Chess Game Analyzer

[Carlos777]

Table:

Code: Select all

Decisive advantage [+- or -+] score >= +3.0 pawns
Moderate advantage &#91;+/- or -/+&#93; score >= +1.5 and score < +3.0
Slight advantage &#91;+/= or =/+&#93; score >= +0.25 and score < +1.5

Sample annotations with summary of number of blunder, mistakes and dubious moves for each side at the end of the game as comments.

[pgn][Event "4th Sinquefield Cup 2016"]
[Site "Saint Louis"]
[Date "2016.08.11"]
[Round "6.1"]
[White "Aronian, Levon"]
[Black "Vachier-Lagrave, Maxime"]
[Result "0-1"]
[Annotator "Brainfish 080816 64 POPCNT (5.0s/pos, thread=1)"]
[BlackElo "2819"]
[ECO "A37"]
[TimeControl "60"]
[WhiteElo "2792"]

1.Nf3 c5 2.c4 Nc6 3.Nc3 e5 4.g3 g6 5.Bg2 Bg7 6.O-O Nge7 7.a3 O-O 8.b4 $10 { -0.08/18 } ( { Better is } 8.d3 d6 9.Rb1 b6 10.b4 Bb7 11.Bg5 $10 { +0.20/17 } ) ( { Not to be underestimated is } 8.Re1 d6 9.d3 b6 10.Bd2 Qd7 11.b4 $10 { +0.12/17 } ) ( { Black is threatening } 8.-- d5 $10 { -0.10/20 } ) 8...d5 9.cxd5 Nxd5 10.Ng5 Nc7 11.Nge4 c4 12.d3 cxd3 13.Bg5 $6 $15 { -0.99/21 } ( { More accurate is } 13.Qxd3 $1 Qxd3 14.exd3 f5 15.Nc5 e4 16.Bb2 $10 { -0.19/18 } ) ( { Also playable is } 13.exd3 f5 14.Bg5 Qd7 15.Nc5 Qf7 16.Bd2 $15 { -0.41/18 } ) ( { Black is threatening } 13.-- dxe2 14.Qxe2 $15 { -1.47/19 } ) 13...f6 14.Be3 f5 15.Bg5 Qd4 16.Be3 Qd8 $3 17.Bg5 Qd7 $3 18.Nc5 dxe2 $15 { -0.69/21 } ( { More accurate is } 18...Qf7 19.Rc1 dxe2 20.Nxe2 f4 21.Nxb7 Bxb7 $15 { -1.31/18 } ) ( { White is threatening } 18...-- 19.Nxd7 Bxd7 $18 { +6.59/18 } ) 19.Nxe2 Qxd1 20.Rfxd1 $15 { -1.05/19 } ( { Better is } 20.Raxd1 e4 21.f3 b6 22.Nb3 Ba6 23.Rfe1 $15 { -0.80/17 } ) ( { Black is threatening } 20.-- Qxe2 $19 { -16.16/18 } ) 20...f4 $15 { -0.69/19 } ( { More accurate is } 20...e4 21.Rab1 Nb5 22.a4 Ncd4 23.Nxd4 Nxd4 $15 { -1.24/17 } ) 21.gxf4 $15 { -1.42/19 } ( { More accurate is } 21.Nc3 h6 22.Bxc6 bxc6 23.Be7 Re8 24.Bd6 $15 { -0.72/18 } ) ( { Also sufficient is } 21.Bxc6 bxc6 22.Nc3 a5 23.Rac1 Bg4 24.Rd6 $15 { -1.03/18 } ) ( { Black is threatening } 21.-- f3 22.Bf1 fxe2 23.Bxe2 Nd4 24.Bc4+ Kh8 $19 { -3.83/21 } ) 21...h6 22.b5 Nxb5 23.Rab1 Nxa3 24.Rb3 hxg5 $15 { -0.91/19 } ( { More accurate is } 24...Nc2 $1 25.Nxb7 N2d4 26.Nxd4 Nxd4 27.Rxd4 exd4 $17 { -2.05/17 } ) ( { But not } 24...exf4 $2 { considering } 25.Bxf4 Nc2 26.Bd6 Re8 27.Nxb7 $15 { -0.83/17 } ) ( { White is threatening } 24...-- 25.Nxb7 Bg4 26.Bd5+ Kh7 27.f3 Bf5 28.Rxa3 $14 { +1.18/19 } ) 25.Rxa3 $2 $17 { -1.94/21 } ( { More accurate is } 25.Nxb7 Ne7 26.Nd8 Bf5 27.Bxa8 Bc2 28.Rbd3 $15 { -0.98/17 } ) ( { Black is threatening } 25.-- Nc2 $19 { -3.70/20 } ) 25...exf4 26.Bd5+ Kh7 27.Ne4 Kh6 28.Nd6 Bf6 29.Bxc6 bxc6 $7 30.Nxc8 Raxc8 $15 { -1.34/21 } ( { More accurate is } 30...Rfxc8 31.Rd7 Kh5 32.Kg2 Kg4 33.h3+ Kf5 $17 { -1.85/21 } ) ( { White is threatening } 30...-- 31.Nd6 $14 { +1.08/21 } ) 31.Rd7 g4 32.Nxf4 Rcd8 $15 { -0.96/21 } ( { Better is } 32...Bb2 33.Ra4 c5 34.h4 gxh3 35.Ne6 Rf5 $15 { -1.32/21 } ) ( { Also promising is } 32...Be5 33.Nd3 Bb8 34.Rc3 a5 35.Rc4 Rcd8 $15 { -1.23/21 } ) 33.Raxa7 Rxd7 34.Rxd7 c5 $1 35.Ne6 $2 $17 { -1.62/23 } ( { More accurate is } 35.Kg2 Bh4 36.Nd3 c4 37.Ne5 Rxf2+ 38.Kh1 $15 { -0.90/21 } ) ( { Another possibility is } 35.Ra7 Bd4 36.Ne6 Bxf2+ 37.Kg2 Rf6 38.Ra4 $15 { -1.24/21 } ) ( { Black is threatening } 35.-- Bd4 $17 { -1.68/24 } ) 35...Rc8 $7 36.Rc7 $4 $19 { -3.15/26 } ( { A nice try could be } 36.Rf7 Be5 37.f4 gxf3 38.Kf2 c4 39.Rxf3 $17 { -1.83/20 } ) ( { Black is threatening } 36.-- c4 37.Rd2 Rc6 38.Rc2 Rxe6 39.Rxc4 Kh5 $19 { -3.64/23 } ) 36...Rxc7 $7 37.Nxc7 c4 $1 38.Kf1 Kg5 $3 39.Nd5 Be5 40.Ke2 Bxh2 41.Ke3 Be5 42.Nb4 $19 { -5.05/29 } ( { A nice try could be } 42.Ke4 c3 43.Ne3 Bf4 44.Nc2 Bc7 45.Kd3 $19 { -3.91/26 } ) ( { Black is threatening } 42.-- Bb8 $19 { -4.36/29 } ) 42...Bd6 43.Nd5 Bc5+ 44.Ke2 Kf5 45.Nc3 Ke5 46.Nb5 $19 { -7.61/27 } ( { A nice try could be } 46.Kf1 Bd4 47.Na2 c3 48.Nb4 Bc5 49.Nd3+ $19 { -5.75/24 } ) ( { Black is threatening } 46.-- Bd4 $19 { -6.93/28 } ) 46...Bb4 47.Ke3 g5 $1 48.Nc7 Bc5+ 49.Ke2 c3 { WBlunder: 1, WMistake: 2, WDubious: 1, BBlunder: 0, BMistake: 0, BDubious: 0 } 0-1[/pgn]

It is a subjective matter, but in my opinion, it should be:

Decisive: >= 2.0
Moderate: >=1.0 and < 2.0
Slightly adv.: >=0.25 and < 1.0

OTOH, I don't understand why in that game, black's 16th and 17th moves deserved double exclamation marks.

[Ferdy]

OTOH, I don't understand why in that game, black's 16th and 17th moves deserved double exclamation marks.

Double exclamation mark is given to a move of a player when player move and engine move are the same, and its complexity value thru number of root move changes during analysis is greater than or equal to 3.
A high number of move changes indicate that this position is not easy to evaluate as even the engine has difficulties resolving it.

For black's 16th move. Starting at depth 9 compare the pv move, if it changes count it.

position fen r1b2rk1/ppn3bp/2n3p1/4pp2/1P1qN3/P1NpB1P1/4PPBP/R2Q1RK1 b - - 3 16
go movetime 5000
info depth 1 seldepth 1 multipv 1 score cp 36 nodes 214 nps 107000 tbhits 0 time 2 pv d4d7
info depth 2 seldepth 2 multipv 1 score cp 134 nodes 328 nps 109333 tbhits 0 time 3 pv d4d8 e2d3 f5e4 c3e4
info depth 3 seldepth 4 multipv 1 score cp 152 nodes 478 nps 119500 tbhits 0 time 4 pv d4d7 e2d3 f5e4
info depth 4 seldepth 4 multipv 1 score cp 152 nodes 614 nps 122800 tbhits 0 time 5 pv d4d7 e2d3 f5e4 c3e4
info depth 5 seldepth 6 multipv 1 score cp 146 nodes 1629 nps 232714 tbhits 0 time 7 pv d4d8 b4b5 c6d4 e3d4 f5e4
info depth 6 seldepth 7 multipv 1 score cp 85 nodes 3632 nps 330181 tbhits 0 time 11 pv d4d8 e4c5 e5e4 d1b3 g8h8 e2d3 e4d3
info depth 7 seldepth 8 multipv 1 score cp 75 nodes 7466 nps 439176 tbhits 0 time 17 pv d4d7 e4c5 d7f7 g2c6 b7c6 e2d3 e5e4
info depth 8 seldepth 10 multipv 1 score cp 90 nodes 9341 nps 467050 tbhits 0 time 20 pv d3e2 d1e2 d4d7 e2a2 d7e6 e4c5
info depth 9 seldepth 15 multipv 1 score cp 98 nodes 23991 nps 631342 tbhits 0 time 38 pv d4d8 e3g5 d3e2 d1e2 d8d7 e4c5
info depth 10 seldepth 15 multipv 1 score cp 134 nodes 62263 nps 788139 tbhits 0 time 79 pv d4d7 e4c5 d7f7 b4b5 d3e2 d1e2
info depth 11 seldepth 18 multipv 1 score cp 85 nodes 113887 nps 831291 tbhits 0 time 137 pv d4d8 e3g5 d8d7 e4c5 d7f7 g2c6
info depth 12 seldepth 19 multipv 1 score cp 117 nodes 186597 nps 914691 tbhits 0 time 204 pv d4d8 e3g5 d8d7 e4c5 d7f7 e2d3
info depth 13 seldepth 20 multipv 1 score cp 103 nodes 235625 nps 920410 tbhits 0 time 256 pv d4d7 e4c5 d3e2 c5d7 e2d1q f1d1
info depth 14 seldepth 21 multipv 1 score cp 87 nodes 373488 nps 924475 tbhits 0 time 404 pv d4d8 e3g5 d8d7 e4c5 d3e2 c5d7
info depth 15 seldepth 22 multipv 1 score cp 90 nodes 558598 nps 979996 tbhits 0 time 570 pv d4d8 e3g5 d8e8 e4d6 e8e6 d6c8
info depth 16 seldepth 25 multipv 1 score cp 81 nodes 1134161 nps 1007247 hashfull 508 tbhits 0 time 1126 pv d4d7 e4c5 d3e2
info depth 17 seldepth 28 multipv 1 score cp 115 nodes 1907301 nps 1005430 hashfull 734 tbhits 0 time 1897 pv d4d8 e3g5 d8d7
info depth 18 seldepth 32 multipv 1 score cp 102 nodes 2557481 nps 994354 hashfull 852 tbhits 0 time 2572 pv d4d8 e3g5 d8d7 e4c5
info depth 19 seldepth 32 multipv 1 score cp 127 nodes 4541472 nps 965654 hashfull 974 tbhits 0 time 4703 pv d4d8 e3g5
info nodes 4808571 time 5000
bestmove d4d8 ponder e3g5

depth 9, pv d4d8
depth 10, pv d4d7 ---> number of changes = 1
depth 11, pv d4d8 ---> number of changes = 2
and so on

Number of changes = 6, so this is !!
if number of changes is 2, it will be given !

[Carlos777]

OTOH, I don't understand why in that game, black's 16th and 17th moves deserved double exclamation marks.

Double exclamation mark is given to a move of a player when player move and engine move are the same, and its complexity value thru number of root move changes during analysis is greater than or equal to 3.
A high number of move changes indicate that this position is not easy to evaluate as even the engine has difficulties resolving it.

I disagree. With this reasoning, you would get exclamation marks in positions where there are many good moves and not necessarily difficult to get. In the example, Qd8 or Qd7 are almost forced and are not brilliant moves.

I think an exclamation mark should be related to a sound material sacrifice, a move found after x time or y plies (difficult positions) and/or when the move is the only winning one according to the TB or engine's evaluation.

[Guenther]

OTOH, I don't understand why in that game, black's 16th and 17th moves deserved double exclamation marks.

Double exclamation mark is given to a move of a player when player move and engine move are the same, and its complexity value thru number of root move changes during analysis is greater than or equal to 3.
A high number of move changes indicate that this position is not easy to evaluate as even the engine has difficulties resolving it.

For black's 16th move. Starting at depth 9 compare the pv move, if it changes count it.

position fen r1b2rk1/ppn3bp/2n3p1/4pp2/1P1qN3/P1NpB1P1/4PPBP/R2Q1RK1 b - - 3 16
go movetime 5000
info depth 1 seldepth 1 multipv 1 score cp 36 nodes 214 nps 107000 tbhits 0 time 2 pv d4d7
info depth 2 seldepth 2 multipv 1 score cp 134 nodes 328 nps 109333 tbhits 0 time 3 pv d4d8 e2d3 f5e4 c3e4
info depth 3 seldepth 4 multipv 1 score cp 152 nodes 478 nps 119500 tbhits 0 time 4 pv d4d7 e2d3 f5e4
info depth 4 seldepth 4 multipv 1 score cp 152 nodes 614 nps 122800 tbhits 0 time 5 pv d4d7 e2d3 f5e4 c3e4
info depth 5 seldepth 6 multipv 1 score cp 146 nodes 1629 nps 232714 tbhits 0 time 7 pv d4d8 b4b5 c6d4 e3d4 f5e4
info depth 6 seldepth 7 multipv 1 score cp 85 nodes 3632 nps 330181 tbhits 0 time 11 pv d4d8 e4c5 e5e4 d1b3 g8h8 e2d3 e4d3
info depth 7 seldepth 8 multipv 1 score cp 75 nodes 7466 nps 439176 tbhits 0 time 17 pv d4d7 e4c5 d7f7 g2c6 b7c6 e2d3 e5e4
info depth 8 seldepth 10 multipv 1 score cp 90 nodes 9341 nps 467050 tbhits 0 time 20 pv d3e2 d1e2 d4d7 e2a2 d7e6 e4c5
info depth 9 seldepth 15 multipv 1 score cp 98 nodes 23991 nps 631342 tbhits 0 time 38 pv d4d8 e3g5 d3e2 d1e2 d8d7 e4c5
info depth 10 seldepth 15 multipv 1 score cp 134 nodes 62263 nps 788139 tbhits 0 time 79 pv d4d7 e4c5 d7f7 b4b5 d3e2 d1e2
info depth 11 seldepth 18 multipv 1 score cp 85 nodes 113887 nps 831291 tbhits 0 time 137 pv d4d8 e3g5 d8d7 e4c5 d7f7 g2c6
info depth 12 seldepth 19 multipv 1 score cp 117 nodes 186597 nps 914691 tbhits 0 time 204 pv d4d8 e3g5 d8d7 e4c5 d7f7 e2d3
info depth 13 seldepth 20 multipv 1 score cp 103 nodes 235625 nps 920410 tbhits 0 time 256 pv d4d7 e4c5 d3e2 c5d7 e2d1q f1d1
info depth 14 seldepth 21 multipv 1 score cp 87 nodes 373488 nps 924475 tbhits 0 time 404 pv d4d8 e3g5 d8d7 e4c5 d3e2 c5d7
info depth 15 seldepth 22 multipv 1 score cp 90 nodes 558598 nps 979996 tbhits 0 time 570 pv d4d8 e3g5 d8e8 e4d6 e8e6 d6c8
info depth 16 seldepth 25 multipv 1 score cp 81 nodes 1134161 nps 1007247 hashfull 508 tbhits 0 time 1126 pv d4d7 e4c5 d3e2
info depth 17 seldepth 28 multipv 1 score cp 115 nodes 1907301 nps 1005430 hashfull 734 tbhits 0 time 1897 pv d4d8 e3g5 d8d7
info depth 18 seldepth 32 multipv 1 score cp 102 nodes 2557481 nps 994354 hashfull 852 tbhits 0 time 2572 pv d4d8 e3g5 d8d7 e4c5
info depth 19 seldepth 32 multipv 1 score cp 127 nodes 4541472 nps 965654 hashfull 974 tbhits 0 time 4703 pv d4d8 e3g5
info nodes 4808571 time 5000
bestmove d4d8 ponder e3g5

depth 9, pv d4d8
depth 10, pv d4d7 ---> number of changes = 1
depth 11, pv d4d8 ---> number of changes = 2
and so on

Number of changes = 6, so this is !!
if number of changes is 2, it will be given !

Hi Ferdinand,

I also think this should be refined. Practically it changes only between two candidate moves (and with the same piece too!) from depth nine on and I don't think this can be really called a high level of complexity.
(Moreover double exclamation marks should be given very sparsely)

My suggestions:
With todays hardware I don' think depths below 14 or 15 should be considered for the annotation process at all.
Then the level of complexity (for !!) should be perhaps 4 or more different candidate moves and only if it does not concern the same piece to move.
Also that move should not appear already in early depths and later changed back.
For Humans those exclamation marks really look crazy here.

Guenther

[Ferdy]

OTOH, I don't understand why in that game, black's 16th and 17th moves deserved double exclamation marks.

Double exclamation mark is given to a move of a player when player move and engine move are the same, and its complexity value thru number of root move changes during analysis is greater than or equal to 3.
A high number of move changes indicate that this position is not easy to evaluate as even the engine has difficulties resolving it.

I disagree. With this reasoning, you would get exclamation marks in positions where there are many good moves and not necessarily difficult to get.

Stockfish is very strong, when it changes its move often starting at depth 9 there is something in the position.

In the example, Qd8 or Qd7 are almost forced and are not brilliant moves.

It may look simple, but finding the best move among some tempting candidate moves is the most difficult one to decide over the board, the fact that Stockfish at 5s per move also experience difficulties just proves that this position is not easy to assess. But in the end Stockfish agrees with the human move, so with the current method, I give a !! to the human move, he deserves it. There are only few times that engines like Stockfish get confused about the position. Another condition that I used is that the score >= +0.25. Now I am changing it to score >= +1.5, with this, lesser !! are appended.

[Ferdy]

OTOH, I don't understand why in that game, black's 16th and 17th moves deserved double exclamation marks.

Double exclamation mark is given to a move of a player when player move and engine move are the same, and its complexity value thru number of root move changes during analysis is greater than or equal to 3.
A high number of move changes indicate that this position is not easy to evaluate as even the engine has difficulties resolving it.

For black's 16th move. Starting at depth 9 compare the pv move, if it changes count it.

position fen r1b2rk1/ppn3bp/2n3p1/4pp2/1P1qN3/P1NpB1P1/4PPBP/R2Q1RK1 b - - 3 16
go movetime 5000
info depth 1 seldepth 1 multipv 1 score cp 36 nodes 214 nps 107000 tbhits 0 time 2 pv d4d7
info depth 2 seldepth 2 multipv 1 score cp 134 nodes 328 nps 109333 tbhits 0 time 3 pv d4d8 e2d3 f5e4 c3e4
info depth 3 seldepth 4 multipv 1 score cp 152 nodes 478 nps 119500 tbhits 0 time 4 pv d4d7 e2d3 f5e4
info depth 4 seldepth 4 multipv 1 score cp 152 nodes 614 nps 122800 tbhits 0 time 5 pv d4d7 e2d3 f5e4 c3e4
info depth 5 seldepth 6 multipv 1 score cp 146 nodes 1629 nps 232714 tbhits 0 time 7 pv d4d8 b4b5 c6d4 e3d4 f5e4
info depth 6 seldepth 7 multipv 1 score cp 85 nodes 3632 nps 330181 tbhits 0 time 11 pv d4d8 e4c5 e5e4 d1b3 g8h8 e2d3 e4d3
info depth 7 seldepth 8 multipv 1 score cp 75 nodes 7466 nps 439176 tbhits 0 time 17 pv d4d7 e4c5 d7f7 g2c6 b7c6 e2d3 e5e4
info depth 8 seldepth 10 multipv 1 score cp 90 nodes 9341 nps 467050 tbhits 0 time 20 pv d3e2 d1e2 d4d7 e2a2 d7e6 e4c5
info depth 9 seldepth 15 multipv 1 score cp 98 nodes 23991 nps 631342 tbhits 0 time 38 pv d4d8 e3g5 d3e2 d1e2 d8d7 e4c5
info depth 10 seldepth 15 multipv 1 score cp 134 nodes 62263 nps 788139 tbhits 0 time 79 pv d4d7 e4c5 d7f7 b4b5 d3e2 d1e2
info depth 11 seldepth 18 multipv 1 score cp 85 nodes 113887 nps 831291 tbhits 0 time 137 pv d4d8 e3g5 d8d7 e4c5 d7f7 g2c6
info depth 12 seldepth 19 multipv 1 score cp 117 nodes 186597 nps 914691 tbhits 0 time 204 pv d4d8 e3g5 d8d7 e4c5 d7f7 e2d3
info depth 13 seldepth 20 multipv 1 score cp 103 nodes 235625 nps 920410 tbhits 0 time 256 pv d4d7 e4c5 d3e2 c5d7 e2d1q f1d1
info depth 14 seldepth 21 multipv 1 score cp 87 nodes 373488 nps 924475 tbhits 0 time 404 pv d4d8 e3g5 d8d7 e4c5 d3e2 c5d7
info depth 15 seldepth 22 multipv 1 score cp 90 nodes 558598 nps 979996 tbhits 0 time 570 pv d4d8 e3g5 d8e8 e4d6 e8e6 d6c8
info depth 16 seldepth 25 multipv 1 score cp 81 nodes 1134161 nps 1007247 hashfull 508 tbhits 0 time 1126 pv d4d7 e4c5 d3e2
info depth 17 seldepth 28 multipv 1 score cp 115 nodes 1907301 nps 1005430 hashfull 734 tbhits 0 time 1897 pv d4d8 e3g5 d8d7
info depth 18 seldepth 32 multipv 1 score cp 102 nodes 2557481 nps 994354 hashfull 852 tbhits 0 time 2572 pv d4d8 e3g5 d8d7 e4c5
info depth 19 seldepth 32 multipv 1 score cp 127 nodes 4541472 nps 965654 hashfull 974 tbhits 0 time 4703 pv d4d8 e3g5
info nodes 4808571 time 5000
bestmove d4d8 ponder e3g5

depth 9, pv d4d8
depth 10, pv d4d7 ---> number of changes = 1
depth 11, pv d4d8 ---> number of changes = 2
and so on

Number of changes = 6, so this is !!
if number of changes is 2, it will be given !

Hi Ferdinand,

I also think this should be refined. Practically it changes only between two candidate moves (and with the same piece too!) from depth nine on and I don't think this can be really called a high level of complexity.
(Moreover double exclamation marks should be given very sparsely)

My suggestions:
With todays hardware I don' think depths below 14 or 15 should be considered for the annotation process at all.
Then the level of complexity (for !!) should be perhaps 4 or more different candidate moves and only if it does not concern the same piece to move.
Also that move should not appear already in early depths and later changed back.
For Humans those exclamation marks really look crazy here.

Guenther

I am basically relying on how Stockfish feels about the position. We already knew its strength (we are now at a point where some of its moves are difficult to understand and it is increasing) and to think that it gets confused on some positions is just crazy.

I will probably increase it to 4 or more for !! and 3 for !. Also when I increase the score to >= +1.5 (was >= +0.25), I have already seen lesser !! and !.

[EvgeniyZh]

No, if score is +6, than a move which lowers the score to +3 is not a blunder, it is an inaccuracy.

This is what I found in Lichess.

Blunder: 300 centipawns
Mistake: 100 centipawns
Inaccuracy: 50 centipawns

Source: https://github.com/ornicar/lila/blob/ma ... la#L55-L58


Lichess Q&A
https://en.lichess.org/qa/75/how-are-th ... inaccuracy

On Lichess, values was probably adjusted for the value for the average level of their games : 1700 level in 10 minutes games.
I'd want to have analysis for a higher levels : 2000 level in 2 hours.

I am now implementing a different approach to annotation. It is not based on the difference of the score, but is based on the impact of the error of the move.

Code: Select all

1. Blunder [??] - a move that would result to a decisive advantage to the opponent
2. Mistake [?] - a move that would result to a moderate advantage to the opponent.
3. Dubious [?!] - a move that would result to a slight advantage to the opponent.

Table:

Code: Select all

Decisive advantage [+- or -+] score >= +3.0 pawns
Moderate advantage &#91;+/- or -/+&#93; score >= +1.5 and score < +3.0
Slight advantage &#91;+/= or =/+&#93; score >= +0.25 and score < +1.5

Sample annotations with summary of number of blunder, mistakes and dubious moves for each side at the end of the game as comments.

Missing checkmate in one to draw won't be even dubious?

[Ferdy]

No, if score is +6, than a move which lowers the score to +3 is not a blunder, it is an inaccuracy.

This is what I found in Lichess.

Blunder: 300 centipawns
Mistake: 100 centipawns
Inaccuracy: 50 centipawns

Source: https://github.com/ornicar/lila/blob/ma ... la#L55-L58


Lichess Q&A
https://en.lichess.org/qa/75/how-are-th ... inaccuracy

On Lichess, values was probably adjusted for the value for the average level of their games : 1700 level in 10 minutes games.
I'd want to have analysis for a higher levels : 2000 level in 2 hours.

I am now implementing a different approach to annotation. It is not based on the difference of the score, but is based on the impact of the error of the move.

Code: Select all

1. Blunder [??] - a move that would result to a decisive advantage to the opponent
2. Mistake [?] - a move that would result to a moderate advantage to the opponent.
3. Dubious [?!] - a move that would result to a slight advantage to the opponent.

Table:

Code: Select all

Decisive advantage [+- or -+] score >= +3.0 pawns
Moderate advantage &#91;+/- or -/+&#93; score >= +1.5 and score < +3.0
Slight advantage &#91;+/= or =/+&#93; score >= +0.25 and score < +1.5

Sample annotations with summary of number of blunder, mistakes and dubious moves for each side at the end of the game as comments.

Missing checkmate in one to draw won't be even dubious?

Good point. I will revise the system.

[corres]

An issue about the programs: They work on .PGN file which it contains English characters only.
Any help?
Robert

[gbtami]

An issue about the programs: They work on .PGN file which it contains English characters only.
Any help?
Robert

PGN spec allows ISO 8859/1 (Latin 1) characters only. See
http://www.saremba.de/chessgml/standard ... e.htm#c4.1