|
< Earlier Kibitzing · PAGE 2 OF 2 ·
Later Kibitzing> |
Dec-17-18
 | | keypusher: <WorstPlayerEver: <moronovich>
Of course. Considered that this is only the beginning. What disturbs me, however, is the way AO is advertised by Google; if it is really that good, that seems unnecessary.> Explain. Clearly, concisely, and coherently, please. |
|
| Dec-17-18 | | The Kingfish: <keypusher: Note to <worstplayerever>: If you think the games are fixed, just say so. The hints and insinuations make you look like a complete idiot.> +1 |
|
Dec-25-18
| | keypusher: This is probably the most mind-blowing of all the A0-SF games. Here are a couple more videos. https://m.youtube.com/watch?v=bo5pl...
Saddler goes too fast but points out some amazing ideas, such as 36.....Re8 37.Rxe8 Nxe8 38.b6 Qf7 39.Ba6. |
|
Dec-25-18
| | keypusher: And here is <adagmator>. He’s very funny (“overextend on the queen side, mess up your kingside, and you’ll have a very good game — that’s what A0 is teaching us”). https://m.youtube.com/watch?v=wui0Y... |
|
Jan-17-19
 | | tpstar: https://youtu.be/iENQXIQ8wH0
[Fritz 10]: 1. Nf3 Nf6 2. c4 e6 3. Nc3 Bb4 4. Qc2 0-0 5. a3 Bxc3 6. Qxc3 a5 7. b4 d6 [last book move] 8. e3 Ne4 [8 ... e5 9. d3 =] 9. Qc2 Ng5 10. b5 Nxf3+ 11. gxf3 Qf6 12. d4 Qxf3 13. Rg1 Nd7 14. Be2 Qf6 15. Bb2 Qh4 16. Rg4 Qxh2 17. Rg3 f5 18. 0-0-0 ["He knows no fear" - 18. Bf1 Qh5 =] Rf7 [18 ... Qxf2!? 19. Rdg1 Rf7 ⩱] 19. Bf3 [=] Qh4 20. Rh1 Qf6 [20 ... Qe7 21. e4 g6 22. Rgg1 =] 21. Kb1 g6 22. Rgg1 [22. Be2 Qe7 ⩲] a4 [22 ... Qe7 23. e4 =] 23. Ka1 [23. Rh6 Nb6 ⩲] Rg7 [23 ... Qe7 24. e4 ⩲] 24. e4 [24. Be2 Qe7 ⩲] f4 [24 ... fxe4 25. Bxe4 Rb8 26. f3 ⩲] 25. c5 Qe7 26. Rc1 Nf6 27. e5 dxe5 28. Rhe1 e4 29. Bxe4 Qf8 30. d5 [30. Qc4!? c6 31. b6 ±] exd5 [=] 31. Bd3 Bg4 32. f3 [32. Qd2 Re8 33. Qxf4 Rxe1 34. Rxe1 Rf7 =] Bd7 [32 ... Bh5!? ∓] 33. Qc3 [=] Nh5 34. Re5 c6 35. Rce1 Nf6 [Weaker is 35 ... cxb5 36. Rxh5 gxh5 37. Qxg7+ Qxg7 38. Bxg7 Kxg7 39. Re7+ Kf6 40. Rxd7 ⩲] 36. Qd4 [36. b6!? =] cxb5 [ ⩱] 37. Bb1 Bc6 38. Re6 Rf7 [38 ... Nh5 ∓] 39. Rg1 [ ⩲] Qg7 40. Qxf4 Re8 [40 ... Bd7!? would keep Black alive 41. Re5 Rff8 ⩲] 41. Rd6 [ ±] Nd7 42. Qc1 Rf6 43. f4 Qe7? [43 ... b4 44. axb4 a3 +-] 44. Rxf6 [+-] Nxf6 45. f5 Qe3 [45 ... Kg7 46. Bd4 b6 47. fxg6 +-] 46. fxg6 Qxc1 47. gxh7+ Kf7 48. Rxc1 Nxh7 49. Bxh7 Re3 50. Rd1 Ke8 51. Ka2 Bd7 [51 ... Rh3 52. Bg6+ Ke7 53. Re1+ Kd8 54. Bf6+ Kc7 55. Be5+ Kd8 56. Rf1 +-] 52. Bd4 [52. Bg8 Kd8 +-] Rh3 53. Bc2 Be6 54. Re1 Kd7 55. Kb2 Rf3 56. Re5 Rg3 57. Re3 Rg2 [57 ... Rg4 58. Kc3 +-] 58. Kc3 Rg4 [58 ... b4+ 59. axb4 a3 60. Ba4+ Ke7 61. Bb3 +-] 59. Rf3 [59. Rh3 Rf4 60. Rh7+ Kc6 +-] Ke8 60. Rf2 Rg3+ 61. Kb4 Rg4 62. Rd2 Bd7 63. Ka5 Rf4 64. Be5 Rf3 65. Rd3 Rf2 66. Bd1 Bc6 67. Kb6 1-0 Ah yes, the glory of using Fritz to analyze two other engines. =) This is the second computer game to ever gain a spot in the Best Games contest, with the first being AlphaZero vs Stockfish, 2017 last year. In this Nimzo-English line, 8 ... e5 looks good for Black trying to capitalize on no Pd4 to contest the center. White casually sacrifices two Pawns on the Kingside and offers a third (18. 0-0-0 Qxf2) with compensation in the Two Bishops plus wide open lines against the Black King. After 20. Rh1 Qf6 things look grim for Black, yet White proceeds patiently with 23. Ka1 followed by relocating Rooks to c1 & e1. Interesting that Fritz prefers Black after the suggestions 32 ... Bh5!? and 38 ... Nh5 instead. 42. Qc1 is an amusing retreat with White's pieces clustered in the corner. The only clear error noted was 43 ... Qe7?, then I will let other analysts determine whether 48 ... Nxh7 was absolutely necessary. Good game. =) |
|
| Mar-26-20 | | Everett: Hey, I’m only 1750 USCF, but I wouldn’t play 17.f5 I’d extricate my Q and get my Q-side pieces into play. Basic chess |
|
Mar-27-20
| | keypusher: < Everett: Hey, I’m only 1750 USCF, but I wouldn’t play 17.f5
I’d extricate my Q and get my Q-side pieces into play. Basic chess> I am nearly as quick to jump to A0’s defense as I am to Botvinnik’s…clearly I like machines. A few posters have criticized 17….f5; on the other hand a 3000+ engine picked it, and GMs annotating the game (there have been quite a few) generally pass over it in silence. There’s a basic justification for it: in the fairly near future, White plans to have both rooks, both bishops and the queen pointed at the pawns on g7 and h7. They will need help. f7…f5 obstructs the b1-h7 diagonal, gives the queen or a sheltered square at f6, and opens f7 for the rook to defend from the side. Also, playing ...f5 doesn't rule out extricating the queen (which SF does immediately after) or developing the queenside pieces. I had SF10 look at various possibilities around move 17 for eight hours or more. Left to its own devices, it always plays …f5 sooner or later, and …g6 as well. Top lines at move 17:
17... Qh1+ 18. Bf1 f5 19. O-O-O Qh4 20. Kb1 g6 21. Bd3 Rf7 22. Rdg1 Kf8 23. c5 Nf6 24. Qc1 Ne4 25. Rh1 Qe7 26. Rxg6 Nxf2 27. Qg1 hxg6 28. Rh8+ Kg7 29. Qh2 Qg5 30. d5+ e5 31. Rh7+ Kg8 32. Rh8+ = (0.00) Depth=57/74 0:484:41 52685 MN 17... f5 18. O-O-O Nf6 19. Bf3 Ne4 20. Bxe4 fxe4 21. Rdg1 g6 22. Qxe4 Rf7 23. d5 e5 24. Rxg6+ hxg6 25. Qxg6+ Kf8 26. Qg8+ Ke7 27. Qg5+ Kd7 28. Qg4+ Ke8 29. Qg8+ Rf8 30. Qg6+ Kd8 31. Qg5+ Ke8 = (0.00) Depth=56/30 0:484:41 52685 MN 17... g6 18. Bf3 Qh4 19. O-O-O Qe7 20. Rh1 Re8 21. Be4 c6 22. f4 f5 23. d5 Nf6 24. dxe6 Qxe6 25. Bf3 Bd7 26. Qh2 Re7 27. Rxg6+ hxg6 28. Qh8+ Kf7 29. Rh7+ Nxh7 30. Qxh7+ Kf8 31. Qh6+ Rg7 32. Qxg7+ Ke8 33. bxc6 Qxe3+ 34. Kc2 Qf2+ 35. Kb1 Qf1+ 36. Kc2
= (0.00) Depth=56/69 0:484:41 52685 MN
17... Re8 18. Bf3 Qh4 19. O-O-O Qe7 20. Rh1 g6 21. Be4 c6 22. f4 f5 23. d5 Nf6 24. dxe6 Qxe6 25. Bf3 Bd7 26. Qh2 Re7 27. Rxg6+ hxg6 28. Qh8+ Kf7 29. Rh7+ Nxh7 30. Qxh7+ Kf8 31. Qh6+ Rg7 32. Qxg7+ Ke8 33. bxc6 Qxe3+ 34. Kc2 Qf2+ 35. Kb1 Qf1+ 36. Kc2
= (0.00) Depth=56/43 0:484:41 52685 MN
Note that White is getting a strong-looking attack in those lines, though SF thinks it can hold. If you make it extricate the queen and try to get the q-side pieces into play, SF thinks that is worse. 17….Qh4 18.Bf3 Qe7 19.0-0-0 Nf6:
Top line:
20. Rh1 g6 21. e4 e5 22. Bd1 Rd8 23. Qd2 Kf8 24. Qh6+ Ke8 25. Bc2 Bg4 26. c5 a4 27. b6 Rdc8 28. f3 Bh5 29. dxe5 dxe5 30. Qg7 cxb6 31. Rg5 Ng4 32. Rhxh5 gxh5 33. fxg4 Rxc5 34. Rxh5 f6 35. Qg8+ Qf8 36. Qxh7 Rac8 37. Qg6+ Kd8 38. Rh2 Kc7 39. Kb1 b5 40. Rf2 Qd8 41. Qxf6 Qxf6 42. Rxf6 Rh8 43. Bd3 Rh3 44. Be2 Rg3 45. Rf5 Re3 46. Bf1 Rxe4 47. g5 b4 48. axb4 Rxb4 49. Ka1 Rb3 50. g6 Rg3 51. Bxe5+ Rxe5 52. Rxe5 Rxg6 53. Kb2 Kb6 54. Rb5+ Kc7 55. Ka3 Rg3+ 56. Kxa4 Kc6 57. Kb4 (0.62) Depth=53/82 0:549:28 59306 MN
After 36.Qxh7: 
click for larger viewInteresting position! No idea what’s going on for some of those moves in the line above, but note that SF feels obligated to get its king away from the kingside pronto. |
|
| Mar-27-20 | | SChesshevsky: <...I wouldn't p!at 17...f5...> Seems the idea SF has is grab the two pawns and try to set up some kind of middle game blockade. So in that case ...f5 does make some sense. It looks like from there SF makes every attempt to just keep the position closed. It's all very passive and allows A0 all the time in the world to open up with at least positional advantage. One thing to note is that when you're two pawns up and the eval doesn't give you a major plus your position is probably so critical it might be busted with even small imprecision. Or busted just because the horizon is just too short. |
|
| Mar-28-20 | | Everett: thanks for your responses <keypusher> and <SChesshevsky> I'm not a pawn grabber either, so I likely wouldn't be in that position anyways! Looks to me Stockfish bit off more than it could chew, grabbing pawns and creating half-open files in front of his castled king with its Q-side sleeping. The entire sequence from Black's 8th-12th would be reprimanded by many, it seems. Anything wrong with 8..e5 and normal development? Anyways, interesting game. |
|
Mar-28-20
| | keypusher: Part I
< SChesshevsky: <...I wouldn't p!at 17...f5...>
Seems the idea SF has is grab the two pawns and try to set up some kind of middle game blockade. So in that case ...f5 does make some sense. It looks like from there SF makes every attempt to just keep the position closed.It's all very passive and allows A0 all the time in the world to open up with at least positional advantage. One thing to note is that when you're two pawns up and the eval doesn't give you a major plus your position is probably so critical it might be busted with even small imprecision. Or busted just because the horizon is just too short.> > Yes, unlike in the game you and I discussed at length, AlphaZero vs Stockfish, 2018 where SF was handed an unfavorable opening setup, here it was on its own. I assume this game is from the January 2018 1,000 game match from the initial starting position that A0 won +155-6=839. As a lot of people pointed out back when the 2017 SF-A0 games came out, SF isn't designed to play without an opening book. I don't mind, because we got games like this out of it, but it was definitely "unfair" to SF to make it play that way. In the Caro-Kann game, you'll remember SF never "thought" it had the advantage. It grabbed a pawn because it thought that was the best continuation available, but it never gave itself a positive eval. In this game, as you may recall, at move 16, at 41 ply SF gives the position a 0.00 evaluation. I did some long searches with SF10 earlier in the opening, specifically after 7.b4 (to see how SF “feels” about the opening in general); after 9….Ng5 (just before A0’s first pawn offer); and after 10.b5 (just after the pawn offer). Then, in response to <Everett>’s query, I ran a long search after 8….e5 and posted a game in which SF played that way — see next post. Basic point is that SF thinks it is slightly worse in the opening. Its top lines at move 7 are: 7... Re8 8. e3 e5 9. Bb2 d6 10. Be2 Bg4 11. d3 Nbd7 12. O-O Bh5 13. h3 h6 14. Rab1 c6 15. Ra1 Qe7 16. Rfb1 Bg6 17. Rc1 Bh5 18. Qc2 Qe6 19. Rd1 Bg6 20. Nd2 Bh5 21. Bxh5 Nxh5 22. d4 Nhf6 23. dxe5 dxe5 24. Qc3 Qf5 25. Rac1 Nb6 26. Qb3 axb4 27. Qxb4 Na4 28. Nf3 Nxb2 29. Qxb2 Ra5 30. Qxb7 Rxa3 31. Qxc6 Ra2 32. Rd6
+ (0.30) Depth=49/71 0:453:00 45826 MN
7... b6 8. Bb2 axb4 9. axb4 Rxa1+ 10. Bxa1 c5 11. e3 Qe7 12. b5 Bb7 13. Be2 d5 14. O-O Nbd7 15. Bb2 Ra8 16. Ra1 Rxa1+ 17. Bxa1 dxc4 18. Qxc4 Ne4 19. Qa4 e5 20. d3 Nd6 21. Nd2 Bd5 22. Bb2 h6 23. Bc3 f6 24. h3 g6 25. Nc4 Kg7 26. Bb2 Be6 27. g4 Bd5 28. Bc3 Bxc4 29. dxc4 Ne4 30. Be1 Nd6 31. Kg2 e4 32. Qa8 Ne5 33. Qd5 Nef7 34. Bc3
+ (0.35) Depth=49/74 0:453:00 45826 MN
7... d6 8. e3 e5 9. Bb2 Bg4 10. Be2 Bh5 11. d3 Re8 12. O-O Nbd7 13. Qc2 h6 14. Bc3 Qe7 15. h3 Bg6 16. e4 axb4 17. axb4 c6 18. Rxa8 Rxa8 19. Ra1 Rxa1+ 20. Bxa1 b6 21. Qb2 b5 22. cxb5 cxb5 23. Nh4 Bh7 24. Qc3 d5 25. Qc8+ Qf8 26. Qxf8+ Kxf8 27. d4 dxe4 28. dxe5 Nd5 29. Bxb5 N7b6 30. g4 Nxb4 31. f4 N6d5 32. Bd4 Nd3 33. f5 N3f4 34. Kh2 Nd3
+ (0.41) Depth=49/73 0:453:00 45826 MN
7... Qe7 8. Bb2 b6 9. g3 axb4 10. axb4 Rxa1+ 11. Bxa1 c5 12. bxc5 bxc5 13. Bg2 d6 14. O-O e5 15. d3 Nc6 16. Qd2 Bg4 17. Rb1 Qd7 18. Ng5 h6 19. Ne4 Nxe4 20. Bxe4 Nd4 21. Bxd4 exd4 22. e3 dxe3 23. Qxe3 Be6 24. h4 Qc7 25. Rb7 Qd8 26. Kg2 d5 27. cxd5 Bxd5 28. Rb5 Bxe4+ 29. dxe4 Re8 30. Rxc5 Qb6 31. Rc4 Qxe3 32. fxe3 g6 33. Kf3 Kg7 34. Rc7 Re5 35. Rd7 Kf6 36. Rd5
+ (0.49) Depth=49/72 0:453:00 45826 MN
7... c6 8. e3
+ (0.40) Depth=48/68 0:453:00 45826 MN |
|
Mar-28-20
| | keypusher: Part II
<Everett>
<The entire sequence from Black's 8th-12th would be reprimanded by many, it seems. Anything wrong with 8..e5 and normal development?> SF “thinks” that 8.e3 is a minor inaccuracy, allowing it to trade knights. I’m just going to post its top line after 8….e5 after a relatively modest search, because (i) one of the lines in my prior post tells you what SF thinks of 8…e5 — it’s going to be a little worse (ii) the game I posted below shows how playing ….e6-e5 worked out in practice. 9. Bb2 Bg4 10. Be2 Nbd7 11. d3 c6 12. h3 Bh5 13. Qc2 Qe7 14. O-O h6 15. Rfc1 Rfc8 16. Bc3 b6 17. Qd1 Qe6 18. Qe1 Bg6 19. Rd1 Qe7 20. Bb2 Bh5 21. Qd2 Ra7 22. Rdc1 Rca8 23. b5 Rc7 24. a4 Re8 25. d4 c5 26. d5 Bxf3 27. Bxf3 e4 28. Bd1 Ne5 29. Be2 Nfd7 30. Qc3 Qf6
+ (0.47) Depth=41/52 0:26:39 2754 MN
Its top lines after 8….Ne4 9.Qc2 Ng5 are below. Note that 10.b5 is on its radar screen, with an 0.00 evaluation, trailing 10.Ng1, 10.Nxg5 and 11.Bb2. 10.b5 is really a remarkable move for a machine to come up with. 10. Ng1 [yes, really!] e5 11. Bb2 axb4 12. axb4 Rxa1+ 13. Bxa1 Na6 14. h4 Ne6 15. Bc3 c5 16. b5 Nb4 17. Qb1 b6 18. Bxb4 cxb4 19. Qxb4 Bb7 20. f3 Nc5 21. Qc3 Qf6 22. d4 Ne6 23. Ne2 Ra8 24. Kf2 h5 25. dxe5 dxe5 26. Ng3 Nc5 27. Qb2 g6 28. Be2 Qe7 29. e4 Qf6 30. Qd2 Bxe4 31. Qg5 Qxg5 32. hxg5 Bb7 33. Nf1 Ne6 34. Ne3 Nxg5 35. Nd5
+ (0.21) Depth=48/72 0:482:26 49437 MN
10. Nxg5 Qxg5 11. h4 Qf6 12. Bb2 e5 13. Qe4 Qe6 14. Be2 f5 15. Qd5 axb4 16. axb4 Rxa1+ 17. Bxa1 Nc6 18. Qxe6+ Bxe6 19. Bc3 Ra8 20. O-O Ra2 21. Kh2 Ne7 22. Kg3 Kf7 23. Ra1 Rxa1 24. Bxa1 g6 25. f4 e4 26. h5 Ng8 27. hxg6+ hxg6 28. c5 Nf6 29. cxd6 cxd6 30. Bb2 Nd5 31. b5 Nf6 32. d3 exd3 33. Bxd3 Ne4+ 34. Bxe4 fxe4 35. Bd4 Ke7 36. b6 Bd5
+ (0.21) Depth=48/77 0:482:26 49437 MN
10. Bb2 Nxf3+ 11. gxf3 axb4 12. axb4 Rxa1+ 13. Bxa1 e5 14. h4 Nd7 15. Be2 Kh8 16. c5 Nf6 17. d4 Re8 18. Bb2 h6 19. dxe5 dxe5 20. c6 Qe7 21. b5 Nd5 22. cxb7 Bxb7 23. Qe4 c6 24. bxc6 Bxc6 25. Bd3 Qb4+ 26. Qxb4 Nxb4 27. Be2 f6 28. O-O Rd8 29. Rd1 Rxd1+ 30. Bxd1 Kh7 31. Be2 Be8 32. Ba3 Nd5 33. Bc4 Nb6 34. Bf1 Bg6 35. Kg2 Bf7 36. Kg3 Kg6 37. Bd3+ Kh5 38. f4 exf4+ 39. Kxf4 Kxh4
+ (0.13) Depth=48/77 0:482:26 49437 MN
10. Nd4 e5 11. h4 Ne6 12. Nxe6 Bxe6 13. Bb2 Nd7 14. Be2 f5 15. f4 h6 16. g3 Qe7 17. O-O e4 18. d3 exd3 19. Bxd3 axb4 20. axb4 c5 21. Rxa8 Rxa8 22. Bxf5 Bxf5 23. Qxf5 Qxe3+ 24. Rf2 Qxg3+ 25. Kf1 Nf6 26. Bxf6 gxf6 27. Qxf6 Qd3+ 28. Kg1 Qe3 29. Kg2 Kh7 30. Qf7+ Kh8 31. Qf6+
= (0.00) Depth=48/69 0:482:26 49437 MN
10. b5 Nxf3+ 11. gxf3 Qf6 12. Bb2 Qxf3 13. Rg1 g6 14. Rg3 Qh5 15. Be2 Qf5 16. Bd3 Qh5
= (0.00) Depth=48/15 0:482:26 49437 MN
<Everett> For an example of how things went after …e5 (played a move earlier), see the remarkable AlphaZero vs Stockfish, 2018 <SChesshevsky>
After 10.b5, 10….Nxf3+ 11.gxf3 Qf6 with an 0.00 evaluation remains SF’s choice even at enormous search depths (see below). Again, as in the Caro-Kann game, it’s not that it thinks “free pawn.” It’s that it thinks that every other move leads to a disadvantage. Which makes sense, right? White has more space and the bishop pair, so he ought to be a little bit better. It's not really a case of the horizon effect. It's a case of an opening not quite working out, saddling SF with some less than ideal choices. I think if SF could talk, that is what it would tell us. (Some of these lines are truncated to save me from having to do Part III.) 10... Nxf3+ 11. gxf3 Qf6 12. Bb2 Qxf3 13. Rg1 g6 14. Rg5 e5 15. Bg2 Qf6 16. f4 c6 17. Kf2 Nd7 18. Kg1 Qe7 19. Rf1 cxb5 20. cxb5 Nb6 21. Rg3 Bd7 22. a4 Rac8 23. Qe4 Rc4 24. Qxb7 Nxa4 25. Ba1 Nc5 26. Qc7 Rb4 27. fxe5 Rxb5 28. exd6 Qe6 29. Bh3 Qe8 30. Bg2 Qe6
= (0.00) Depth=50/81 0:483:56 49127 MN
10... b6 11. Nxg5 Qxg5 12. h4 Qe7 13. Qe4 Ra7 14. h5 Nd7 15. Bd3 f5 16. Qh4 Qxh4 17. Rxh4 e5 [snip] + (0.46) Depth=50/79 0:483:56 49127 MN
10... Nd7 11. Be2 e5 12. Bb2 Nc5 13. O-O Nxf3+ 14. Bxf3 Qg5 15. d4 exd4 16. exd4 Bf5 17. Qd1 Ne4 [snip] + (0.59) Depth=50/79 0:483:56 49127 MN
10... e5 11. Bb2 Nxf3+ 12. gxf3 Qf6 13. f4 Bf5 14. e4 Bg4 15. f5 Qh4 16. Rg1 a4 17. Rg3 Nd7 18. d4 exd4 19. Bg2 Nc5 20. f3 Bxf5 21. exf5 Rfe8+ 22. Kf1 Qxh2 23. Rh3 Qe5 24. f4 Qxf4+ 25. Qf2 Qxf2+ 26. Kxf2 [snip] + (0.77) Depth=50/77 0:483:56 49127 MN |
|
Mar-28-20
| | keypusher: <Everett>
Other examples of ..e5:
AlphaZero vs Stockfish, 2018 AlphaZero vs Stockfish, 2018 |
|
| Mar-28-20 | | Everett: <keypusher>
Thanks for all those! I see you’ve been studying these games quite a bit. I wish I had time to look at this material more closely. What I like to try to do with these various games is to see if my ideas make any sense. So it’s fun to try my hand against these monsters and see what comes up. I remember Karpov taking two beatings at the hands of a young Kramnik in this precise line. It’s quite possible that the early k-side castling is simply too committal, allowing White more space, the two Bs, and a flexible K-placement himself. Of note, White’s DSB is a main feature in many of these games, and by move 6 it’s clear it will pointing at the Black K for all eternity. In any case, I’ll take a look. Thank you! |
|
Dec-02-21
 | | saffuna: Daniel King analysis of this game:
https://www.youtube.com/watch?v=pFt...
King goes through some amazing variations were not played, and marvels how Alphazero doesn't just grab the advantage but is very patient. For example, King suggests SF played 36...cxb5 because of the danger of white playing b6 followed by Ba6: 36...Re8 37 Rxe8 Nxe8 38 Ba6. |
|
Dec-02-21
| | saffuna: Oops. Just saw that <keypusher> pointed out that variation above. |
|
Jan-08-22
 | | An Englishman: Good Evening: All quite baffling to this elderly carbon-based life form with bad eyesight, but most puzzling? 26...Nf6. Thought Black intended to play 26...Nf8, which seemed to provide better protection of the King side pawns. |
|
Jan-08-22
 | | Richard Taylor: I had a book (library book) of many games by A0 but I was bored by the games. Computer games don't interest me that much. I am not sure who can evaluate these things. The point is that we all, right up to Carlson, are humans. Carlsen was asked 'Do you play against computers?' He said: 'No, I don't like being beaten by things that are stupid, and computers are stupid.' Of course one can learn some ideas but there has to be a good reason to sacrifice two (or more) pawns in an opening. And for most of us...well we have known this, or I have...for nearly 60 years! <tp.star> who gave a long commentary here, died a few years ago. Anna I used to watch, she is always the way she is, or was. (But there is no "mystery", f5 looks logical at first for many players, although Black has been greedy and we know what is going to happen, the attack using the two Bs and the open files that Black has opened will take place and White will win. Or the Q will or may get trapped or embarrassed...). She, Anna, was good value actually when showing puzzles and talking about positional ideas. Giri's wife was did tactics and opening analysis. Both are better on the eye than, well, many men, strong players who, usually, with the exception of a few (Sam Shankland) say nothing interesting about the positions they are in...as well as not being very attractive! |
|
| Jan-08-22 | | parmetd: Tpstar passed away?! That's the first time I am hearing of it. Uscf and fide down show him deceased either. |
|
| Jan-08-22 | | TheBirdman33: <A Tribute to Dr. Tony Palmer> https://grandrapidschesscenter.org/... |
|
| Jan-08-22 | | Ironmanth: This game gives me chills. Had seen it at least once before. The artistry and cunning of Alpha Zero in opening lines is scary and breathtaking. Fearless silicon strangulation. Thanks chessgames. Y'all stay safe out there this weekend. |
|
| Jan-08-22 | | Stanco: Watching these monsters games just makes me think of chess as a done deal in a soon time.
But it's good to know how it's done!
Afterwards, humans wrote all these algorithms and all this wouldn't be possible if we don't exist. |
|
Jan-08-22
 | | Check It Out: Fallible yet exciting human chess with all it's errors waiting to be made will not end. AI and engines may get ever closer to "solving" the game, but the human mind will not. TicTacToe (or checkers) is one thing; chess is far too complex for that. |
|
| Jan-08-22 | | Stanco: Think so?
Why not human chess with no errors?
I have no problem with that if I'm given to with no time limit. |
|
Jan-10-22
| | Check It Out: <stanco: Why not human chess with no errors?> Even with you odorous "no time limit" qualifier, your question is stinky. |
|
Mar-17-22
 | | gezafan: This is a great game! At one point Alphazero was down 4 pawns yet won! I think one thing that separates the machines from us is their ability to make long term positional sacrifices that are beyond the horizon that we can see. |
|
 |
|
< Earlier Kibitzing · PAGE 2 OF 2 ·
Later Kibitzing> |
Kibitzing
