That meant Houdini was next in line as the top engine, but Rybka’s disqualification opened up opportunities for other engines like Stockfish to rise in prominence. Fritz was also a well-known engine in 2011, but it was generally considered weaker than Rybka and Houdini. Overall, by this time engines had become significantly stronger than human players with the best engines exceeding 3000 Elo.
In those days I only used the free engines and there were several strong free and open-source engines available. Stockfish was one of the most widely used.
Houdini 1.5a was another extremely strong engine as was Rybka 2.3.2a. Critter, Gull, Protector and Crafty were also very strong free engines.
Many of these engines could be used with free graphical user interfaces such as Arena and Scid, but I was using Fritz 12. It was released in November of 2009. I don't remember the cost, but I found the program in a long defunct office supply store whose name I have forgotten, but not the price. I bought the CD for $20 which was a fraction of the price from the USCF. It was my go-to program until I replaced it with Fritz 17 in 2021. It's still on my laptop and I actually use it to conduct Shootout analysis because it works better for that than the newer Fritz programs.
When playing the following game I used several different free engines and on occasion felt they were suggesting less than optimal moves.
In my old post-mortem analysis it was very difficult for me to determine what the best moves were, but my feeling was that black missed several chances to draw.
Black’s trouble seemed to stem from the decision at move 52 to trade his two R’s for my Q and a P. Despite the engine evaluations giving white a nearly 3 Pawn advantage, my gut feeling was that by keeping the R’s on he would have made it much more difficult for me to squeeze out a win.
All that is moot because today it’s clear that he was already dead lost at move 52 and he had a bad position for many moves before that. It was interesting to analyze the game using today’s Stockfish and comparing it to the old free Rybka engine.
[Event "Server Tournament"]
[Site "?"]
[Date "2011.??.??"]
[Round "?"]
[White "Tartajubow"]
[Black "Opponent"]
[Result "1-0"]
[ECO "E46"]
[Annotator "Stockfish/Rybka"]
[PlyCount "161"]
[EventDate "2011.??.??"]
[WhiteTeam ""]
[BlackTeam ""]
[WhiteTeamCountry ""]
[BlackTeamCountry ""]
{E46: Nimzo-Indian: Rubinstein: 4 e3 0-0 5 Ne2 without early Bd3} 1. c4 e6 2.
Nc3 Bb4 3. d4 Nf6 4. e3 O-O 5. Nge2 {This move was a specialty of Samuel
Reshevsky. White will first play a3 to kick the B away, before moving his N on
e2 to a more active square.} Nc6 {Judging by statistics, this rarely played
move isn’t the best. A note says that the database I was using at the time
had 6 master games in it and that all were lost by black. I suspect that black
played it because it was Rybka's top choice. 5...d5 is ny far the most popular
move.} 6. a3 {Black almost always plays 6...Bxc3+, but preserving the B with
this retreat is hardly bad.} Be7 (6... Bxc3+ 7. Nxc3 d5 8. b3 dxc4 9. bxc4 e5
10. d5 {Feuerstein,A-Denker,A New York 1957}) 7. d5 Ne5 8. Ng3 a6 {After this
there are no more games in the database so we are "out of the book" and white,
must come up with a plan. I considered several moves here before finally
deciding on my next move which has the idea of gaining time by harrassing thr
Ns.} (8... exd5 {is better.} 9. cxd5 d6 10. f4 Ng6 11. Be2 Re8 12. O-O Bf8 13.
Bc4 Ng4 14. Qe1 {The position is about equal. Jelen,I (2340)-Krasevec,A Bled
1992}) 9. f4 Ng6 10. h4 {The most aggressive...the N gets driven back.} Re8 (
10... Ne8 11. h5 Nh4 12. h6 g6 {is Rybka2.3.2a (multi-processor)'s preferred
line which it says is equal, but black's position just doesn't look right.
Stockfish agreesm giving white almost a 1.5 P advantage.}) 11. h5 Nf8 12. h6 {
White has to press on. Simple development seems to allow black to undermine
white’s P-center, bring his pieces out and get the initiative} (12. Bd3 h6 {
Apparently eignes, even the old Rybka, recognize the value of h6 so take steps
to prevent it.} 13. Qf3 d6 {and as cramed as black is, both Rybka and
Stockfish 17.1 evaluate the position as equal.}) 12... g6 13. Be2 {Best
according to Rybkam but Stockfish's 13.b4 is more logical. Black’s next move
represent a huge difference betweebn engines then and now! Black plays Rybka;s
top choice with an equal evaluation, Stocjfish agrees that the position is
equal, but only of black plays 13...b5. It thinks Rybka’s choice stinks.}
exd5 14. cxd5 Bc5 15. e4 d6 16. Qd3 c6 {An attempt to undermine white;s center.
It doesn't show up as a top choice for eother engine, but it's not a bad move.
Now Rybla suggests I play either 17.dxc6 or 17.b4 with equality. I played the
most aggressive move I could come up with which turns out to be Stockfish's
top choice.. Kudos to me!} 17. f5 Qb6 (17... gxf5 {is met ny} 18. Bg5 {and
black is lost after} Re5 19. Qf3 cxd5 20. Nxd5 N8d7 21. Nxf6+ Nxf6 22. b4 Bd4
23. Rd1 Bb6 24. Nh5) 18. Rf1 {[%mdl 2048] White is in control.} (18. dxc6 {
The move in th game is Stockfish's choice, Rybka likes this.} bxc6 19. Rf1 {
White is slightly better, but Rubka thinks the position is equal.}) 18... Bd4
19. Bg5 gxf5 {This can't be good. Better (slightly) wpuld have been 19...N8d7}
20. O-O-O {Offering the e-Pawn in order to unclog the K[side, Black has little
choice but to accept it.} Nxe4 21. Ngxe4 Rxe4 22. Bf3 Be3+ 23. Bxe3 {Black
faces a tough decision. Should he simplify by trading Qs or will keeping them
on give him better chances. It seems the choice is a matter of preferance
according to Rybka which gives white only a minimal edge after either capture.
Stockfish, on the ohter hand, thinks white is considerably betyter after 23...
Qxe3, but practically winning after taking with the R.} Rxe3 24. Qd2 Bd7 25.
Na4 Qa7 26. Qb4 {This move is far better than Rybka's recommendation of 26.
dxc6. The threat of Nb6 entices black into making a series of exchanges that
benefits white.} cxd5 27. Bxd5 Bxa4 28. Qxa4 Re5 29. Kb1 {[%mdl 1024] A safety
precaution.} Qe3 30. Qh4 {Intending Qf6 and mate.} (30. Bxb7 {looks plausible,
but it's deceiving.} Rb8 31. Qxa6 Qb3 32. Qxd6 Rb5 33. Qd2 {Threatening Qg5+
and mate, so...} R5xb7 (33... R8xb7 34. Qg5+ Ng6 35. Rd8#) 34. Qg5+ Ng6 35.
Rd8+ Rxd8 36. Qxd8+ Nf8 {and white has to take the draw with} 37. Qg5+) 30...
Ne6 31. Rf3 Qe2 32. Rfd3 {Black has defended against the mate threat soi now
white goes after the d-Pawn.} Re3 33. Bf3 {This tactical shot was seen by
Rybkam but it completely musevaluated the position thinking that white is only
slightly better when, in fact, white is better to the tune of nearly two Ps.}
Qxd3+ 34. Rxd3 Rxd3 35. Qh3 Rc8 36. Qxf5 Re3 37. Qd5 (37. Bxb7 {is a blunder...
} Re1+ 38. Ka2 Rcc1 {with equal chance after 39.Qg4+, but not the double
question mark worthy} 39. Bxa6 Ra1+ 40. Kb3 Nd4+) 37... Re1+ 38. Ka2 Rc6 39.
Qd2 Rcc1 40. Qxd6 a5 41. Qb8+ (41. Bxb7 {has been rendered impossible thanks
to nlack's last move.} a4 42. Qg3+ Kf8 43. Qd6+ Kg8 (43... Ke8 44. Bc6+ Rxc6
45. Qxc6+ Ke7 46. Qxa4 {White wins.}) 44. Qg3+ {and draws.}) 41... Nf8 42. Kb3
{It's still not safe to take the P.} Re6 43. Bd5 Rb6+ 44. Ka4 Rf1 45. Kxa5 {
Finally...all that remains os to picj up the b-Pawn.} Ra6+ 46. Kb4 Rb6+ 47. Kc3
Rxh6 48. Qxb7 Rd6 {Rybka is still giving black a fighting chance in evaluating
the position as only 1.5 Ps in white's favor. That's hard to belive
considering the B is superior to the N and f7 is very weak and white has two
connected passed Ps. White has a decisive advantage} 49. Bc4 Rc1+ 50. Kb3 Rd7
51. Qb6 Rb1 52. a4 {[%mdl 32]} Rd2 {Black is lost no matter what, but
Stockfish gives this a question mark.} 53. Qf6 {Rybka has finally recognized
that black is lost and actually recommended black's next move which is utterly
hopeless. But then so was anything else.} Rdxb2+ 54. Qxb2 Rxb2+ 55. Kxb2 {
[%mdl 4096] The remainder of the game requires no engine assistance!} Ng6 56.
a5 Ne7 57. a6 Nc6 58. Bd5 Na7 59. Kc3 Kf8 60. Kb4 Ke7 61. Kc5 Kd7 62. Kb6 Nc8+
63. Kb7 Nd6+ 64. Kb8 Nb5 65. Bxf7 Nd4 66. a7 Nc6+ 67. Kb7 Nxa7 68. Kxa7 Ke7 69.
Bh5 Ke6 70. Bg4+ {Rybka give white a 5 Pawn advantagem but Stockfish is
announcing mate in 33 moves.} Kf6 71. Bh3 Kg5 72. Kb6 Kf6 73. Kc6 Ke5 74. Kd7
Kf6 75. Kd6 h6 76. Kd5 Kg5 77. Ke5 h5 78. g3 Kg6 79. Kf4 Kf6 80. Bg2 Kg6 81.
Bf3 {Black resoigned. The finish is instructive though.} (81. Bf3 h4 82. g4 (
82. gxh4 {would be wrong! the B is the wrong color to force black out of the
corner.} Kh6 83. Kf5 Kg7 84. Kg5 Kh7 85. Be4+ Kg7 86. Bd3 Kh8 87. Kh6 Kg8 88.
Bf5 Kh8 {and you get the point. Advancing the P is no help.} 89. Kg6 Kg8 90. h5
Kh8 91. h6 Kg8 92. h7+ Kh8 93. Kg5 Kg7 {etc.}) 82... Kf6 83. g5+ Kg6 84. Be4+
Kf7 85. Kf5 Ke7 86. g6 h3 87. Ke5 h2 88. Bh1 Ke8 89. Ke6 Kf8 90. Kf6 Ke8 91. g7
Kd7 92. g8=Q Kc7 93. Qc4+ Kb6 94. Qb4+ Kc7 95. Ke7 Kc8 96. Qb7#) 1-0
No comments:
Post a Comment