Not logged inRybka Chess Community Forum
Up Topic The Rybka Lounge / Computer Chess / TCEC Season 21 - Division P
- By MrKris (****) Date 2021-06-23 01:17

Selected live chat:
Moderator28-Month Subscriber (2-Year Badge)VerifiedNightbot
: Shortest game ever in TCEC was the 9 moves draw (with 2 move book) in Season 10 Stage 1 of Texel vs Gull 1. b4 d5 2. Bb2 Qd6 3. a3 a5 4. Nf3 axb4 5. Be5 Qb6 6. Bd4 Qd6 7. Be5 Qb6 8. Bd4 Qd6 9. Be5 1/2-1/2 - see

{16...Bg4 threat to R(d1), 30...Bg4xRd1}
: Rook hanging for ten moves comes in handy now.

Game #15 just ended. Eight move book, Sf +2.06 move 11, +2.96 move 13, +4.08 move 16, etc., Mate in 27 move 23, Mates move 35.
[Event "TCEC Season 21 - Division P"]
[Site ""]
[Date "2021.06.22"]
[Round "4.3"]
[White "Stockfish 2021061923"]
[Black "Stoofvlees II a18"]
[Result "1-0"]
[Annotator "live"]

1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e6 6. g4 Be7 7. g5 Nfd7 8. h4 Nc6 {Book.}
9. Be3 {1.15} O-O 10. f4 Qb6 11. a3 {2.06} d5 12. exd5 Nxd4 13. Bxd4 {2.96} Bc5 14. Bxc5 Nxc5 15. Qd4 {3.29} exd5
16. O-O-O {4.08} Bg4 17. Nxd5 {6.14} Qa5 18. h5 {8.08} Ne6 19. Qe5 Qc5 20. b4 Qf2 21. Bd3 {17.60}
Rfd8 22. h6 {63.35} g6 23. Ne7+ {M29} Kf8 24. Nxg6+ {M27} hxg6 25. h7 {M25} Ke7
26. Qf6+ Kd7 27. Qxf7+ Kc6 28. Be4+ Kb5 29. Qxb7+ Qb6 30. h8=Q Bxd1 31. Qe5+ Nc5
32. Bc6+ Kc4 33. Qf7+ Ne6 34. Qfxe6+ Rd5 35. Q6xd5# 1-0
- - By MrKris (****) Date 2021-06-26 02:52
After 48 games, 12 each:

1. Sf  7.5/12  *
2. Lc  7.0/12  *
3. Dr  6.5/12  *
4. Sv  6.5/12
5. Sc  5.5/12
6. Et  5.0/12
7. Al  5.0/12
8. Ig  5.0/12

* undefeated except 1-1 win-loss in opening colors-reversed game pairs

The Stockfish-Stoofvlees, 1-0 by mate on move 35, game has all ready been mentioned twice on this forum.
Sorry, here is another 'take' on it: contrasting it with the colors reversed.
After Stockfish's 23.Ne7+ it said at TCEC "Mate in 29". Stockfish sacrificed its R on d1 - Stoofvlees took it 30...Bx(R)d1 with nothing better to do and totally lost.
r2r2k1/pp2Np1p/4n1pP/4Q1P1/1P3Pb1/P2B4/2P2q2/2KR3R b - - 1 23
Stockfish sacrificed the exchange: 21...Rx(N)c3 22.bxc3 Qc7 23.f5 Rc8 24.Bd2 Qd7 it said 0.67 and went on to draw on move 96.
2r3k1/1b1qbp1p/p2p2pP/1p2pPP1/4P1R1/2P5/P1PBQ3/1K1R4 w - - 3 25
Parent - - By sarona (**) Date 2021-07-04 23:56
Looks like TCEC Season 21 Premier Division is going to be very close.

Too bad K_Dragon has five losses against Stockfish and Lc0.  Splitting some of those points would have made things much tighter than it already is.

Parent - - By MrKris (****) Date 2021-07-05 05:10
Upcoming 2nd game/book, to not lose ground:
K'Dragon must win as white vs. Sf: Rd. 30.4 , Sf 1-0 KDr
r1b1kb1r/1p1n1ppp/pqn1p3/2ppP3/3P4/2PB1N2/PP1N1PPP/R1BQ1RK1 w kq - 3 9
K'Dragon must draw with black vs. Lc0: Rd. 32.4 , KDr 1/2-1/2 Lc0
r1bqk2r/3nppbp/P2p1np1/1BpP4/4P3/2N5/PP3PPP/R1BQK1NR w KQkq - 3 9

TCEC is playing the human standard RR: PlayerA wh. - PlayerB bk., then they play everyone else 1 game alternating colors, then get back to PlayerB wh. - PlayerA bk.
The humans can consider their standings before deciding on their colors-reversed game's opening.
Engines do not need that, I think TCEC should both colors-reversed games consecutively.
Otherwise the games pairs are always 'dangling' with only 1 played.
A trick by TCEC to get viewers to always 'stay tuned' for the 'upcoming exciting 2nd game' - to me it gets old.
Parent - - By sarona (**) Date 2021-07-06 02:13
Engines do not need that, I think TCEC should both colors-reversed games consecutively.
Otherwise the games pairs are always 'dangling' with only 1 played.
A trick by TCEC to get viewers to always 'stay tuned' for the 'upcoming exciting 2nd game' - to me it gets old.

I completely agree with your statement

I would prefer both colors-reversed games to be played one after the other. Lets say its SF v Stoofvlees, K_Dragon v Lc0, Ethereal v Scorpio and Igel v AllieStein.  Play both SF v Stoofvlees games before moving on to Dragon v Lc0.

It always seems that the game I really look forward to watching is being played live at 3 AM in the mountain time zone. . . :cry:
Parent - - By MrKris (****) Date 2021-07-06 17:27
Thanks.  I had not thought it much but the old fashioned way is probably best for humans, a 2nd game would be as different as any other next game so might as well let them complete the round 1st.

Top 3 vs. each other so far, with overall scores:
              Sf      Lc      Dr
Sf  6.5/12          ½½0½½½  101½1½   (26.0/40)
Lc  6.5/12  ½½1½½½          0101½½   (25.5/40)
Dr  5.0/12  010½0½  1010½½           (24.0/40)
Parent - By sarona (**) Date 2021-07-07 00:56 Upvotes 1
After thinking about it, you are right.

Reverse game pairs can be analyzed/viewed/downloaded from the TCEC archive. This way, comparisons can be done during an analytical session. That's probably better anyway than following games for a few hours or more.
Parent - - By MrKris (****) Date 2021-07-08 02:44
(I meant "the old fashioned way is probably best for humans players in a human RR".)

In any case:
your "That's [downloading] probably better anyway than following games for a few hours or more. "
Yes, the only time I watch TCEC is when I feel informed or entertained by the chatters (I remember when they typed BOOM or mooB, now the site has effects for it).

A side note: Engines same book color reversed really worked out well for the last TCEC Cup 8 I thought.

They reduced the requirement to only 1 winning game pair (1.5/2 or 2.0/2) needed to win the match - not sure of the exact rules.

The matches were all 3 games minimum to 8 games maximum. (Hard to read, see the "Score" column in Table 1 and Table 2: )

But the results look "good" to me.
Parent - - By sarona (**) Date 2021-07-11 03:28
your "That's [downloading] probably better anyway than following games for a few hours or more. "
Yes, the only time I watch TCEC is when I feel informed or entertained by the chatters (I remember when they typed BOOM or mooB, now the site has effects for it).

I will often watch the games in the evenings when I am on the road. Staying in hotels/motels or at various site camps.  There is often not much else to do. Especially in northern Alberta, Manitoba or Saskatchewan. I follow CCC as well. Actually prefer CCC because of the shorter time controls. The chat can be, at times, very entertaining as well.

Last month, while away from home, I was going through The Edge section of the forum. Following a lot of the threads dealing with Rybka 3 and the various Ippo clones. Kind of wish I was around then. Missed all of that. I don't even think Houdini had arrived on the scene yet.
Parent - - By MrKris (****) Date 2021-07-11 06:39
You missed a lot. There are still reams on this forum in the regular sections also about many exciting things!

Houdini 1.5 was widely accused of "being Ippo./Ivanhoe".
The Critter author said (in TalkChess I think) something close to "If you call much improved in every important aspect a clone..."

But he later claimed that he could make essentially Houdini from Ivanhoe. He "transferred" it to someone to "win a bet about it" using a forum then took it down. (Of course there were copies 'floating around' for a while.)

Note Ippo./Ivanhoe was not proved to infringe on Rybka but that [paraphrase] 'The authors understood much about Rybka' (evidentially something about uncopyrightable general ideas not using copyrightable code or so).
Lacking such proof its "public domain" status is valid so Houdini's use of it -IF it happened- was okay (he credited ideas from it on his site I believe).

Later the Houdini author warned the Critter author that [paraphrase] 'Any more of my Houdini -so unique that it could only come from a de-compilation- in a future Critter [beyond the last 1.6] I will go to court' [E.U. does not allow de-compilation, U.S. I have no idea]
Note interesting coincidences in the ratings and release dates of [numbers inexact] Critter 0.8?, Houdini 0.7?, Critter 1.0?, Houdini 1.0? through 1.5/1.6.
Parent - By Labyrinth (*****) Date 2021-07-11 21:23 Upvotes 1

>Houdini 1.5 was widely accused of "being Ippo./Ivanhoe".

Not so much Ippo/Ivanhoe, but heavily based on Robbolito, which it almost certainly was.

Also while there wasn't absolute proof, ippo was almost certainly based on reverse engineered Rybka 3 code.

Neither Rybka 3 or Houdini were open source, and reverse engineering them to a sufficient enough point to make a meaningful comparison would be a nightmare. To make matters worse Ippo was full of decompiled garbage, and while Robbolito cleaned it up a lot, it likely added some of its own stuff, so it was sort of impossible to provide the kind of rigorous comparison necessary to satisfy public opinion.

>Lacking such proof its "public domain" status is valid so Houdini's use of it -IF it happened- was okay

In my opinion, people looked the other way because Houdart, despite starting with Robbolito (imo), did a huge amount of development on Houdini and continued to develop it, such that he was making a genuine contribution to computer chess that really couldn't be ignored. A lot of the other engines weren't much stronger than ippo/robbo.

Unfortunately things got pretty ugly on the forum here, as Vas considered these early engines based on ippo to be pirated Rybka so the Rybka forum banned any mention of them. To make matters worse, mods were deleting whole threads instead of removing individual posts or blanking out names and links.

The author of Critter, Richard Vida, is supposedly pretty good at reverse engineering programs, and he often ragged on Houdart who he stated was taking code from other engines such as Stockfish. Then Houdart accused Rich of stealing his code, and yeah, argument.
- By MrKris (****) Date 2021-07-05 05:55
My favorite game I've seen so far, except the Stockfish-Stoofvlees minature above, is .
(Sf only drew as white) Lc0 1-0 Sf after 23.Nc6!!

At TCEC Stockfish played 23...bxc6 and lost on move 68.

Here are the 3 other alternaives from here, Stockfish_21070310, R7 2700X:

[FEN "r1bqn1k1/1p3pbp/1nNp2p1/pP1P4/P4B1P/2N2BP1/5P2/2RQ2K1 b - - 1 23"]

{41/57 18:2 1 wh. 2.85} 23. ... Qc7 24.Be3 bxc6 25.dxc6 Be6 26.Nd5 Bxd5 27.Bxd5 Rb8 28.Bb3 Bb2 29.Rb1 Be5 30.Qf3 h5 31.Bd2 Nf6 32.Bxa5 Kg7 33.Be1 Nc8 34.Qd3 Qb6 35.Rc1 Na7 36.Kg2 d5 37.Rc2 Re8 38.Bb4 Qc7 39.Ba2 Qd8 40.a5 Bc7 41.b6 Bxb6 42.axb6 Qxb6 43.Bc5 Qb5 44.Qxb5 Nxb5

[FEN "r1bqn1k1/1p3pbp/1nNp2p1/pP1P4/P4B1P/2N2BP1/5P2/2RQ2K1 b - - 1 23"]

{41/50 18:21 wh. 4.25} 23. ... Qd7 24.Be3 Nc4 25.Bd4 Ne5 26.Nxe5 dxe5 27.Be3 Bf8 28.Ne4 Qd8 29.d6 Bxd6 30.Nxd6 Qxd6 31.Qxd6 Nxd6 32.Rd1 Nc4 33.Rd8+ Kg7 34.Bc5 Bg4 35.Rxa8 Bxf3 36.Be7 h6 37.Bf8+ Kf6 38.Bxh6 Ke6 39.Bg5 Kf5 40.Re8 Nd6 41.Rd8 Ke6 42.Ra8 b6 43.Rb8 Bd1 44.Rxb6 Bxa4 45.Ra6 Bxb5 46.Rxa5 Kd7 47.Ra7+ Kc6

[FEN "r1bqn1k1/1p3pbp/1nNp2p1/pP1P4/P4B1P/2N2BP1/5P2/2RQ2K1 b - - 1 23"]

{41/49 18:21 wh. 7.84} 23. ... Qf6 24.Be3 bxc6 25.dxc6 Rb8 26.Bxb6 Rxb6 27.Nd5 Qd4 28.Qxd4 Bxd4 29.Rc4 Rxc6 30.bxc6 Bf6 31.Rc1 Kg7 32.Rb1 Be6 33.Rb7 Be5 34.Ra7 Bd4 35.Re7 Kf8 36.Rb7 Nf6 37.Rb8+ Kg7 38.c7 h6 39.Kf1 Be5 40.Bg2 Nxd5 41.Bxd5 Bxd5 42.c8Q g5 43.Qh8+ Kg6 44.Rg8+ Kf5 45.Qh7+ Ke6 46.Qxh6+ Ke7 47.Rxg5 Be6
- By MrKris (****) Date 2021-07-05 23:45
The only black win in the 153 games so far.

102 moves 38. through 140.: both probably overestimate white -it looks drawish to me-, evaluations very slowly decline, both do virtually nothing but shuffle.
Then Stoofvlees errors badly and loses fast.
[Event "TCEC Season 21 - Division P"]
[Site ""]
[Date "2021.07.03"]
[Round "33.2"]
[White "Stoofvlees II a18"]
[Black "KomodoDragon 2747.00"]
[Result "0-1"]
[Annotator "archive"]

1. e4 d5 2. exd5 Qxd5 3. Nc3 Qa5 4. d4 Nf6 5. Bd2 c6 6. Bc4 e6 7. Nf3 Qc7 8. O-O Bd6 9. Re1 O-O
10. Ne4 Nxe4 11. Rxe4 Nd7 12. Qe2 Rd8 13. a4 Nf8 14. a5 Bd7 15. b4 Be8 16. Bb3 b6 17. a6 Kh8
18. g3 Qe7 19. Re1 f6 20. Nh4 Qd7 21. c3 Bf7 22. Bf4 Bxf4 23. gxf4 f5 24. Re5 Bg8 25. Qh5 Rac8
26. Bc4 Rc7 27. Nf3 Qd6 28. Qg5 h6 29. Qg3 b5 30. Bb3 Rb8 31. R5e2 Rb6 32. Ra1 Qd8 33. Ne5 Qc8
34. Rea2 Nd7 35. Nd3 Nf6 36. Nc5 Nd5 37. Re2 Re7
[FEN "2q3bk/p3r1p1/Prp1p2p/1pNn1p2/1P1P1P2/1BP3Q1/4RP1P/R5K1 w - - 16 38"]

38. Rae1 {TCEC 2.32 @-wh. K on g2-@} Qb8 {TCEC 1.06} 39. Re5 Qd6 40. Kf1 Bf7 41. Ke2 Rb8
42. Kd2 Nc7 43. Rg1 Bg8 44. Kc2 Nd5 45. Kb2 {TCEC 2.32 @-wh. K on b2-@} Kh7 {TCEC 1.30} 46. Qf3 Nc7 47. Ra1 Bf7 48. Kc2 Nd5 49. Rg1 Rc8
50. Qg3 Bg8 51. Nd3 Kh8 52. Ree1 Rce8 53. Rg2 Nf6 54. Ne5 Bh7 55. Qf3 Ne4 56. Rgg1 Qc7 57. Ra1 Rb8
58. Nd3 Nf6 59. Kb2 Nd5 60. Rge1 Bg8 61. Ne5 Ree8 62. Bxd5 cxd5 63. Qh5 Bh7 64. Ra5 Re7 65. Qe2 Qb6
66. Kb3 Rc7 67. Nd3 Re7 68. Nc5 Bg6 69. Qe5 Rc8 70. Kc2 Bh5 71. Qe3 Qc6 72. Qh3 Bf7 73. Re3 Kh7
74. Kd2 {TCEC 2.15 @-wh. K on d2-@} Rb8 {TCEC 0.95} 75. Ra1 Rb6 76. Rg1 Qe8 77. Reg3 Bh5 78. R3g2 Bg4 79. Qg3 Kh8 80. Kc2 Rc6 81. Re1 Qf7
82. Qh4 Bh5 83. Rg3 Qe8 84. Rh3 Bg4 85. Rhe3 Kg8 86. Re5 Bh5 87. Qh3 Rd6 88. Qe3 Bf7 89. h3 Kf8
90. Kd2 Rb6 91. Kd3 Bg8 92. Kc2 Qf7 93. Qf3 Qf6 94. Rh1 Bf7 95. h4 Rc7 96. Ree1 Kg8 97. Reg1 Qd8
98. Rh3 Qe8 99. h5 Kh7 100. Kd2 Re7 101. Rh2 Kh8 102. Kc2 Kh7 103. Kd3 Rc7 104. Rh3 Re7
105. Ke2 {TCEC 1.47 @-wh. K on e2-@} Kh8 {TCEC 0.47} 106. Ke1 Kh7 107. Kd2 Rc6 108. Kc2 Rb6 109. Kd3 Rd6 110. Qe2 Rb6 111. Kd2 Qb8
112. Qe3 Qf8 113. Kc2 Kh8 114. Qe5 Be8 115. Kd2 Kh7 116. Qe2 Bf7 117. Rhg3 Be8 118. R3g2 Qf6
119. Rh2 Kh8 120. Qe3 Qf8 121. Qh3 Bf7 122. Rhg2 Be8 123. Rg3 Qg8 124. Qh4 Qf8 125. R3g2 Rc6
126. Kc2 Rec7 127. Kb3 Qe7 128. Qh2 Qd8 129. Rg3 Re7 130. Qh4 Qc7 131. Kc2 Rb6 132. R3g2 Rc6
133. f3 Rf7 134. Re2 Re7 135. Kd2 Bf7 136. Ke3 {TCEC 1.18 @-wh. K on e3-@}  Qd8 {TCEC 0.37} 137. Reg2 e5 138. Kf2 exd4 139. cxd4 Re2+
[FEN "3q3k/p4bp1/P1r4p/1pNp1p1P/1P1P1P1Q/5P2/4rKR1/6R1 w - - 1 140"]

140. Kg3 {TCEC -0.03} g5 {TCEC -0.10 <<<Finally drawish evaluations!} 141. Qh3 {?? Dragon TCEC PV: 141.Qh2 Re3 142.Re2!}
Re3 {TCEC -1.93 PV 142.Rf2} 142. fxg5 {?} hxg5 {TCEC -2.93 PV 143.Rf2} 143. Kf2 {?} Qe8 {TCEC -4.84 Black definitely winning.} 144. Qxf5 Bxh5 145. Rh1 Re2+ 146. Kg1 Re1+
147. Kh2 Rxh1+ 148. Kxh1 Bxf3 149. Qxf3 Rh6+ 150. Rh2 Qe1+ 151. Kg2 Qd2+ 152. Kf1 Qc1+ 153. Kg2 Qc2+
154. Kf1 Rxh2 155. Kg1 Rh4 156. Qf8+ Kh7 157. Qe7+ Kh6 158. Qd6+ Kh5 159. Ne6 Rg4+ 160. Kf1 Qg2+
161. Ke1 Qg1+ 162. Kd2 Rg2+ 163. Kd3 Qf1+ 164. Ke3 Qe2#
- - By MrKris (****) Date 2021-07-06 17:11
A 333 move game. AllieStein would not go into a drawish evaluation.
K'Dragon said 0.00 from move 104 through the end, for a solid 220 consecutive moves!

After 104. Re2 {0.43} Qh7 {0.00 -1st}
3br1k1/3br2q/1p1p1p2/pPpP1p1p/2P2PpP/1P1BP1P1/3QR2R/3N2K1 w - - 84 105

After 229. Kf2 {0.89 , had been lower; white is 2 pawns down with an 'advantage'?} Ke8 {0.00 -for the 123rd consecutive time}
3bk3/3b4/1p1p1p2/1PpP1p1p/2P2PpP/1p4P1/1Nr1BK2/1R6 w - - 8 230

After 328. Kh2 {0.23 -white still a pawn down, next move starts 0.00} Bd7 {0.00 -for the 225th consecutive time}
8/b2b4/1p1p1pk1/1PpP1p1p/2P2PpP/6P1/2R2N1K/r2B4 w - - 91 329
Parent - - By sarona (**) Date 2021-07-11 04:23 Upvotes 1
I noticed in the reverse game, Allie decided to play 9...cxd5 (instead of 9...c5) and wound up getting mated in 56 moves. :confused:
Parent - By MrKris (****) Date 2021-07-12 04:47
I guess some of these might be worth watching :smile:

In fact, at one point I wondered if TCEC and CCC should play more of just the top 3.
But of course there are a lot of instructive games (or 'what not to play vs. a top engine' :smile:).

Did you know that at first TCEC meant "Thorensen Chess Engine Championship".
He began it on his own early i7 quad core when many people had less.
His main point was, in addition to a good live site, he seemed the only one diligent enough to do all the qualifiers through final carefully and continuously at long TC for how ever long it took.
- By MrKris (****) Date 2021-07-12 03:02
50 games, 25 openings, G/90'+9", Premier 3rd vs. 4th match. For the official 'Season 3rd' place.

Dragon & Stoofvlees will be in the Intrafinal.
Head-to-head: the most one sided were in a 5-way tie with 4 wins and 4 draws:
Sf-Et, Lc-Sc, Lc-Ig, Dr-Sv, Sv-Et.

Stockfish & Lc0 will be in the Superfinal.
These two had by far the least losses: Sf 2, Lc 3, Dr 6, Sv 13 ...

S21 Div P - 1 game remaining

1 Sf  36.5/55 Sf-Ig in progress
2 Lc  36.5/56
3 Dr  33.5/56
4 Sv  28.0/56
5 AS  25.5/56
6 Sc  24.0/56
7 Ig  22.0/55 Sf-Ig in progress
8 Et  18.0/56

The shortest win was the Stockfish - Stoofvlees game already mentioned.
Here is the 2nd shortest:
White was winning at move 21 or so.
- 1st Lc0 pawn sacrifice 12.Rg1
- 2nd Lc0 pawn sacrifice 16.Bd3
[Event "TCEC Season 21 - Division P"]
[Site ""]
[Date "2021.07.08"]
[Round "47.1"]
[White "LCZero 0.28.0-rc1_69146"]
[Black "AllieStein v0.8-120f959_PS1.7"]
[Result "1-0"]
[Annotator "archive"]

1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. f3 O-O 6. Be3 e5 7. d5 Nbd7 8. Qd2 Nh5 {Book.}
9. g4 {TCEC 0.42} Nf4 10. O-O-O Qh4 11. Nge2 h5 12. Rg1 Nb6 13. Nxf4 exf4 14. Bf2 Qxh2 15. gxh5 Qxh5
16. Bd3 Nd7 17. Rh1 {TCEC 1.36} Qxf3 18. Rh2 Bxc3 19. bxc3 Ne5 20. Bd4 Bd7 21. Rdh1 {TCEC 3.12} Rfe8 22. Bc2 a5
23. Qe1 Qg3 24. Bf2 Qg4 25. Rh4 f5 26. Rxg4 fxg4 27. Bd4 g3 28. Rh4 g5 29. Rh5 Kg7 30. Qf1 Kf6
31. Qh1 Rg8 32. Bd1 Be8 33. Rh6+ Bg6 34. Bh5 Ke7 35. Bxg6 Raf8 36. Rh7+ Kd8 37. Qh3 g2 38. Qe6 Rf7
39. Bxf7 Rf8 40. Qf6+ Kc8 41. Be6+ Kb8 42. Qxf8# 1-0

Dragon had the only black win in 224 games.
Dragon won with white in that reversed, for the only won-with-both-colors in the 112 opening game pairs.

TCEC has an excellent Stats button - upper right - that shows game pair and more info., also links to the games and relevant game lists.
- By MrKris (****) Date 2021-07-12 04:28
Nightbot [official TCEC chatter]
: Now DivP, then updates+testing+!InfraFinal, then updates+testing+Sufi, then Bonus games ( !VVLTC , LCZeroCPU vs DivP), then Cup9, FRC4 and Swiss2 (current plan, can change, also !4k bonus and/or !r-mobility (armageddon) demo or bonus might be inserted somewhere (1 engine minimum with reasonable support needed)) ⭐

Last game: Stockfish 1-0 Igel
'Proof that the a- and h-pawns just get in the way.'

I had given this up for a draw somewhere before:
35. Re1 {TCEC 0.53, PV 35...Qe6} Qf8?! 36. Bh3 {TCEC 0.76}

White is mounting a strong K-side attack: 43. Bf5! {preparing the next move} Bc8 44. Bb1! {very strong on this diagonal} Rb6 {also Sf's PV} 45. a5!! {TCEC 1.54, PV 45...Rd6}
Igel chased white's a-pawn, 'forgetting' to help defend its 6th rank in the coming attack as in Sf's PV, with 45...Rb5?? :
2b2rk1/2n1qpp1/p4b1p/Pr1pN3/1ppP1P1P/4PQP1/1P2N1RK/1B3R2 w - - 1 46

Note that black is in effect a whole R down on most of the board, all of the c- through h-files.
Then Stockfish sacrificed its h-pawn with 46. g4!! {TCEC 2.69} with essentially a winning evaluation, especially vs. a weaker opponent and no ending oddities such as OCB's.

The h-file (allowing white's R on the 7th), strong white B and useless black R themes came to a glaring conclusion after 57. Rxc7 {TCEC 6.42 +-, recovering the N it sac'ed earlier, from h7 protected by its Bb1}
2b2rk1/2R5/p4p2/3pP3/1pp2PP1/4P3/1P6/rB3RK1 b - - 0 57

Black played the same as Sf's Pv 57...Rxb1 which of course just delayed the end.
Up Topic The Rybka Lounge / Computer Chess / TCEC Season 21 - Division P

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill