardcoded to NOT do that on its own volition.
@@alsy0055 yes they are coded to win, but for Leela Ai, it prefers to simplify situations. I forgot youtube video which showcases leela's weird behavior but there was one when she had 3 pawns up in king endgame and she threw 2 pawns even though she could promote all three to queens.
@@lalajun5902 i dont even care about the argument anymore, but the way you use 'she' as pronoun while i used 'it' as pronoun gives me an AI existential crisis lmao.
I would assume so. The neural network as no need to understand how to extend the game from a lost position, in fact it would take up space in the network and wouldn't be worth learning. leela might actually play in such away that it's clones might make mistakes, you will notice stock fish is incapable of playing hope chess, however leela might have learned some hope chess strategies in training, as it's clones would have made mistakes some times.
Do people think dragon is worth the cost? I like that you can have the option of Monte Carlo or brute force. However, I currently use Lc0 and that is fine for my level of chess.
5:55 "down a piece"... well, the A8 rook is blocked in by the B8 knight and the knight has no opportunity to move without walking into a horrid mating attack. And is it just me or has black back rank hardly moved anything at all? White might technically be down materially but when you can't move you knight without walking into mate and that knight is blocking a rook in the corner, you're not actually up material, you are just loosing. At that point saccing is obvious to an experienced player. When you are up tempo, your opponents king is unsafe, and you have their position locked in, you make as many forcing moves as possible so you can hold pressure and maintain advantage. It just so happens that saccing is about the most forcing thing you can do outside of a check.
Someone pointed out that the game of chess is solved for fewer than 7 pieces on the board, if leela takes the pawn, now your loss is guaranteed, so leela might just make any other move to avoid that.
@@-zelda- Yes. they defiantly do Things like throwing away all their pieces just to delay mate, a human would look for a trap. In the case of leela, moves they make after defeat is guaranteed way as well be random as they don't even try and delay mate
@@gwentarinokripperinolkjdsf683 Engines expect perfect play from the opponent, delaying mate is not doing "silly moves". So again: no, engines don't do silly moves.
Pretty important too. They usually have faster time controls and better CPUs which means that CPU engines have an advantage there (or you could also say that GPU engines have an advantage at TCEC)
Judging from the nasty raucous clicks at 1:49 , I think my man James Canty are using blue mechanical switches on his keyboard. Therefore, he just lost a fan.
...Kf7 is such a dumb move. It almost feels like engines get worse the further they can look ahead if they are already in a losing position. Imagine playing an engine with rook odds or something to see how well you can do, and the engine just starts gifting you more pieces instead of still trying to win because it knows it's lost anyway.
Someone else had a nice idea why this may have happened in a comment above - to do with tablebase and chess being solved for 7 pieces or less. Do have a look.
I like Canty's analysis
Check out his TH-cam channel its really good
🔥🔥🔥
@@sheddrickkstraker369 thanks bro 💯
Who doesnt?🔥
@@JamesCantyIII Yo wassup
Canty always the best commentator for these analyses!
🔥🔥🔥
"He sacked 2 peices! 12345.. 1234... yeah 2 peices" fantastic
Hey guys I use cheat engine with smartphone fist 5 move engine then random move try to exchange major pieces bring opponent to endgame and try to help with engine opponent become mad.
I never get caught I have lots of id .
I use smartphone for cheating .
This is so much fun hahahahaha
i think i know why lc0 plays kf7. table base is available when 7 pieces left on the board. if rxh6 by lc0 you have 7 pieces and white would guaranteed a win. i think lc0 logic is; rxh6 is not the best move simply because it leads to losing tablebase, even if dragon can reduce it to 7 pieces on next few moves, lc0 just hardcoded to NOT do that on its own volition.
Nice! This is a good point.
The sky is the limit 😂 it's finished not necessary to make crook straight by means of hook or crook..just kidding but yeah enough can't catch up with engines yet better enjoy a bit than sacrifice the king😂
King f7 is like lighting a cigarette when you are about to explode. Like in those 80's and 90's action movies we love (or maybe it's just me) LOL.
I watched your solo commentary for the first time. I watched your duo commentaries before and could easily notice missing some deep lines. I know you aren’t gm or super gm but your solo one was much better than many super gms commentary. Well prepared, nicely explained
Enjoyed that analysis, well done.
Thanks John!
I like how passionate Canty
Is about the games
Always fun to watch Canty slap down those engine moves for our viewing pleasure 😆
Sometimes in endgames the engines will desperately try to keep pieces on the board so any losing 6 piece position is to them a dead loss no matter how many complex moves are left.
@ 01:39, it is +1.3 for white in stockfish, so why don't you say I don't, it's just an artificial intelligence thing instead of "it's hanging without being hanging" explanation.
Thanks for the great commentation.
"when you are down a piece the rule is sac another one said nobody" 🙂 . James commentary makes these sacrifices even more enjoyable .
Edit :- Kf7 was a weird move though . Don't know why Leela played it in the end even though the game was anyway lost .
because the probability to be in stalemate is greater with Rook and Queen
🤣🤣🤣🔥🔥🔥
@@luciangv3252 computers dont go for stalemate tho right? they only play to win.
@@bendeguire6965 Leela isn't a computer; it's an AI. And I believe she does weigh in stalemate probabilities when losing.
@@Nobody-Nowhere-Nothing oh cool thanks
I see NM James Canty, I like the video.
Seeing the lines these computers come up with will never get old.
Sick Canty content as usual!
Tal said:
„If there are two pieces under attack, offer a third one!“ (exept a fork)
That was his magic. Because one of these pieces has an extra ride and goes devastating! 🤪
When the engine realized it is about lose, it will give up! When it is +75 as you mentioned, there is no point of continuing for the engine!
they are coded to always play to win is what i thought. EDIT alpha zero as an AI unlike computers will go for draws.
?
@@bendeguire6965 When the bar says +75 for white, basically any move for black is losing. We as human might say this move would make more sense but in reality for the engine even those moves are bad for black. I see your point to play always for win or draw but computers do not make mistakes. Once the game is a one-sided situation the black pieces are in a forced checkmate condition and as far as I followed these games there is no concept of lose with dely vs lose immediately.
@@saman007 Moves made by engines in losing positions are effectively random.
Great video, James!
Amazing video Love your video 😻 hoping weekly engine games
I'm a NM Canty fan 4 life. Also, INSANE GAME
Great commentary!!!
The short answer to why leela played kf7 is that it simply played a move that wasnt "losing in the simplest way". while it may seem easier for a human to just simplify position, for an engine that uses database and calculation, "concepts" arent really a thing for them. This can be seen in the early 2000s when engines like komodo could be easily exploited into playing a "worst" move when facing pawn fortress situations.
thanks man fun game and commentary enjoyed it!
Canty presenting super engines duel get's always an insta like...
Cantys you are very good on what you are doing, keep it up 👍 1 of the best out there for sure 👍
I always think of these sacrifices but I never win
I thought all this artifical food I been eating would make me ai 🤔
becasue is so scary, u need to make a perfect moves.
Need more Canty III.
The game reminded me of the series where Leela blasted apart Stockfish, often locking down development of one side of the board (typically queen side) despite being -4 - -6 score wise from pieces alone, the active pieces were just too much.
I might have wrecked blacks defence to get 3 past, connected and central pawns, that to me feels quite logical, but the rest of the game and especially that willing knight and pawn sacrifice is amazing. I assume the whole purpose of that is just to slow down blacks development further and get the queen away, but it takes some vision, to put it mildly.
The chess engines really have become a wonder to watch.
Thats a crazy crazy line.
Crazy game!
Very good commentary.
Nice guy!
GRRREATINGS from Germany
Bro I died when he said en croissant
At 11:38, i think leela has evolved that if it knows it will lose endgames in less than 20 moves, it will just throw harder so the lose will be quicker
no they are coded to win. they don't know how to lose, fast or slow. i think i know why lc0 plays kf7. table base is available when 7 pieces left on the board. if rxh6 by lc0 you have 7 pieces and white would guaranteed a win. i think lc0 logic is; rxh6 is not the best move simply because it leads to losing tablebase, even if dragon can reduce it to 7 pieces on next few moves, lc0 just hardcoded to NOT do that on its own volition.
@@alsy0055 yes they are coded to win, but for Leela Ai, it prefers to simplify situations. I forgot youtube video which showcases leela's weird behavior but there was one when she had 3 pawns up in king endgame and she threw 2 pawns even though she could promote all three to queens.
@@lalajun5902 i dont even care about the argument anymore, but the way you use 'she' as pronoun while i used 'it' as pronoun gives me an AI existential crisis lmao.
@@alsy0055 "its 2022 bro,even AI's have gender rights"💀
The king is in the middle of the street here guys.
In the spirit of Mikhail Tal
@ 01:42, I would never be either Bd2 or d5, it would Be3 for example. What is wrong with that?
Love these
I like guys gangster guys like Canty 😂😂
6:50 Qf3 is not The best move :0
Didn't know that Lc0 analysis was available on chesscom
It is not
Game 23 watched ❤
I like en croissant move too.
Cool commentary.
I like this one
Interesting: does Leela-0 play a bit randomly when the position is clearly lost? (where Stockfish and Dragon would go for the longest mating sequence)
I would assume so. The neural network as no need to understand how to extend the game from a lost position, in fact it would take up space in the network and wouldn't be worth learning.
leela might actually play in such away that it's clones might make mistakes, you will notice stock fish is incapable of playing hope chess, however leela might have learned some hope chess strategies in training, as it's clones would have made mistakes some times.
i love this dude
Do people think dragon is worth the cost? I like that you can have the option of Monte Carlo or brute force. However, I currently use Lc0 and that is fine for my level of chess.
It is not
Considering SF is still currently the best and is free and open source... Id say no
On immortal chess forum people give it out for free. Keep it on the low though. Lol.
Insane game
5:55 "down a piece"... well, the A8 rook is blocked in by the B8 knight and the knight has no opportunity to move without walking into a horrid mating attack. And is it just me or has black back rank hardly moved anything at all? White might technically be down materially but when you can't move you knight without walking into mate and that knight is blocking a rook in the corner, you're not actually up material, you are just loosing. At that point saccing is obvious to an experienced player. When you are up tempo, your opponents king is unsafe, and you have their position locked in, you make as many forcing moves as possible so you can hold pressure and maintain advantage. It just so happens that saccing is about the most forcing thing you can do outside of a check.
en croissant is gold
Leela says: I've already lost, but my program doesn't allow me to resign. I'm going to have a tantrum. Kf7!!!
Someone pointed out that the game of chess is solved for fewer than 7 pieces on the board, if leela takes the pawn, now your loss is guaranteed, so leela might just make any other move to avoid that.
When engines calculate that the game is over,they usualy do silly moves. Just to kill the game.
no
@@-zelda- Yes. they defiantly do Things like throwing away all their pieces just to delay mate, a human would look for a trap.
In the case of leela, moves they make after defeat is guaranteed way as well be random as they don't even try and delay mate
@@gwentarinokripperinolkjdsf683 Engines expect perfect play from the opponent, delaying mate is not doing "silly moves". So again: no, engines don't do silly moves.
How does CCC compare to TCEC in terms of importance in the world of computer chess?
Pretty important too. They usually have faster time controls and better CPUs which means that CPU engines have an advantage there (or you could also say that GPU engines have an advantage at TCEC)
Smooth talker
I keep forgetting Is this guy James canty the 1st, 2nd, or 3rd? I need someone to remind me so that I can stop forgetting his name
Link PGN ?!
I like to see your videos on speed 1.5
LOL King f7 shows computers really are blind kung fu masters
Please change the logo on your chair, it makes you like monster with your ears, so sorry but thats true
lol wtf
0:40
If we go so far that we see Leela and Dragon then why not Stockfish 15 vs Dragon?
Because SF 15 hasn't been released yet
Judging from the nasty raucous clicks at 1:49 , I think my man James Canty are using blue mechanical switches on his keyboard. Therefore, he just lost a fan.
The modern is a terrible opening for engines, the consistently refute it.
...Kf7 is such a dumb move. It almost feels like engines get worse the further they can look ahead if they are already in a losing position. Imagine playing an engine with rook odds or something to see how well you can do, and the engine just starts gifting you more pieces instead of still trying to win because it knows it's lost anyway.
Engines expect best play for the opponent. If "gifting" you more pieces leads to taking more time to get mated, they are going to do it
@@-zelda- I understand that. It's still dumb
Someone else had a nice idea why this may have happened in a comment above - to do with tablebase and chess being solved for 7 pieces or less. Do have a look.
NM? I thought Canty was a GM...🤔
Dragon deez nutz across the board?
yes
Is this Tall
Most chess wins are Pyrrhic in nature........hmmmm 🤔🤨
Very nice;)
he talk to children ? or to stupid ? ... Had some confidence, man !
Hello 💪💪
Best commenter
first
first comment
Love these