**AlphaZero vs Stockfish **

On December 5, 2017, Google DeepMind announced that it had developed an automated learning program — AlphaZero (hereafter AZ)– which trained itself to play Go, and chess, and Shogi in only a few hours and then decisively beat the world’s best programs at each:

- beat Elmo at Shogi: +90 =2 -8
- beat AlphaGO at Go: +60 =0 -40 (there are no draws in Go)
- beat Stockfish at Chess: +28 =72 -0.

This result follows DeepMind’s previous high-profile successes in mastering Go, decades before many experts thought that would be possible:

- March 2016: AlphaGo defeated 18-time Go World Champion Lee Sedol: 4-1.
- May 2017: AlphaGo defeated current world #1 Go player Ke Jie: 3-0.
- October 2017: AlphaGo Zero defeated AlphaGo by a score of 100-0.

The paper, which has not been peer-reviewed, describing the experiment, along with 10 games between AlphaZero and Stockfish can be downloaded **here.**

We’ll have more on this in a few days, including:

Q: Did AZ really take only 4 hours to learn chess?

A: Yes. Not counting the years of development in creating self-learning algorithms and the super-fast hardware to run them.

Q: Are these 10 games a fair representation of AZ’s chess ability?

A: Obviously not: AZ scored 64% in the match, but scores 100% in the published sample.

Q: Was this a fair test for Stockfish?

A: No. There are objections to the lack of opening book, no endgame tablebases, the unusual time control, the exceptionally low level of memory cache allowed to Stockfish relative to the number of threads it ran on, the fact that they used a year-old version of the constantly changing program…

Q: So, is AlphaZero *not* stronger than Stockfish?

A: It is clearly *much* stronger, though probably by less than the 64-36 score indicates.

Q: But their paper says SF was analyzing way more positions per second than AZ, doesn’t that mean SF had an advantage?

A: No. It means SF and AZ spend their calculating time differently: SF looks at lots of positions, AZ applies more time-consuming evaluations to each position.

Q: When will I be able to buy AlphaZero for chess?

A: Don’t hold your breath. Google DeepMind did not sell the programs it made for Go, and (even if it did) they all run on hardware that goes about 1000 times faster than you could run on a desktop CPU with an AMD Ryzen 9.

Q: Will AlphaZero StockPicker revolutionize investing?

A: Not for long, if at all. Stock markets, unlike boardgames, are chaotic type-2 systems: choices made by agents affect the system in unpredictable ways…

Q: Was it just a coincidence that the first round of the 2017 London Chess Classic was played at the home of Google DeepMind the same week as their paper about AlphaZero was published?

A: :-)

**Chess Games First**

There are many interesting things about this story, some of which we may return to here. But, as chess-lovers, the most interesting thing about the story is that there is now a chess-playing entity which is several orders of magnitude better than the best publicly-available programs, which are themselves several orders of magnitude better than the best human players, and that (some of) its games are available. So what does it play like??

All 10 games, with notes, can be replayed in the viewer below. If you need some encouraging, how about **7-time Russian Champion Peter Svidler**:

“The games were absolutely fantastic, phenomenal…. I’m not amazed with the fact that it learned chess, but I was stunned by the games’ quality.”

I think all 10 games are worth playing through, but if you’re short of time, then play at least the following three:

**game 3**: a QID where AZ sacrifices a pawn and an exchange to trap Black’s Q on h8 and win by zugzwang.**game 10**: another QID, where AZ allows a N to be captured at move 19 just to gain a development lead which doesn’t turn into a (humanly) clear win until after move 33.**game 7**: Karpov 2.0: no fireworks, but a Karpov-like win where AZ didn’t seem to do anything… until its opponent was lost.

**() – ()**

Round: Result:

[Event "AlphaZero vs. Stockfish"] [Site "?"] [Date "2017.12.04"] [Round "1.1"] [White "Stockfish 8"] [Black "AlphaZero"] [Result "0-1"] [ECO "C65"] [Annotator "John Upper"] [PlyCount "134"] [EventDate "2017.??.??"] [Source "CFC Newsfeed"] [SourceDate "2017.12.08"] [WhiteTeam "Norway"] [BlackTeam "England"] [WhiteTeamCountry "NOR"] [BlackTeamCountry "ENG"] {The Power of the Bishop Pair} 1. e4 e5 2. Nf3 Nc6 3. Bb5 Nf6 4. d3 Bc5 5. Bxc6 dxc6 6. O-O Nd7 7. Nbd2 O-O 8. Qe1 $146 (8. Nc4 f6 9. Kh1 a5 10. Nh4 Nb6 11. Be3 Bd4 12. Nxb6 cxb6 13. Qc1 Be6 14. f4 exf4 15. Bxd4 Qxd4 16. Rxf4 c5 17. Rf1 a4 $15 {1/2-1/2 (46) Yu,Y (2729)-Cheparinov,I (2689) Doha 2016}) 8... f6 9. Nc4 Rf7 10. a4 Bf8 11. Kh1 Nc5 12. a5 Ne6 13. Ncxe5 fxe5 14. Nxe5 Rf6 15. Ng4 Rf7 16. Ne5 Re7 17. a6 c5 18. f4 Qe8 19. axb7 Bxb7 $15 20. Qa5 Nd4 21. Qc3 Re6 22. Be3 Rb6 23. Nc4 Rb4 24. b3 {Diagram [#]} a5 $1 {Black could keep the a-pawn with ...a6, but it would be indefinitely weak (tying down the LSB) and Black's Ra8 doesn't have particularly good prospects since there are no open file. Sacing the a-pawn looks good for both the Ra8 and the LSB.} 25. Rxa5 (25. Bxd4 cxd4 26. Qxd4 a4 $15) 25... Rxa5 26. Nxa5 Ba6 27. Bxd4 Rxd4 28. Nc4 Rd8 29. g3 h6 30. Qa5 Bc8 31. Qxc7 {Diagram [#]White has four pawns and a N for the B pair... but even my laptop SF8 rates Black as better. White would have much better drawing chances if his f and g-pawns were back on the second rank.} Bh3 $1 $15 32. Rg1 Rd7 33. Qe5 Qxe5 34. Nxe5 Ra7 35. Nc4 g5 36. Rc1 Bg7 37. Ne5 Ra8 38. Nf3 Bb2 39. Rb1 Bc3 40. Ng1 Bd7 41. Ne2 Bd2 42. Rd1 (42. Kg2 Ra2 $19) 42... Be3 43. Kg2 Bg4 44. Re1 Bd2 45. Rf1 Ra2 $19 (45... Bxe2 46. Rf2 Bxd3 47. cxd3 Be3 $15) 46. h3 (46. Ng1 Be3 $19) 46... Bxe2 47. Rf2 Bxf4 48. Rxe2 Be5 { White has three pawns for the B, but Black's B completely controls the b, c, d, and e-pawns.} 49. Rf2 Kg7 50. g4 Bd4 51. Re2 Kf6 52. e5+ (52. Kf3 Ke6 (52... Ke5 $4 53. c3 $1 $14) 53. Kg2 Ke5 54. Kf3 Ra1 $19) 52... Bxe5 53. Kf3 Ra1 54. Rf2 Re1 55. Kg2+ Bf4 {It might be silly to talk about "style" here, but AZ so often plays for domination by the minors, even when other options seem as good; as here} (55... Ke6 {also wins, but gives White's R some freedom,} 56. Rf8 Re2+ 57. Kf3 Rxc2 $19) 56. c3 Rc1 57. d4 Rxc3 58. dxc5 Rxc5 59. b4 Rc3 60. h4 Ke5 61. hxg5 hxg5 62. Re2+ Kf6 63. Kf2 Be5 64. Ra2 Rc4 65. Ra6+ Ke7 66. Ra5 Ke6 67. Ra6+ Bd6 0-1 [Event "AlphaZero vs. Stockfish"] [Site "?"] [Date "2017.12.04"] [Round "1.2"] [White "Stockfish 8"] [Black "AlphaZero"] [Result "0-1"] [ECO "C65"] [Annotator "John Upper"] [PlyCount "174"] [EventDate "2017.??.??"] [Source "CFC Newsfeed"] [SourceDate "2017.12.08"] [WhiteTeam "Norway"] [BlackTeam "England"] [WhiteTeamCountry "NOR"] [BlackTeamCountry "ENG"] {NO DRAW FOR YOU! AZ avoids repetitions twice before outplaying SF8} 1. e4 e5 2. Nf3 Nc6 3. Bb5 Nf6 4. d3 Bc5 5. Bxc6 dxc6 {Diagram [#]} 6. O-O (6. Nbd2 O-O 7. Qe2 Re8 8. Nc4 Nd7 9. Rg1 Bb4+ 10. Bd2 Bxd2+ 11. Qxd2 Qf6 {Giri,A (2762) -Jakovenko,D (2721) Palma de Mallorca 2017 (½-½, 35)}) 6... Nd7 7. c3 O-O 8. d4 Bd6 9. Bg5 Qe8 10. Re1 f6 11. Bh4 Qf7 $146 (11... Nb6 12. Nbd2 Bg4 13. Qc2 Kh8 14. Bg3 Nd7 15. Nh4 Qf7 {1/2-1/2 (31) Balogh,C (2638)-Parligras,M (2603) Achaea 2017}) 12. Nbd2 a5 13. Bg3 Re8 14. Qc2 Nf8 15. c4 c5 16. d5 b6 17. Nh4 g6 18. Nhf3 Bd7 19. Rad1 Re7 20. h3 Qg7 21. Qc3 Rae8 22. a3 h6 {Diagram [#]NO DRAW FOR YOU An extremely blocked position, the kind that has long been a problem for engines. SF8 doesn't seem to see anything to do and starts what looks like an attempt at repetition...} 23. Bh4 Rf7 24. Bg3 Rfe7 25. Bh4 Rf7 26. Bg3 a4 {... and Black declines!? 20 years ago, a paranoid GM as White might accuse the the programmers behind Black of outside interference, but today everybody knows there are less nefarious explanations: . 1. Black had more time to analyze this position, and may have come to a different evaluation about it because of the extra analysis; . 2. AZ might have created its equivalent of a "contempt value", such that it rates draws from an equal position as a sub-optimal outcome, and so will play into slightly inferior lines to avoid repetitions in positions it may otherwise rate as 0.00. } 27. Kh1 Rfe7 28. Bh4 Rf7 29. Bg3 Rfe7 30. Bh4 g5 {...and again, Black avoids the repetition.} 31. Bg3 Ng6 32. Nf1 Rf7 33. Ne3 Ne7 34. Qd3 {At 26 ply (less than what SF8 should have reached in this match) my version of SF8 rates this as a mistake, preferring both Qc2 (where it eyes a4) and Nd2 (possibly heading to c3).} h5 35. h4 Nc8 {Beginning a nice maneuver to put the N on d6 and activate the B on h6.} 36. Re2 g4 37. Nd2 Qh7 38. Kg1 Bf8 39. Nb1 Nd6 40. Nc3 Bh6 41. Rf1 Ra8 42. Kh2 Kf8 43. Kg1 Qg6 {Diagram [#] I don't know what Black would do if White sits with Kh2, but the move played in the game (f2-f4) leads to mass liquidations and an endgame where the Nd6 and the targets on c4 and e4 leave Black clearly better.} 44. f4 $6 (44. Rc2 Bxe3 45. fxe3 $17) (44. Kh2 $5) 44... gxf3 45. Rxf3 Bxe3+ 46. Rfxe3 Ke7 47. Be1 Qh7 48. Rg3 Rg7 49. Rxg7+ Qxg7 50. Re3 Rg8 51. Rg3 Qh8 52. Nb1 Rxg3 53. Bxg3 Qh6 54. Nd2 Bg4 {Diagram [#]} 55. Kh2 (55. Kf2 Bd1 {the B goes to b3 attacking c4; if White takes then Black's Q will get into White's position.}) 55... Kd7 (55... Bd1 $17 56. Nf1 Bb3 57. Ne3 Qg6 58. Nf5+ Nxf5 59. exf5 Qg4 $19) 56. b3 axb3 57. Nxb3 Qg6 58. Nd2 Bd1 59. Nf3 Ba4 60. Nd2 Ke7 61. Bf2 Qg4 62. Qf3 {Diagram [#]The rest is easier than it might look: White's c4 and e4 pawns are too weak.} Bd1 63. Qxg4 (63. Qd3 Qe2) 63... Bxg4 64. a4 Nb7 65. Nb1 Na5 66. Be3 Nxc4 67. Bc1 Bd7 68. Nc3 c6 69. Kg1 cxd5 70. exd5 Bf5 71. Kf2 Nd6 72. Be3 {Diagram [#]} Ne4+ 73. Nxe4 Bxe4 74. a5 bxa5 75. Bxc5+ Kd7 76. d6 Bf5 (76... Bc6 77. Bb6 a4 78. Bc5 Ke6 $19) 77. Ba3 Kc6 78. Ke1 Kd5 79. Kd2 Ke4 80. Bb2 Kf4 81. Bc1 Kg3 82. Ke2 a4 83. Kf1 Kxh4 84. Kf2 Kg4 85. Ba3 Bd7 86. Bc1 Kf5 87. Ke3 Ke6 0-1 [Event "AlphaZero vs. Stockfish"] [Site "?"] [Date "2017.12.04"] [Round "1.3"] [White "AlphaZero"] [Black "Stockfish 8"] [Result "1-0"] [ECO "E15"] [Annotator "John Upper"] [PlyCount "119"] [EventDate "2017.??.??"] [Source "CFC Newsfeed"] [SourceDate "2017.12.08"] [WhiteTeam "England"] [BlackTeam "Norway"] [WhiteTeamCountry "ENG"] [BlackTeamCountry "NOR"] {QID ZUGZWANG Six of the 10 published games in the AZ SF8 match were Queen's Indians, all with AZ as White. This is the one that will go into anthologies.} 1. Nf3 Nf6 2. c4 b6 3. d4 e6 4. g3 Ba6 5. Qc2 c5 6. d5 $1 { Diagram [#] This pawn sac is the modern mainline: it frees f5 for a piece and tactics on the long diagonal make it awkward for Black to finish queenside development.} exd5 7. cxd5 Bb7 8. Bg2 Nxd5 9. O-O Nc6 10. Rd1 Be7 11. Qf5 (11. Qa4 Nf6 12. Nh4 O-O 13. Nc3 g6 14. Bg5 Rb8 15. Bf4 Rc8 16. Bh6 Re8 17. Qf4 Na5 18. Nf5 Bxg2 19. Nxe7+ Qxe7 20. Kxg2 Qe5 21. e3 d5 22. Bg5 Ne4 23. Rxd5 Qxf4 24. gxf4 Nxc3 25. bxc3 Kg7 26. Rd7 $16 {1-0 (83) Aronian,L (2786)-Nakamura,H (2790) Moscow 2016}) 11... Nf6 12. e4 g6 13. Qf4 O-O 14. e5 Nh5 {Diagram [#] It may be worth reminding readers that neither engine was using an opening book, but were finding these moves as the result of calculation.} 15. Qg4 (15. Qc4 Re8 16. Nc3 Ng7 17. Nd5 Ne6 18. Be3 Nb4 19. Nf4 Qc8 20. a3 Bxf3 21. Bxf3 Nc6 22. Qc3 Ncd4 23. Bxd4 cxd4 24. Qd3 Rb8 25. Rac1 Qd8 26. Nxe6 dxe6 27. Bc6 Rf8 28. Qxd4 Qxd4 29. Rxd4 Rfc8 $11 {1/2-1/2 (41) Shirov,A (2715)-Leko,P (2707) Skopje 2015}) 15... Re8 $146 (15... d5 16. exd6 Bf6 17. Nc3 Nd4 18. Nxd4 Bxg2 19. Nf5 Bc6 20. d7 Qc7 21. Nd5 Bxd5 22. Rxd5 Rfd8 23. Be3 Bxb2 24. Nh6+ Kf8 25. Rad1 Bg7 26. Rxh5 gxh5 27. Qf5 {1-0 (27) Carlsen,M (2690)-Ivanchuk,V (2750) Monte Carlo (rapid) 2007}) 16. Nc3 Qb8 17. Nd5 Bf8 18. Bf4 Qc8 19. h3 Ne7 20. Ne3 Bc6 21. Rd6 Ng7 22. Rf6 {Diagram [#] Materially, Black is up a pawn, but those dark squares are very weak, and it is hard to imagine how it will get the Rs into play.} Qb7 23. Bh6 Nd5 24. Nxd5 Bxd5 25. Rd1 Ne6 26. Bxf8 Rxf8 27. Qh4 {threatening Ng5, with mate on h7 and a skewer on Bxd5.} Bc6 28. Qh6 Rae8 29. Rd6 {I can't remember seeing a game with two R outposts like this.} Bxf3 30. Bxf3 {Diagram [#]White has a terrific bind. Black's only mobile piece is the Q, otherwise the tactics all go White's way (see notes). What impresses me most about the next phase of the game is that White moves all of its pieces from their dominating positions, and ends up leaving Black even MORE tied up than before!} Qa6 {Going for counterplay with its only active piece.} (30... Qc7 {Tries to hold on.} 31. Bd5 Ng7 (31... Nd4 $2 32. Rxg6+ hxg6 33. Rxg6#) 32. Qd2 {is similar to the game, but with Black a bit less active;} (32. f4 $2 Nf5 $2 (32... Re6 {only move.}) 33. Rxf7 Rxf7 (33... Nxh6 34. Rfxd7+ $18) 34. Bxf7+ Kxf7 35. Qxh7+ $18) 32... Nf5 $4 {fails to} 33. Rxf5 $18) 31. h4 Qa5 32. Rd1 c4 33. Rd5 Qe1+ 34. Kg2 c3 35. bxc3 Qxc3 36. h5 Re7 37. Bd1 {Diagram [#]} Qe1 (37... Ng7 38. hxg6 hxg6 39. Bb3 Rxe5 (39... Nf5 40. Rxf5 gxf5 41. Qg5+ $18) (39... Re6 40. Rf3 Qc6 41. Qd2 $18) 40. Rxd7 Re7 41. Rf4 $1 $18 Rxd7 (41... Qc6+ 42. Kh2 $1) 42. Rh4 Nh5 43. Qxg6+ Qg7 44. Qxh5 Rd6 45. Rg4 Rg6 46. Rxg6 $18) 38. Bb3 Rd8 39. Rf3 Qe4 40. Qd2 Qg4 41. Bd1 Qe4 (41... Qxh5 42. Rxf7 $18 {the Q can't safely defend itself and the Re7.}) 42. h6 Nc7 43. Rd6 Ne6 (43... Qxe5 $4 44. Re3 Qg5 45. f4 $18) 44. Bb3 Qxe5 45. Rd5 Qh8 46. Qb4 Nc5 {Diagram [#]} 47. Rxc5 $3 {SF8 and Komodo10 do not rank this move among their top 4; Houdini4 does, but all three rate the resulting position as equal.} bxc5 48. Qh4 $1 {I wonder if it saw that the c5 pawn would leave the B with a nice outpost on c4!?} (48. Qxc5 d6 49. Bxf7+ Rxf7 50. Qd5 Rdf8 51. Rxf7 Rxf7 52. Qa8+ Rf8 53. Qd5+ $11 { with a perpetual.}) 48... Rde8 {only move.} (48... Ree8 49. Bxf7+ Kf8 50. Bc4+ $18) 49. Rf6 $1 {Diagram [#]Another picture of complete domination.} Rf8 50. Qf4 a5 51. g4 d5 52. Bxd5 Rd7 53. Bc4 a4 {Computers always choose to lose ugly.} (53... Rd4 $1 54. Rxf7 Rxf4 55. Rg7# {if SF8 had allowed this, would we suspect it had a sense of beauty?}) 54. g5 a3 {If White started marching his K around the board right now, then we would know that real AI had arrived, because it would show AZ had developed a sense of humour. Or sadism.} 55. Qf3 Rc7 56. Qxa3 Qxf6 57. gxf6 Rfc8 58. Qd3 Rf8 59. Qd6 Rfc8 60. a4 1-0 [Event "AlphaZero vs. Stockfish"] [Site "?"] [Date "2017.12.04"] [Round "1.4"] [White "AlphaZero"] [Black "Stockfish 8"] [Result "1-0"] [ECO "C11"] [Annotator "John Upper"] [PlyCount "189"] [EventDate "2017.??.??"] [Source "CFC Newsfeed"] [SourceDate "2017.12.08"] [WhiteTeam "England"] [BlackTeam "Norway"] [WhiteTeamCountry "ENG"] [BlackTeamCountry "NOR"] {BAD OPENING? This is the only game of the 10 published where SF8 clearly plays worse than it would have with an ordinary opening book.} 1. d4 e6 2. Nc3 Nf6 3. e4 d5 4. e5 Nfd7 5. f4 c5 6. Nf3 Nc6 ({Game 9 continued:} 6... cxd4 7. Nb5 Bb4+ 8. Bd2 Bc5 9. b4 Be7 10. Nbxd4 Nc6 11. c3 a5 12. b5 $14 {1-0 (52) AlphaZero-Stockfish 8 AlphaZero vs. Stockfish 2017}) 7. Be3 Be7 8. Qd2 a6 9. Bd3 {Diagram [#]} c4 $2 {Computers generally like these space-gaining tempo-gaining moves, but humans who play the French believe that (unless White has caslted long!) it is wrong to give up the pressure on d4 and potential c-file play since then White's space advantage on the kingside is more important than Black's on the queenside. . It would be interesting to know what AZ played as Black if they had this position, but such a game (if played) was not included in the 10 games selected by the google team. FWIW, during the 2016 TCEC SuperFinal between SF8 and Houdini 5, the Opening Book did not allow the engines to play this error, continuing: 7...cxd4 8.Nxd4 Bc5 9.Qd2 0-0 10.0-0-0 a6 to reach a tabiya of the Classical/Steinitz French.} (9... b5 10. O-O Qb6 11. dxc5 Bxc5 12. Bxc5 Nxc5 13. Qf2 Nd7 14. a3 Ke7 15. Qxb6 Nxb6 16. Ne2 Bd7 17. b3 f6 18. Ned4 Nxd4 19. Nxd4 fxe5 20. fxe5 $14 {1-0 (50) Caruana,F (2787)-Nakamura,H (2793) Saint Louis 2015}) 10. Be2 b5 11. a3 Rb8 12. O-O O-O 13. f5 a5 14. fxe6 fxe6 {Diagram [#]} 15. Bd1 $1 {The star move of the game: undeveloping to reroute the Nc3 to the kingside.} b4 {Giving up a pawn for play that doesn't compensate.} 16. axb4 axb4 17. Ne2 c3 18. bxc3 Nb6 19. Qe1 Nc4 20. Bc1 bxc3 21. Qxc3 Qb6 22. Kh1 Nb2 23. Nf4 Nxd1 24. Rxd1 Bd7 25. h4 Ra8 26. Bd2 Rfb8 27. h5 Rxa1 28. Rxa1 Qb2 29. Qxb2 Rxb2 30. c3 {Diagram [#]} Rb3 31. Ra8+ Rb8 32. Ra2 Rb3 33. g4 Ra3 34. Rb2 Kf7 35. Kg2 Bc8 36. Rb6 Ra6 37. Rb1 Ke8 38. Kg3 h6 39. Ng6 Ra3 40. Rb6 Bd7 41. g5 hxg5 42. Kg4 Bd8 43. Rb2 Bc8 44. Nxg5 Ra1 45. Nf3 Ra3 46. Be1 Ba5 {Diagram [#]} 47. Rf2 $1 {offers to sac two pawns to make the h-pawn an unstoppable passer.} Ra1 (47... Bxc3 48. Bxc3 Rxc3 49. Nfh4 Nxd4 50. Rf8+ Kd7 51. Rf7+ Kc6 52. Rxg7 $18) 48. Bd2 Bd8 49. Rh2 Ne7 50. Bg5 Nf5 51. Bxd8 Kxd8 52. Rb2 Rc1 53. Ngh4 {Diagram [#]} Nxh4 (53... Rxc3 54. Nxf5 exf5+ 55. Kf4 Rc1 56. Rg2 Rc7 57. Rg6 $18 {and White's pieces dominate Blacks', again.}) 54. Nxh4 Bd7 55. Rb8+ Bc8 56. Ng2 Rxc3 57. Nf4 Rc1 58. Ra8 Kd7 59. Kf3 Rc3+ 60. Kf2 Ke7 61. Kg2 Kf7 62. Ng6 Ke8 63. Ra1 Rc7 (63... Rd3 64. Rf1 Rxd4 65. Rf8+ Kd7 66. Rf7+ $18 {is like the note at move 47, minus one pair of Ns.}) 64. Kh3 Rf7 65. Kg4 Kd8 66. Nf4 Bd7 67. Ra7 Kc8 68. Kg3 Re7 69. Nd3 Kb8 70. Ra6 Bc8 71. Rb6+ Kc7 72. Rd6 Kb8 73. Nc5 g6 74. h6 Rh7 75. Nxe6 Rxh6 {Diagram [#]} 76. Nf4 Rh1 77. Nxd5 Rh3+ 78. Kf4 Rh4+ 79. Ke3 Rh3+ 80. Kd2 Bf5 81. Ne7 Rh2+ 82. Ke3 Bh3 83. Nxg6 Rh1 84. Nf4 Bg4 85. Rf6 Kc7 86. Nd3 Bd7 87. d5 Bb5 88. Nf4 Ba4 89. Kd4 Be8 90. Rf8 Rd1+ 91. Kc5 Rc1+ 92. Kb4 Rb1+ 93. Kc3 Bb5 94. Kd4 Ba6 95. Rf7+ 1-0 [Event "AlphaZero vs. Stockfish"] [Site "?"] [Date "2017.12.04"] [Round "1.5"] [White "AlphaZero"] [Black "Stockfish 8"] [Result "1-0"] [ECO "E17"] [Annotator "John Upper"] [PlyCount "233"] [EventDate "2017.??.??"] [Source "CFC Newsfeed"] [SourceDate "2017.12.08"] [WhiteTeam "England"] [BlackTeam "Norway"] [WhiteTeamCountry "ENG"] [BlackTeamCountry "NOR"] {Bg5!! vs QUEENSIDE A spectacular move lets AZ finish development and establish a game-winning bind.} 1. d4 Nf6 2. c4 e6 3. Nf3 b6 4. g3 Bb7 5. Bg2 Be7 6. O-O O-O 7. d5 exd5 8. Nh4 c6 9. cxd5 Nxd5 10. Nf5 Nc7 11. e4 Bf6 (11... d5 12. Re1 Bf6 13. Nc3 Bc8 14. Qf3 Re8 15. Bd2 Nd7 16. exd5 Ne5 17. Qh5 g6 18. Nh6+ Kg7 19. Qd1 cxd5 {unclear.} {(0-1, 75) Moiseenko,A (2657)-Anton Guijarro,D (2650) Doha, 2016.}) 12. Nd6 Ba6 13. Re1 Ne8 14. e5 Nxd6 15. exf6 (15. exd6 Bc4 16. Nd2 Be6 17. Ne4 Na6 18. Qf3 Nb4 19. Nxf6+ {1/2-1/2 (19) Castaneda,G (2518) -Admiraal,M (2301) Wijk aan Zee 2012}) 15... Qxf6 16. Nc3 Nb7 $146 (16... Bc4 17. Bf4 Nf5 18. Ne4 Qg6 19. h4 h6 20. h5 $16 {1/2-1/2 (40) Podzielny,K (2408) -Emunds,H (2179) Germany 2011}) 17. Ne4 Qg6 18. h4 h6 19. h5 Qh7 20. Qg4 Kh8 { Diagram [#]} 21. Bg5 $3 {Developing a piece and threatening Nf6! The tactics are simple enough that even a laptop-powered SF8 can see that the Bg5 can't be taken, but it rates the position after ...f5 as 0.00.} f5 $1 (21... hxg5 $4 22. Nxg5 Qh6 23. Nxf7+ Rxf7 24. Re8+ Kh7 25. Be4+ {is mating.}) (21... d5 {lets White demonstrate one point of Bg5} 22. Nf6 $3 Qc2 (22... Qd3 {covers more squares, but makes no difference since it can be dislodged with Rad1.}) 23. Bxh6 gxh6 24. Qf4 Kg7 25. Qe5 $18 (25. Ne8+ $18)) 22. Qf4 $1 Nc5 (22... hxg5 23. Nxg5 Qg8 (23... Qxh5 $2 24. g4 $1 $18 Qg6 (24... Qxg4 $4 25. Qh2+ $18) 25. gxf5 Rxf5 26. Qxf5 $18) 24. Re7 {threatening h5-h6.} Rf6 25. Qe5 d6 26. Qc3 Nc5 27. h6 Rxh6 28. Nf7+ Kh7 29. Nxh6 Kxh6 30. Qd2+ $18) 23. Be7 (23. Nxc5 hxg5 24. Qd6 $14) 23... Nd3 24. Qd6 Nxe1 25. Rxe1 fxe4 {Diagram [#]} 26. Bxe4 (26. Bxf8 Bd3 27. Bxe4 Bxe4 28. Bxg7+ Kxg7 29. Qe5+ $1 Kg8 (29... Kf8 30. Rxe4 Na6 31. Qf6+ $18) 30. Qe8+ Kg7 31. Rxe4 $18) 26... Rf5 {only move.} 27. Bh4 Bc4 28. g4 Rd5 29. Bxd5 Bxd5 30. Re8+ Bg8 $18 {Diagram [#]Black is up a piece and a pawn, but is so completely dominated by White that it can't develop the queenside without losing material.} 31. Bg3 c5 32. Qd5 d6 33. Qxa8 Nd7 34. Qe4 Nf6 (34... Qxe4 35. Rxe4 Nf6 36. Ra4 a5 37. Bxd6 $18) 35. Qxh7+ Kxh7 36. Re7 Nxg4 37. Rxa7 Nf6 38. Bxd6 Be6 39. Be5 Nd7 {Diagram [#]} 40. Bc3 {The Stockfish build I use rates White's next few moves as mistakes, but not nearly serious enough to jeopardize White's win.} (40. Bc7 $18) 40... g6 41. Bd2 gxh5 42. a3 Kg6 43. Bf4 Kf5 44. Bc7 h4 45. Ra8 h5 46. Rh8 Kg6 47. Rd8 Kf7 48. f3 Bf5 49. Bh2 h3 50. Rh8 Kg6 51. Re8 Kf7 52. Re1 Be6 53. Bc7 b5 54. Kh2 Kf6 55. Re3 Ke7 56. Re4 Kf7 57. Bd6 Kf6 58. Kg3 Kf7 59. Kf2 Bf5 60. Re1 Kg6 61. Kg1 c4 62. Kh2 {Diagram [#]} h4 (62... Kf7) 63. Be7 Nb6 64. Bxh4 Na4 65. Re2 Nc5 66. Re5 Nb3 67. Rd5 Be6 68. Rd6 Kf5 69. Be1 Ke5 70. Rb6 Bd7 71. Kg3 Nc1 72. Rh6 Kd5 73. Bc3 {Diagram [#]} Bf5 (73... Ne2+ 74. Kf2 Nxc3 75. bxc3 {and White advances the f-pawn.}) 74. Rh5 Ke6 75. Kf2 Nd3+ 76. Kg1 Nf4 77. Rh6+ Ke7 78. Kh2 Nd5 79. Kg3 Be6 80. Rh5 Ke8 81. Re5 Kf7 82. Bd2 Ne7 83. Bb4 Nd5 84. Bc3 Ke7 85. Bd2 Kf6 86. f4 Ne7 87. Rxb5 Nf5+ 88. Kh2 Ke7 89. Ra5 Nh4 90. Bb4+ Kf7 91. Rh5 Nf3+ 92. Kg3 Kg6 93. Rh8 Nd4 94. Bc3 Nf5+ 95. Kxh3 Bd7 96. Kh2 Kf7 97. Rb8 Ke6 98. Kg1 Bc6 99. Rb6 Kd5 100. Kf2 Bd7 101. Ke1 Ke4 102. Bd2 Kd5 103. Rf6 Nd6 104. Rh6 Nf5 105. Rh8 Ke4 106. Rh7 Bc8 107. Rc7 Ba6 108. Rc6 Bb5 {Diagram [#]} 109. Rc5 Bd7 (109... Ba6 110. Re5+ {wins the N.}) 110. Rxc4+ Kd5 111. Rc7 Kd6 112. Rc3 Ke6 113. Rc5 Nd4 114. Be3 Nf5 115. Bf2 Nd6 116. Rc3 Ne4 117. Rd3 1-0 [Event "AlphaZero vs. Stockfish"] [Site "?"] [Date "2017.12.04"] [Round "1.6"] [White "AlphaZero"] [Black "Stockfish 8"] [Result "1-0"] [ECO "E17"] [Annotator "John Upper"] [PlyCount "139"] [EventDate "2017.??.??"] [Source "CFC Newsfeed"] [SourceDate "2017.12.08"] [WhiteTeam "England"] [BlackTeam "Norway"] [WhiteTeamCountry "ENG"] [BlackTeamCountry "NOR"] {BISHOP PAIR, worth 2 pawns!?} 1. d4 Nf6 2. Nf3 e6 3. c4 b6 4. g3 Be7 5. Bg2 Bb7 6. O-O O-O 7. d5 exd5 8. Nh4 c6 9. cxd5 Nxd5 10. Nf5 Nc7 11. e4 Bf6 12. Nd6 Ba6 13. Re1 Ne8 14. e5 Nxd6 15. exf6 Qxf6 16. Nc3 Bc4 17. h4 h6 18. b3 Qxc3 19. Bf4 Nb7 20. bxc4 Qf6 ({I wonder if SF8 avoided} 20... Qxc4 {because it allows White to draw with} 21. Rc1 Qxa2 22. Ra1 Qc4 23. Rc1) 21. Be4 Na6 22. Be5 Qe6 23. Bd3 {Diagram [#]} f6 $1 {Black doesn't want to loosen the kingside light sqaures, but there's no other way to save the Q from Bxg7. For instance,} ( 23... Qh3 24. Bxg7 $3 (24. Qh5 {also works.}) 24... Kxg7 25. Qh5 {and there's no good way to save the Q from getting trapped by Bf5 since} f5 26. Re7+ Kf6 27. Rae1 {is a mating attack.}) 24. Bd4 Qf7 25. Qg4 {Diagram [#]White is down two pawns but has the B pair, light square weaknesses around Black's K, and Black's Ns can't both get into the game via c5.} Rfd8 26. Re3 Nac5 27. Bg6 Qf8 28. Rd1 Rab8 ({I don't see what's wrong with this:} 28... Ne6 $5 29. Bc3 Nbc5) 29. Kg2 $5 (29. Rf3) 29... Ne6 30. Bc3 Nbc5 31. Rde1 Na4 32. Bd2 Kh8 33. f4 Qd6 34. Bc1 {Diagram [#]} Nd4 (34... Nf8 $2 35. f5 $18 {followed by Re7 and Bxh6.}) (34... Nac5 35. f5 Qd4 36. Qe2 {wins a piece by threatening to trap the Q with Rd1.}) 35. Re7 f5 36. Bxf5 Nxf5 37. Qxf5 Rf8 38. Rxd7 Rxf5 39. Rxd6 Rf7 $16 { Diagram [#] Black has avoided mate, and material is equal, but White has the more active Rs the better minor piece, and chances attack by advancing its kingside majority.} 40. g4 Kg8 41. g5 hxg5 42. hxg5 Nc5 43. Kf3 Nb7 44. Rdd1 Na5 45. Re4 c5 46. Bb2 Nc6 47. g6 Rc7 48. Kg4 Nd4 {Diagram [#]AZ is winning, but plays some very strange "nothing" moves along the way.} 49. Rd2 $5 { Can anyone see a reason not to take on d4 right away?} (49. Bxd4 cxd4 (49... Rd8 $2 50. Bc3 Rxd1 $2 51. Re8#) 50. Rdxd4 {transposes to the game, but with the R on b8 instead of f8; does that make a difference? I guess it did to AZ.}) 49... Rf8 50. Bxd4 cxd4 51. Rdxd4 Rfc8 52. Kg5 Rf8 (52... Rc5+ {doesn't help.} 53. Re5 Rxe5+ 54. fxe5 Kf8 55. Rd7 $18 Rxc4 56. Rf7+ $18) 53. Rd2 Rc6 54. Rd5 Rc7 55. f5 Rb7 56. a3 $5 {playing for zugzwang after ...a6 then a4?} Rc7 57. a4 a6 58. Red4 Rcc8 59. Re5 Rc7 60. a5 Rc5 61. Rxc5 bxc5 62. Rd6 Ra8 63. Re6 Kf8 64. Rc6 Ke7 65. Kf4 Kd7 66. Rxc5 Rh8 67. Rd5+ Ke7 68. Re5+ Kd7 69. Re6 Rh4+ 70. Kg5 1-0 [Event "AlphaZero vs. Stockfish"] [Site "?"] [Date "2017.12.04"] [Round "1.7"] [White "AlphaZero"] [Black "Stockfish 8"] [Result "1-0"] [ECO "E16"] [Annotator "John Upper"] [PlyCount "199"] [EventDate "2017.??.??"] [Source "CFC Newsfeed"] [SourceDate "2017.12.08"] [WhiteTeam "England"] [BlackTeam "Norway"] [WhiteTeamCountry "ENG"] [BlackTeamCountry "NOR"] {KARPOV 2.0 A smooth, almost tactics-free win, which reminds me of Kramnik's description of Karpov: "When Karpov had an advantage he could maintain the status quo and thereby mysteriously increase his advantage!" Black allows itself to get stuck with an inflexible queenside and a poor LSB, and when White starts to advance on the kingside, every pawn move Black makes gets expolited by White's pieces.} 1. d4 Nf6 2. c4 e6 3. Nf3 b6 4. g3 Bb7 5. Bg2 Bb4+ 6. Bd2 Bxd2+ 7. Qxd2 d5 8. O-O O-O 9. cxd5 exd5 10. Nc3 Nbd7 {Diagram [#]} 11. b4 (11. Rac1 Re8 12. Rfd1 c6 (12... a6 {this game shows the kind of bind Black should avoid:} 13. Qc2 Nf8 14. b4 Qe7 15. Qb3 Ne6 16. e3 Rab8 17. Ne5 Red8 18. Na4 Ng5 19. Rc2 Nge4 20. Rdc1 Ne8 $16 {1/2-1/2 (55) Beliavsky,A (2679)-Kasimdzhanov,R (2640) Pune 2004}) 13. Qc2 Qe7 14. Nd2 Nf8 15. e4 dxe4 16. Ndxe4 Ne6 17. Qa4 Nxe4 18. Bxe4 b5 {0-1 (38) Markus,R (2662)-Cori Tello,J (2609) Baku 2016}) 11... c6 12. Qb2 $146 (12. a4 a5 13. b5 c5 14. Rfc1 Re8 15. dxc5 bxc5 16. e3 {½-½ Stohl,I (2556)-Parker,J (2570) Kallithea 2002}) 12... a5 $6 13. b5 c5 14. Rac1 Qe7 15. Na4 Rab8 16. Rfd1 c4 {Diagram [#]Black has a bad piece on b7. The rest of the game looks like Karpov vs NoName GM c.1977.} 17. Ne5 Qe6 18. f4 Rfd8 19. Qd2 Nf8 20. Nc3 Ng6 21. Rf1 Qd6 22. a4 Rbc8 23. e3 Ne7 24. g4 Ne8 25. f5 f6 26. Nf3 Qd7 27. Qf2 Nd6 28. Nd2 Rf8 29. Qg3 Rcd8 30. Rf4 Nf7 31. Rf2 Rfe8 32. h3 Qd6 33. Nf1 Qa3 34. Rcc2 h5 35. Qc7 Qd6 {Diagram [#]} 36. Qxd6 (36. Qxb7 $5 {looks playable, but not in the Karpov-style like the rest of the game.} Rb8 37. Qa6 Ra8 38. Nxd5 $3 Rxa6 (38... Nxd5 39. Qb7 $18 ) 39. Nxe7+ Qxe7 40. bxa6 Qa7 41. Rxc4 Qxa6 42. Rfc2 {unclear.}) 36... Rxd6 37. Ng3 h4 {the position is so closed that even though White has to go through contortions to win the h-pawn, it can.} (37... hxg4 $5 38. hxg4 Ng5) 38. Nh5 Ng5 39. Rf1 Kh7 40. Nf4 Rdd8 41. Kh2 Rd7 42. Bh1 $1 {Diagram [#] No sense of shame.} Rd6 (42... Rh8 {saves the h4 pawn...} 43. Ng2 Kg8 44. Nf4 $14 {but maybe the Rh8 will be offside?}) 43. Ng2 g6 $1 44. Nxh4 gxf5 45. gxf5 Rh8 46. Nf3 Kg7 (46... Nxf5 $4 47. Nxg5+ fxg5 48. Rxf5 $18) 47. Nxg5 fxg5 48. Rg2 Kf6 { Diagram [#] The next phase feels even more like Karpov: lots of very "small" (one square) moves which infintesimially improve White's coordination and keep the extra pawn.} 49. Rg3 Re8 50. Bf3 Rdd8 51. Be2 Rf8 52. Bg4 Nc8 53. Bf3 Rfe8 {Of course, taking the f5 pawn with the K loses the Bb7.} 54. h4 Rh8 (54... Rxe3 $4 {drops the R since after} 55. hxg5+ {the passers will win if Black doesn't take one of them, but} Kxf5 56. Bh5+ $18) 55. h5 Rhe8 56. Bg2 Ne7 57. h6 Rh8 58. Rh3 Rh7 59. Kg1 Ba8 60. Nd1 g4 61. Rh5 g3 62. Nc3 Ng8 63. Ne2 { Diagram [#]The rest is too easy.} Rxh6 64. Nxg3 Rxh5 65. Nxh5+ Kf7 66. Kf2 Nf6 67. Nxf6 Kxf6 68. Rh1 c3 69. Rc1 Rh8 70. Rxc3 Kxf5 71. Rc7 Kf6 72. Bf3 Rg8 73. Rh7 Rg6 74. Bd1 Rg8 75. Rh6+ Ke7 76. Rxb6 Kd7 77. Rf6 Ke7 78. Rh6 Rg7 79. Rh8 Bb7 80. Rh5 Kd6 81. Rh3 Rf7+ 82. Ke1 Bc8 83. Rh6+ Kc7 84. Rc6+ Kb8 85. Rd6 Bb7 86. b6 Ba6 87. Rxd5 Rf6 88. Rxa5 Rxb6 89. Kd2 Bb7 90. Rb5 Rf6 91. Bb3 Kc7 92. Re5 Ba6 93. Kc3 Rf1 94. Bc2 Rh1 95. a5 Kd6 96. e4 Bf1 97. Rf5 Bg2 98. Rf4 Rc1 99. Kb2 Rh1 100. a6 1-0 [Event "AlphaZero vs. Stockfish"] [Site "?"] [Date "2017.12.04"] [Round "1.8"] [White "AlphaZero"] [Black "Stockfish 8"] [Result "1-0"] [ECO "E16"] [Annotator "John Upper"] [PlyCount "135"] [EventDate "2017.??.??"] [Source "CFC Newsfeed"] [SourceDate "2017.12.08"] [WhiteTeam "England"] [BlackTeam "Norway"] [WhiteTeamCountry "ENG"] [BlackTeamCountry "NOR"] {DOES STOCKFISH HATE ITS LSB? One standard theme in the QGD-family openings is the problems Black has in developing the LSB. SF8 lost horribly in several games in this match for just this reason.} 1. d4 Nf6 2. c4 e6 3. Nf3 b6 4. g3 Bb7 5. Bg2 Bb4+ 6. Bd2 Be7 7. Nc3 c6 8. e4 d5 {Diagram [#]} 9. e5 (9. exd5 cxd5 10. Ne5 O-O 11. O-O Nc6 12. cxd5 (12. Bf4 Na5 13. Rc1 dxc4 14. Bxb7 Nxb7 15. Nxc4 {(1-0, 41) Radjabov,T (2724)-Eljanov,P (2739) Geneva, 2017.}) 12... Nxe5 13. d6 Nc6 $1 14. dxe7 Qxe7 15. Bg5 h6 16. d5 {unclear.} {Anand,V (2792)-Carlsen,M (2863) WCh g5, 2014 ½-½ (39).}) 9... Ne4 10. O-O (10. cxd5 cxd5 11. O-O O-O { ½-½ Nikolaidis,I (2524)-Halkias,S (2519) Athens 2001} 12. Re1 Nxd2 13. Qxd2 Ba6 14. a3 Nc6 15. b4 Bc4 16. Qe3 b5 17. Nd2 Qb6 18. Nxc4 bxc4 19. Rad1 { ½-½ (43) Karpov,A (2780)-Tiviakov,S (2625) Linares 1995}) 10... Ba6 11. b3 Nxc3 12. Bxc3 dxc4 13. b4 b5 {Once again, AZ is down material but its opponent's pieces are sadly passive.} 14. Nd2 O-O 15. Ne4 Bb7 16. Qg4 Nd7 17. Nc5 Nxc5 18. dxc5 {Diagram [#] Once again, SF8 ends up with a pseudo-B shut in on b7. Compare with the Karpov-Tiviakov game in the notes to move 10.} a5 19. a3 axb4 20. axb4 Rxa1 21. Rxa1 Qd3 22. Rc1 Ra8 23. h4 Qd8 24. Be4 Qc8 25. Kg2 $1 Qc7 26. Qh5 g6 (26... h6 $5 {also gives White a hook, though maybe not one that allows the R to get into the game so easily.}) 27. Qg4 Bf8 (27... h5 { changes the focus but doesn't stop the attack,} 28. Qf4 Bf8 29. g4) 28. h5 Rd8 29. Qh4 {Diagram [#]} Qe7 (29... Be7 30. Qh3 Bg5 $2 31. hxg6 $1 fxg6 32. f4 Be7 33. Rh1 Bf8 34. Bxg6 $18) 30. Qf6 {Diagram [#] Calling Black's bluff: without the Q to plug some holes there's no way for Black to survive with such passive Bs.} Qe8 (30... Qxf6 31. exf6 Bh6 $2 32. Rh1 $1 g5 (32... Bd2 33. Rd1) 33. g4 e5 34. Ra1 $1 {Black can't allow the R to enter, but can't survive the pure B ending with weaknesses on h7, e5, c6.}) 31. Rh1 Rd7 32. hxg6 fxg6 33. Qh4 Qe7 34. Qg4 Rd8 35. Bb2 Qf7 (35... Rd2 36. Bxg6 hxg6 37. Qxg6+ Bg7 38. Bc1 $18) 36. Bc1 c3 37. Be3 Be7 38. Qe2 Bf8 39. Qc2 Bg7 40. Qxc3 {Diagram [#]} Qd7 (40... Qc7 41. Bg5 $1) 41. Rc1 $1 {A weird-looking move, but reasonable: White wants to play Bg5 to control the d-file, but doesn't want to allow Black time for ... Qd4 (then ...Rd7) which would keep control of the d-file, so AZ defends the Qc4 first.} Qc7 42. Bg5 Rf8 43. f4 h6 44. Bf6 $1 {A temporary sac which ensures the Qs get traded.} Bxf6 45. exf6 Qf7 46. Ra1 Qxf6 47. Qxf6 Rxf6 48. Ra7 Rf7 49. Bxg6 $18 {Material is only nominally equal, since the thing on b7 is useless and White has multiple ways to win.} Rd7 50. Kf2 Kf8 51. g4 Bc8 52. Ra8 Rc7 53. Ke3 h5 54. gxh5 Kg7 55. Ra2 Re7 56. Be4 e5 57. Bxc6 exf4+ 58. Kxf4 Rf7+ 59. Ke5 Rf5+ 60. Kd6 Rxh5 61. Rg2+ Kf6 62. Kc7 Bf5 63. Kb6 Rh4 64. Ka5 $5 {Another strange-looking but possibly interesting choice: in what way might it be better than Kxb5?} Bg4 65. Bxb5 Ke7 66. Rg3 Bc8 67. Re3+ Kf7 68. Be2 1-0 [Event "AlphaZero vs. Stockfish"] [Site "?"] [Date "2017.12.04"] [Round "1.9"] [White "AlphaZero"] [Black "Stockfish 8"] [Result "1-0"] [ECO "C11"] [Annotator "John Upper"] [PlyCount "103"] [EventDate "2017.??.??"] [Source "CFC Newsfeed"] [SourceDate "2017.12.08"] [WhiteTeam "England"] [BlackTeam "Norway"] [WhiteTeamCountry "ENG"] [BlackTeamCountry "NOR"] {DOES STOCKFISH REALLY HATE ITS LSB? Another game where SF8 gets a garbage LSB. Maybe they could try a match where both engines had no LSBs!?} 1. d4 e6 2. e4 d5 3. Nc3 Nf6 4. e5 Nfd7 5. f4 c5 6. Nf3 cxd4 {Diagram [#]} 7. Nb5 Bb4+ 8. Bd2 Bc5 9. b4 Be7 10. Nbxd4 Nc6 11. c3 a5 12. b5 Nxd4 13. cxd4 Nb6 14. a4 Nc4 15. Bd3 Nxd2 {Diagram [#]} 16. Kxd2 $1 {the obvious move, but still cute.} Bd7 17. Ke3 b6 18. g4 h5 {Diagram [#]} 19. Qg1 (19. g5 g6 20. Qb1 Bb4 21. Bxg6 (21. Nh4 $5) 21... fxg6 22. Qxg6+ Kf8 23. Nh4 {unclear.}) 19... hxg4 20. Qxg4 Bf8 21. h4 Qe7 22. Rhc1 g6 23. Rc2 Kd8 24. Rac1 Qe8 (24... Rc8 25. Qg2 $16 {and the White Q heads to the c-file.}) 25. Rc7 Rc8 26. Rxc8+ Bxc8 27. Rc6 Bb7 28. Rc2 (28. Rxb6 $4 Kc7 29. Rd6 Bxd6 30. exd6+ Kxd6 $15) 28... Kd7 29. Ng5 Be7 {Diagram [#] Once again, SF8 has a terrible LSB. Since it can (at best) hope to sacrifice itself on b5, White's next move maybe shouldn't be thought of as a piece sacrifice, since White's up a piece already...} 30. Bxg6 $1 Bxg5 {only move.} (30... fxg6 $4 31. Qxe6+ Kd8 32. Qxb6+ Kd7 33. Qc7#) (30... Rg8 $2 31. Nxe6 $1 Rxg6 (31... fxe6 32. Bxe8+ {with check, saving the Qg4.}) 32. Rc7#) 31. Qxg5 fxg6 32. f5 $3 Rg8 (32... exf5 $2 33. Qf6 {creeping in along the 6th} Qf8 34. Qxb6 $18) (32... gxf5 $2 33. Qg7+ $18 {wins everything.}) 33. Qh6 Qf7 {Diagram [#]} 34. f6 { Materially, Black is up a B for a pawn, but that's only if you count that immobile thing on b7 as "a piece".} (34. fxe6+ Kxe6 (34... Qxe6 $4 35. Qh7+ $18 ) 35. Rf2 Qe7 36. Rf6+ {unclear.}) 34... Kd8 (34... Rc8 35. Rxc8 Bxc8 36. Kf4 $1 Ke8 37. Kg5 $18) 35. Kd2 (35. Rg2 Qc7) 35... Kd7 (35... Qf8 36. Qh7 $18) 36. Rc1 Kd8 {Diagram [#]} 37. Qe3 $1 ({With play only on the kingside White can win a pawn, but can't break through for a win:} 37. Rg1 Bc8 38. h5 g5 {only move.} (38... gxh5 $2 39. Rxg8+ Qxg8 40. Qg7 $18) 39. Rxg5 Rxg5 40. Qxg5 Ke8 41. h6 Qh7 $11 42. Qg7 $4 Qxg7 43. fxg7 Kf7 44. Ke3 Bd7 45. Kf4 Bxb5 $1 $19 {and it turns out that the LSB is good for something after all.}) 37... Qf8 {Home laptops find the rest of the moves:} 38. Qc3 Qb4 39. Qxb4 axb4 40. Rg1 b3 41. Kc3 Bc8 42. Kxb3 Bd7 43. Kb4 Be8 {Diagram [#]} 44. Ra1 $1 {Creating an entry point for the R...} Kc7 45. a5 Bd7 46. axb6+ Kxb6 47. Ra6+ Kb7 48. Kc5 {... and the K.} Rd8 49. Ra2 Rc8+ 50. Kd6 Be8 51. Ke7 g5 52. hxg5 1-0 [Event "AlphaZero vs. Stockfish"] [Site "?"] [Date "2017.12.04"] [Round "1.10"] [White "AlphaZero"] [Black "Stockfish 8"] [Result "1-0"] [ECO "E17"] [Annotator "John Upper"] [PlyCount "111"] [EventDate "2017.??.??"] [Source "CFC Newsfeed"] [SourceDate "2017.12.08"] [WhiteTeam "England"] [BlackTeam "Norway"] [WhiteTeamCountry "ENG"] [BlackTeamCountry "NOR"] {Topalov prep + Tal Vision. This game features a mind-boggling sac on move 19, followed by an amazing use of the whole board to threaten the Black K while preventing Black from finishing queenside development.} 1. Nf3 Nf6 2. d4 e6 3. c4 b6 4. g3 Bb7 5. Bg2 Be7 6. O-O O-O 7. d5 exd5 8. Nh4 c6 9. cxd5 Nxd5 10. Nf5 Nc7 11. e4 d5 {Diagram [#]} 12. exd5 (12. Re1 Bf6 13. Nc3 Bc8 14. Qf3 Re8 15. Bd2 Nd7 16. exd5 Ne5 17. Qh5 g6 18. Nh6+ Kg7 19. Qd1 cxd5 {Moiseenko,A (2657) -Anton Guijarro,D (2650) Doha 2016 0-1 (75)}) (12. Nc3 Bf6 13. exd5 Nxd5 14. Nxd5 cxd5 15. Bf4 (15. Ne3 Na6 16. Nxd5 Nc5 {1/2-1/2 (16) Bobras,P (2565) -Macieja,B (2600) Germany 2008}) 15... Bxb2 16. Rb1 Bf6 17. Re1 Na6 18. Bd6 ( 18. Nd6 Bc6 19. Bxd5 Bxd5 20. Qxd5 Qd7 $15) 18... Re8 19. Qg4 g6 20. Nh6+ Kg7 21. Nf5+ Kg8 22. Nh6+ Kg7 23. Nf5+ $11 {1/2-1/2 (23) Ragger,M (2700)-Vitiugov, N (2721) Germany 2017}) 12... Nxd5 13. Nc3 Nxc3 14. Qg4 $146 g6 15. Nh6+ Kg7 16. bxc3 Bc8 17. Qf4 Qd6 18. Qa4 g5 {Diagram [#]} 19. Re1 $5 {Mind-boggling!! Another amazing sac for development/domination. It's worth noting that even at low ply, SF8 and the other top engines all rate White as having almost full compensation for the piece.} Kxh6 20. h4 f6 21. Be3 Bf5 22. Rad1 Qa3 23. Qc4 { White is down a piece and a pawn, but Black's Ra8 and Nb8 are missing in action.} b5 (23... Na6 24. Be4 Bxe4 25. Qxe4 Rae8 26. hxg5+ fxg5 27. Kg2 $1 $18 {and Black is getting mated.}) 24. hxg5+ fxg5 25. Qh4+ Kg6 26. Qh1 {Diagram [#] } Kg7 27. Be4 Bg6 28. Bxg6 hxg6 29. Qh3 Bf6 {only move.} (29... a6 $2 {just to illustrate:} 30. Bd4+ Bf6 31. Bxf6+ Rxf6 32. Rd8 $18) 30. Kg2 Qxa2 31. Rh1 Qg8 {only move.} {Diagram [#]} 32. c4 $5 {There are a few Tal games where he shows this kind of board vision, but I'm not Tal and even with computer assisstance I can't get my head around this position. I think the point of c4 is to stop ...Qa2+ after f2-f4... or deny it the use of c4 in some line.} (32. f4 g4 33. Qxg4 Bxc3 {unclear.} 34. Rd6 Rf6 35. Qh4 {unclear.}) 32... Re8 {Black can't move the N without allowing a killing Rd7+, and can't get the Ra8 into play with Ra7 (which would be taken by the Be3). White doesn't have any immediate threats, so Black has a lot of options. The notes cover a few, starting with the crudest, which illustrate some of the tactics:} (32... Na6 $4 33. Rd7+ Rf7 34. Qh6#) (32... c5 33. Bxc5 Nc6 34. Rd7+ Ne7 35. Bxe7 Bxe7 36. Rxe7+ Rf7 37. Qe6 $1 Rd8 (37... Raf8 38. Qe5#) 38. Rxf7+ Qxf7 39. Rh7+ $18) (32... a5 33. Rd6 a4 (33... bxc4 34. Qg4 Rf7 $2 35. Bd4 $18 {and there's no good defence to Qxg5.}) 34. Qh6+ Kf7 35. Rxf6+ Kxf6 36. Qxg5+ Ke6 37. Bc5 $18 Kd7 38. Bxf8 $1 {(was this the point of c3-c4? if the pawn was back on c4 then ...Qd5+ would win for Black.)} Qxf8 39. Rh7+ $18) (32... bxc4 $2 33. f4 $1 $18 g4 (33... a5 34. fxg5 Be5 35. Bc5 $18) 34. Qxg4 Rf7 35. f5 $1 $18) 33. Bd4 (33. Qh6+ Kf7 34. Bxg5 Qg7 35. Qh4 {only move.} Re5 36. Bxf6 Qxf6 37. Qh7+ Qg7 38. Qh4 {threatening Rd8, and may be just a draw after ...Qf6.}) 33... Bxd4 34. Rxd4 Rd8 (34... bxc4 35. g4 $3 $18 { threatening Qc3.}) 35. Rxd8 Qxd8 36. Qe6 {Diagram [#]} Nd7 {Black finally finishes developing the queenside, but at the cost of an exchange:} 37. Rd1 Nc5 38. Rxd8 Nxe6 39. Rxa8 Kf6 40. cxb5 cxb5 41. Kf3 Nd4+ 42. Ke4 Nc6 43. Rc8 Ne7 44. Rb8 Nf5 45. g4 Nh6 46. f3 Nf7 47. Ra8 Nd6+ 48. Kd5 Nc4 49. Rxa7 Ne3+ 50. Ke4 Nc4 51. Ra6+ Kg7 52. Rc6 Kf7 53. Rc5 Ke6 54. Rxg5 Kf6 55. Rc5 g5 56. Kd4 1-0

..

AlphaZero

Source: Canadian Chess