You're right that Lc0 plays like a human, well a superhuman, and that way we can actually learn some strategies. With SF, we mostly learn how bad we all are at Chess.
@@youuuuuuuuuuutube I hear the phrase that lczero plays like a human being tossed around a lot... Or it plays positional Whereas SF plays tactical... Truth of the matter is that they both play very alien chess. The results of the search they perform is what you are witnessing. Neither one of them are thinking like a human would about the game. At least SF is thinking "what if I played my best move against what I think your best move is" and it is getting help from its Neural Net to consider the resulting positions as being more or less favorable. Sometimes it will get very far in the calculation and realize that... "Damn I'm losing right now and there's nothing I can do about it" Bro, Leela, thinks like: imagine if I just set up a random five piece end game table base that I know is winning for me and ask you to recreate the game that got us here. Because, I already played all the games that could have gotten us here and can easily recall them with all their nuances, I know which moves are the best for you to play to get to here and I know which moves are a trap that lead to worse positions for you. You're playing minesweeper and Leela already calculated where the mines most likely will be with each move. You're getting the crumbs from those calculations every move and thinking you can learn its strategies? Lol good luck. No human does this much calculating on every move... We're more likely to remember patterns or sequences of moves... These two take these ideas to their nth degrees.
@@Jozarovschesschannel Yes Danny/Josip, I find engines absolute fascinating and *additive!* I do appreciate top human games but I also love chess on another level.
OMG Leela is such a positional monster! Absolutely fascinating to watch these games played at such an insanely high level. What a time to be a chess fan! In the old days, people would get these masterpieces once a month in a magazine. We get dozens, if not hundreds, in a single day! Thanks for bringing these beautiful masterpieces to us, Jozarov!
Absolutely! I love your videos and I can see from your view counts that many others do as well. Keep up the good work and enthusiasm! It's fantastic entertainment. -A fan in the US
Dear Jozarov, I would like to mention that Stockfish 17 adopted the exact same strategy of blocking the queenside and then evakuate his king to the queenside in order to attack on the kingside in it's own game against Stormphrax in the TCEC Swiss 7 Event, where the exact same opening was played, the Mikenas-Carls-Variation, where Stockfish played in the style of Capablanca, suffocating his enemy! And furthermore, Stockfish 17 played the exact same sacrifice with it's knight hop on f5!
It would be nice if you could make a note of where the pre-arranged moves end and where the engines take over. For additional clarity, you could refer to the sides as "White" and "Black" instead of which engines are playing, reserving their names until they start making moves on their own. Overall, your commentary is very nice. And I enjoy watching Leela's games because I always seem to learn something from them. Even if it's minor, and my chess skills are terrible. But it's still fun.
Leela has defeated Stockfish twice to win the TCEC Grand Championship, three times to take the TCEC Cup, and now twice in a row to win the TCEC Swiss. So let's not say Stockfish has never had *any* competition. "The first time Stockfish did not win a TCEC event in a long time". Well, since TCEC Swiss 6 in February, seven months ago.
Leela was developed on alpha zero principles. It's an extremely creative and smart thinking engine which understands some positions better than stockplankton.
Hi Jozarov, hope you're doing well. Even though LC0 won the event, please note that Stockfish and LC0 did not play the exact same openings (for example in round 1 Stockfish played the old benoni which at computer level is a dead lost opening nowadays, while LC0 played the Nimzowitsch Larsen attack). My point being is that in order to make a fair competition, both engines should play the same pre arranged openings. I think we will have to wait until the season 27 superfinal to see which one is actually stronger.
@@shrirangbondale2263 Yes, you are right that they play the same opening twice but that still makes my point valid: the Old Benoni with black pieces at computer level is basically lost. Stockfish won with white and lost with black, thus scored only one point. The Nimzo Larsen does not result in the same evaluation: it is holdable with black pieces: Lc0 scored 1.5 after these 2 games and this generates a half point imbalance mainly driven by the fact that they played different openings.
Here's what's going on,once black pushed those two center pawns it's my game and white played the elo factor and what black doesn't understand about locking and closing positions and pawns and using the gate they never took why when it happened and used that understanding I was the one who knows what this creation was and is for I'm the chief and head programmer and they still do not understand so let the chief and head programmer show you a true authentic original and first true deep real insight since 2005,when there's a place of change that is greater than the whole thing could be where the arrangement becomes the situation of it's self in that place holds the key can be the greatest understanding you'll ever need.😮
Jozarov was so far the most complete when it comes to explaining those disgusting engine lines. I heard one famous streamer mentioned something like "Jozarov explained it on his video".. I guess he's watching your videos bro😂😂😂
@@Jozarovschesschannel 4 years ago, Agadmator And Rajavov played against you bro. It was even broadcasted live on Agadmator's channel. Have you seen it?
@@Jozarovschesschannel I'd love to see your analysis about that game, you played against raja, and at some point, raja said, " oh my God. I have to work on my chess. 😂😂
@@mgoogyinot the best explanation. LC0 was created based on the concept of A0 using the publication. It employs a zero learning algorithm and a Monte Carlo tree structure search like A0. The difference between it and A0 besides LC0’s many optimizations and upgrades is that is uses matrix mathematics used in GPU hardware to calculate whereas A0 used much more expensive and highly specific hardware called TPUs tensor processor units to do tensor math. Modern GPUs have tpu chips in their architecture to do these calculations. I’m sure Leela has been optimized to utilize these functions. A0 and Leela were made differently from different sources but based on the same research. But they use totally different hardware.
Stockfish 17 is clearly stronger, and I'm not saying this as an opinion, but rather as someone who followed its development and tested its strength along the way on my computer. Between the release of each version of Stockfish, development versions are released, and people can download and test those versions. So, you can witness the next version being increasingly stronger than the last stable one
I know that’s what I was thinking it seemed Stockfish 16 was winning a large majority but we’ve seen many a Stockfish 17 game where the fish struggles but the goal would be to make the engine stronger maybe LC0 is also improving
Moussakaba maybe you are right but to me SF17 seems to be less stabile than SF16.1. Those blunders in the middlegame/endgame are at least unexpected from a versione that should be stronger.
Great Game LC0 a great control on the board it was kind of funny how Lc0 starts an attack from one side of board than the other I knew Bxh5! sacrifice was coming from early on to open the Komodo’s king defense Thanks Jozarov for the game and info 😊♟️‼️
You're right that Lc0 plays like a human, well a superhuman, and that way we can actually learn some strategies. With SF, we mostly learn how bad we all are at Chess.
@@youuuuuuuuuuutube I hear the phrase that lczero plays like a human being tossed around a lot... Or it plays positional Whereas SF plays tactical... Truth of the matter is that they both play very alien chess. The results of the search they perform is what you are witnessing. Neither one of them are thinking like a human would about the game. At least SF is thinking "what if I played my best move against what I think your best move is" and it is getting help from its Neural Net to consider the resulting positions as being more or less favorable. Sometimes it will get very far in the calculation and realize that... "Damn I'm losing right now and there's nothing I can do about it"
Bro, Leela, thinks like: imagine if I just set up a random five piece end game table base that I know is winning for me and ask you to recreate the game that got us here. Because, I already played all the games that could have gotten us here and can easily recall them with all their nuances, I know which moves are the best for you to play to get to here and I know which moves are a trap that lead to worse positions for you. You're playing minesweeper and Leela already calculated where the mines most likely will be with each move. You're getting the crumbs from those calculations every move and thinking you can learn its strategies? Lol good luck. No human does this much calculating on every move... We're more likely to remember patterns or sequences of moves... These two take these ideas to their nth degrees.
Agadmator:Do not watch engine chess games it's like chess coke.
Also Jozarov: Here's some chess coke
It is really like an adiction ;)
@@Jozarovschesschannel Yes Danny/Josip, I find engines absolute fascinating and *additive!* I do appreciate top human games but I also love chess on another level.
If that's the case then some of the young rising stars are definitely high, they routinely play what used to be considered "computer moves".
@@JerehmiaBoazEngines play the best moves more so than any GM which puts the engines on a higher echelon of chess
@@michaelmassaro4375 I'm a stockfish contributor, I know how strong engines are.
OMG Leela is such a positional monster! Absolutely fascinating to watch these games played at such an insanely high level. What a time to be a chess fan! In the old days, people would get these masterpieces once a month in a magazine. We get dozens, if not hundreds, in a single day! Thanks for bringing these beautiful masterpieces to us, Jozarov!
Thanks for this beautiful comment
Absolutely! I love your videos and I can see from your view counts that many others do as well. Keep up the good work and enthusiasm! It's fantastic entertainment. -A fan in the US
Gorgeous. And the concept of the closing the posistion and going for the queenside castle in the kings indian is magnificent
Dear Jozarov, I would like to mention that Stockfish 17 adopted the exact same strategy of blocking the queenside and then evakuate his king to the queenside in order to attack on the kingside in it's own game against Stormphrax in the TCEC Swiss 7 Event, where the exact same opening was played, the Mikenas-Carls-Variation, where Stockfish played in the style of Capablanca, suffocating his enemy!
And furthermore, Stockfish 17 played the exact same sacrifice with it's knight hop on f5!
I will cover the game also
Indeed a winderful strategy from Leela. it is good to see the fish finally having some real competition. Thanks Josip.
In this game white had a huge advantage in space that eventually translated to won material.
I hope Lc0 continue and beating more times Stockfish
Ésa máquina juega terriblemente bien.gracias Jozarov.Saludos.
I discovered your channel gew days ago, I'm hooked. Thanks for that great content, for these dry analysis of those beautiful games. GG!!
Thanks for joining
It would be nice if you could make a note of where the pre-arranged moves end and where the engines take over. For additional clarity, you could refer to the sides as "White" and "Black" instead of which engines are playing, reserving their names until they start making moves on their own.
Overall, your commentary is very nice. And I enjoy watching Leela's games because I always seem to learn something from them. Even if it's minor, and my chess skills are terrible. But it's still fun.
Ok, will do that in the future
Love me some engine chess!
In the silly chase after ELO in bullet games, they managed to ruin it.
I said SF 17 will lose - because it is weaker than SF 16.1 - unfortunately...
а Stockfish 11 в домашних условиях помогает изучать дебют Ферзя - играя с самим собой но первые два-три хода выставляются -
партия 1 -
1. e4 e5 2. Qe2 Nc6 3. Nf3 Nf6 4. c4 Bc5 5. h3 O-O 6. d3 d6 7. g3 a6 8. Bg2 b5 9. b3 b4 10. O-O Bd7 11. Qd1 a5 12. Kh2 Nd4 13. a4 Ne6 14. Nbd2 Bc6 15. Ne1 Bd4 16. Rb1 Bb7 17. Ndf3 Bc3 18. Nc2 c6 19. Bd2 Bd2 20. Qd2 Qe7 21. Ne3 Qd7 22. Nf5 c5
23. Ng1 Ra6 24. f4 Nd4 25. Ne3 Qd8 26. f5 Ra8 27. g4 Nd7 28. Qf2 h6 29. h4 Kh7 30. Nf3 Rc8 31. g5 hg5 32. Bh3 g4 33. Ng5 Kh8 34. Bg4 Nf6
35. Rg1 Qe7 36. h5 Rg8 37. Bd1 Rcf8 38. Qh4 Nd7 39. Rb2 Bc6 40. Kh1 Qd8 41. Rh2 Nf6 42. Qg3 Qd7 43. h6 g6 44. Nxf7+ Qxf7 45. fxg6 Rxg6 46. Qxg6 Qxg6 47. Rxg6 Ne6 48. Nd5 Bxd5 49. exd5 Nf4 50. Rg3 Ng8 51. Bg4 Ne7 52. Be6 Rf6 53. Rg7 Nf5 54. Bxf5 Rxf5 55. Rd7 Nxd3 56. Rxd6 Rf8 57. Re6 Ne1 58. Re2 Nd3 59. Re3 Nf4 60. Ra6 Nh5 61. Rxa5 Kh7 62. Rxc5 Kxh6 63. Rc6+ Nf6 64. Rxe5
1-0
партия 2 - 1. e4 e5 2. Qe2 Nc6 3. Nf3 Nf6 4. c4 Bc5 5. h3 d6 6. d3 O-O 7. g3 a6 8. Bg2 Re8 9. Nc3 Nd4 10. Nxd4 Bxd4 11. O-O h6 12. Qc2
Bd7 13. Ne2 Ba7 14. b3 a5 15. Kh2 b5 16. f4 bxc4 17. bxc4 Rb8 18. f5 Qc8 19. g4 Bc5 20. Bd2 Qa6 21. Bf3 a4 22. h4 a3 23. Rab1 Qa4 24. Qc1 Nh7 25. g5 hxg5 26. hxg5 Qa7 27. Rxb8 Rxb8 28. Kg2 Rb2 29. Rh1 Be8 30. Rh3 Bg1 31. d4 Bxd4 32. Qh1 Bg1 33. Kg3 Qf2+ 34. Kg4 Rxd2 35. Rxh7 Kf8 36. Rh8+ Ke7 37. f6+ gxf6 38. gxf6+ Kd7 39. Nxg1 Qf1 40. Ne2 Qxh1 41. Rxh1 Kc6 42. Nc3 Kc5 43. Rc1 Bd7+ 44. Kg3 c6 45. Be2 Be6 46. Kf2 Kd4 47. Kf3 Rb2 48. Bf1 Rh2 49. c5 Rh6 50. cxd6 Rxf6+ 51. Ke2 Bc4+ 52. Kd2 Rxd6 53. Kc2 Bxf1 54. Rd1+ Kc5 55. Rxf1 Rd7 56. Rf3 Rc7 57. Nb1 Ra7 58. Nd2 Kd4 59. Rd3+ Kc5 60. Rc3+ Kb5 61. Nb1 Kb4 62. Rxc6 Kb5 63. Rc8 Rd7 64. Kb3 Ka6 65. Nxa3 Rd4 66. Rc4 Rd3+ 67. Rc3 Rd4 68. Re3 Rd1 69. Nc4 Rb1+ 70. Ka4 Rf1 71. Kb4 Ka7 72. Ra3+ Kb8 73. Nd6 Kc7 74. Kc5 Rc1+ 75. Nc4 Rc2 76. Ra6 Kb7 77. Ra5 Kc7 78. a4 1-0
партия 3 - 1. e4 e5 2. Qe2 Nc6 3. Nf3 Be7
4. c4 Nf6 5. d3 d6 6. h3 Nd7 7. Be3 O-O 8. Nc3 Nc5 9. g3 f5 10. exf5 Bxf5 11. O-O-O Ne6 12. Bg2 a5 13. Kb1 h6 14. a3 Bf6 15. g4 Bh7 16. h4 Ncd4 17. Qe1 Qe8 18. g5 Bd8 19. Rh3 Nf5 20. gxh6 gxh6 21. Nd2 c6 22. Be4 Kh8 23. Qh1 Nf4 24. Rf3 Bc7 25. Rg1 Rd8 26. Ka1 Rf7 27. h5 Qf8 28. d4 d5 29. cxd5 cxd5 30. Bxf5 Rxf5 31. Bf4 ef4 32. Rh3 Qf7 33. Nf3 Re8 34. Rc1 Bb8 35. Qd1 Qd7 36. a4 Rf6 37. Rh1 Rb6 38. Rg1 Rb4 39. b3 Be4 40. Nh4 f3 41. Qd2 Kh7 42. Ka2 Qh3 43. Nxe4 Rxe4 44. Qc3 Rb6 45. Qc5 Rf6 46. Qxd5 Qe6 47. Qxb7+ Qe7 48. Qxe7+ Rxe7 49. Kb1 Re4 50. Rce1 Re2 51. Nxf3 Rxe1+ 52. Nxe1 Rxf2 53. Rh1 Kg8 54. Nc2 Bd6 55. Rg1+ Kf7 56. Ne3 Bb4 57. Ng4 Rd2 58. Nxh6+ Ke6 59. Re1+ Kd7 60. Nf7 Rh2 61. Re5 Rh3 62. Kc2 Rc3+ 63. Kb2 Rd3 64. Rb5 Kc7 65. Ne5 Rh3 66. Nc4 Bc3+ 67. Kc2 Bxd4 68. Rd5 Bf6 69. Rc5+ Kd7 70. Nxa5 Ba1 71. Rd5+ Ke6 72. Rb5 Bh8 73. Nc4 1-0
Leela has defeated Stockfish twice to win the TCEC Grand Championship, three times to take the TCEC Cup, and now twice in a row to win the TCEC Swiss. So let's not say Stockfish has never had *any* competition. "The first time Stockfish did not win a TCEC event in a long time". Well, since TCEC Swiss 6 in February, seven months ago.
Leela was developed on alpha zero principles. It's an extremely creative and smart thinking engine which understands some positions better than stockplankton.
Hi Jozarov, hope you're doing well. Even though LC0 won the event, please note that Stockfish and LC0 did not play the exact same openings (for example in round 1 Stockfish played the old benoni which at computer level is a dead lost opening nowadays, while LC0 played the Nimzowitsch Larsen attack). My point being is that in order to make a fair competition, both engines should play the same pre arranged openings. I think we will have to wait until the season 27 superfinal to see which one is actually stronger.
They play same opening from white and black to get clear results as far as I know
Leela is great at longer formats
@@shrirangbondale2263 Yes, you are right that they play the same opening twice but that still makes my point valid: the Old Benoni with black pieces at computer level is basically lost. Stockfish won with white and lost with black, thus scored only one point. The Nimzo Larsen does not result in the same evaluation: it is holdable with black pieces: Lc0 scored 1.5 after these 2 games and this generates a half point imbalance mainly driven by the fact that they played different openings.
It's about time.
Stockfish didn't win neither the 1st nor 2nd TCEC swiss events. It's a pretty luck based event, Stockfish not winning is not a surprise.
d6 is known to be a bad move, how could a 3500 play it??
opening moves are set by humans to make games more interesting
Here's what's going on,once black pushed those two center pawns it's my game and white played the elo factor and what black doesn't understand about locking and closing positions and pawns and using the gate they never took why when it happened and used that understanding I was the one who knows what this creation was and is for I'm the chief and head programmer and they still do not understand so let the chief and head programmer show you a true authentic original and first true deep real insight since 2005,when there's a place of change that is greater than the whole thing could be where the arrangement becomes the situation of it's self in that place holds the key can be the greatest understanding you'll ever need.😮
Jozarov was so far the most complete when it comes to explaining those disgusting engine lines. I heard one famous streamer mentioned something like
"Jozarov explained it on his video"..
I guess he's watching your videos bro😂😂😂
Which one
@@Jozarovschesschannel as far as I can recall, it was Agadmator, not so sure but I guess it was him..He's watching your videos bro.
He is my brother in law
@@Jozarovschesschannel 4 years ago, Agadmator And Rajavov played against you bro. It was even broadcasted live on Agadmator's channel. Have you seen it?
@@Jozarovschesschannel I'd love to see your analysis about that game, you played against raja, and at some point, raja said, " oh my God. I have to work on my chess. 😂😂
imagine the queen's last move would be F7, not F8. Just for fun making stalemate
Stockfish works like a calculator leela is much more creative in its games
Stockfish devs are not happy 💀
They don't care
Is Lc0 advance version of alpha0 ?
No. Lc0, stockfish and other top engines are trained similarly as alpha0 learned to play.
@@mgoogyinot the best explanation. LC0 was created based on the concept of A0 using the publication. It employs a zero learning algorithm and a Monte Carlo tree structure search like A0. The difference between it and A0 besides LC0’s many optimizations and upgrades is that is uses matrix mathematics used in GPU hardware to calculate whereas A0 used much more expensive and highly specific hardware called TPUs tensor processor units to do tensor math. Modern GPUs have tpu chips in their architecture to do these calculations. I’m sure Leela has been optimized to utilize these functions.
A0 and Leela were made differently from different sources but based on the same research. But they use totally different hardware.
@@chrisdavis2161 Thanks for the clarification.
There is no proof of SF17 loss. Title is wrong.
Since Lc0 is at 1st place
Will we use Lc0 eval bars?
As good as chess gets
Until next year
Juste wooooooowww
C'est plutôt stockfish 16 pas 17!
LC0 vs. Komodo vs. the clickbaity headline.
#parse error
Stockfish 17
Stockfish 17 is clearly stronger, and I'm not saying this as an opinion, but rather as someone who followed its development and tested its strength along the way on my computer. Between the release of each version of Stockfish, development versions are released, and people can download and test those versions. So, you can witness the next version being increasingly stronger than the last stable one
I know that’s what I was thinking it seemed Stockfish 16 was winning a large majority but we’ve seen many a Stockfish 17 game where the fish struggles but the goal would be to make the engine stronger maybe LC0 is also improving
@@moussakaba7676finally someone who knows ❤
Moussakaba maybe you are right but to me SF17 seems to be less stabile than SF16.1. Those blunders in the middlegame/endgame are at least unexpected from a versione that should be stronger.
Great Game LC0 a great control on the board it was kind of funny how Lc0 starts an attack from one side of board than the other I knew Bxh5! sacrifice was coming from early on to open the Komodo’s king defense Thanks Jozarov for the game and info 😊♟️‼️