Skip to content
  • 3.0.5
  • 0f3db21
  • Compare
    Choose a tag to compare
    Search for a tag
  • 3.0.5
  • 0f3db21
  • Compare
    Choose a tag to compare
    Search for a tag

@vshcherbyna vshcherbyna released this Apr 28, 2021

This version of Igel won League 1 tournament in TCEC Season 21 - https://en.wikipedia.org/wiki/TCEC_Season_21#League_1

What's new:

  • Train a new network using Igel 2.6.0 (HCE), Igel 2.9.0 (NNUE) and Igel 3.0.0 (NNUE) data: ign-1-d593efbd
  • Better time management in sd time control
  • Skip tt cutoff for null moves in search
  • Skip null move prunning on tt hit
  • Vote for best thread in SMP mode
  • Less aggressive time use in non repeated tc

Regression run against Igel 3.0.0:

Long Time Control No Increment

ELO   | 51.95 +- 2.85 (95%)
SPRT  | 60.0+0.0s Threads=1 Hash=64MB
LLR   | 0.00 (-2.94, 2.94) [0.00, 0.00]
Games | N: 20548 W: 5296 L: 2246 D: 13006

Long Time Control With Increment

ELO   | 40.20 +- 2.74 (95%)
SPRT  | 60.0+0.6s Threads=1 Hash=64MB
LLR   | 0.00 (-2.94, 2.94) [0.00, 0.00]
Games | N: 20408 W: 4600 L: 2249 D: 13559
1 person reacted
Assets 7
  • 3.0.0
  • 1f2116f
  • Compare
    Choose a tag to compare
    Search for a tag
  • 3.0.0
  • 1f2116f
  • Compare
    Choose a tag to compare
    Search for a tag

@vshcherbyna vshcherbyna released this Apr 4, 2021

What's new:

  • Train a new network using Igel 2.6.0 (HCE) and Igel 2.9.0 (NNUE) data: ign-1-139b702b
  • Remove 'EvalFile' option - as of Igel 3.0.0 the network file is always embedded in the executable
  • Fix in check extensions
  • Use incbin on Linux to inject network file into the binary
  • Tune null move prunning conditions
  • Tune probcut prunning conditions
  • Tune quiets prunning
  • Fix bogus condition for history calculation
  • Implement more aggressive null move prunning
  • Stop extension of captures
  • Prevent explosion of history extensions
  • Implement better time management for sudden death time control
  • Count not played quiets for LMP prunning
  • Remove dummy 100 multiplier for history ordering calculation
  • Fix compilation issues with gcc 10 and Android NDK's Clang
  • Remove unused counter move table
  • Fix out of boundary access for history array
  • Fix issue with singular extensions - reported by ChizhovVadim (author of CounterGo chess engine)
  • Fix crash when 'ucinewgame' command is not issued - reported by Ipmanchess
  • Fix crash when thread position is not initialized - reported by Ipmanchess

Regression run against Igel 2.9.0:

Short Time Control No Increment

ELO   | 115.05 +- 3.88 (95%)
SPRT  | 10.0+0.0s Threads=1 Hash=8MB
LLR   | 0.00 (-2.94, 2.94) [0.00, 0.00]
Games | N: 20652 W: 9913 L: 3314 D: 7425

Short Time Control With Increment

ELO   | 75.88 +- 3.53 (95%)
SPRT  | 10.0+0.1s Threads=1 Hash=8MB
LLR   | 0.00 (-2.94, 2.94) [0.00, 0.00]
Games | N: 20904 W: 8061 L: 3567 D: 9276

Long Time Control No Increment

ELO   | 83.16 +- 4.04 (95%)
SPRT  | 60.0+0.0s Threads=1 Hash=64MB
LLR   | 0.00 (-2.94, 2.94) [0.00, 0.00]
Games | N: 13508 W: 4814 L: 1641 D: 7053

Long Time Control With Increment

ELO   | 58.22 +- 3.81 (95%)
SPRT  | 60.0+0.6s Threads=1 Hash=64MB
LLR   | 0.00 (-2.94, 2.94) [0.00, 0.00]
Games | N: 13540 W: 4023 L: 1775 D: 7742
Assets 7
  • 2.9.0
  • 299c0f5
  • Compare
    Choose a tag to compare
    Search for a tag
  • 2.9.0
  • 299c0f5
  • Compare
    Choose a tag to compare
    Search for a tag

@vshcherbyna vshcherbyna released this Dec 25, 2020

What's new:

  • Switch to own network based on Igel search and evaluation (ign-0-9b1937cc)
  • NNUE dynamic eval scaling

Regression run against Igel 2.8.0 at 60s+0.6:

ELO   | 23.78 +- 3.56 (95%)
SPRT  | 60.0+0.6s Threads=1 Hash=64MB
LLR   | 0.00 (-2.94, 2.94) [0.00, 0.00]
Games | N: 16112 W: 4102 L: 3001 D: 9009

Regression run against top engines at tc=all/20s+0.2:

Rank Name                          Elo     +/-   Games   Score   Draws
   0 Igel 2.9.0 64 BMI2 AVX2       -84       7    6000   38.2%   40.8%
   1 Stockfish 12 64-bit           387      25    1000   90.3%   18.3%
   2 Komodo 14 64-bit              122      16    1000   66.9%   46.4%
   3 Ethereal 12.75 64-bit          83      16    1000   61.8%   46.9%
   4 Fire 7.1 64-bit                21      16    1000   53.0%   46.3%
   5 Xiphos 0.6 64-bit               5      16    1000   50.7%   45.0%
   6 Komodo 14 MCTS 64-bit         -13      16    1000   48.2%   41.6%
Assets 6
  • 2.8.0
  • d419f97
  • Compare
    Choose a tag to compare
    Search for a tag
  • 2.8.0
  • d419f97
  • Compare
    Choose a tag to compare
    Search for a tag

@vshcherbyna vshcherbyna released this Sep 23, 2020

What's new:

  • Fix null move pruning bug (54 elo)
  • Use 250cp as lazy eval threshold
  • Improve time management for sudden death time controls
  • Use Dietrich Kappe's Night Nurse improved net (internal version ninu-0.2.5b.nnue)

Regression run against Igel 2.7.0 at tc=10+0.1:

Score of Igel 2.8.0 64 BMI2 AVX2 vs Igel 2.7.0 64 BMI2 AVX2: 1012 - 411 - 1380  [0.607] 2803
Elo difference: 75.67 +/- 9.15

Regression run against top engines in tc=all/60+0.6:

Rank Name                          Elo     +/-   Games   Score   Draws
   0 Igel 2.8.0 64 BMI2 AVX2        53       6    8000   57.6%   40.7%
   1 Stockfish 11 64-bit           258      26     500   81.5%   33.0%
   2 Komodo 14 64-bit              110      21     500   65.3%   51.8%
   3 Ethereal 12.25 64-bit          36      22     500   55.2%   48.4%
   4 Komodo 14 MCTS 64-bit          15      22     500   52.2%   48.4%
   5 Xiphos 0.6 64-bit               1      22     500   50.1%   48.6%
   6 Fire 7.1 64-bit               -14      22     500   48.0%   47.6%
   7 RofChade 2.202 64-bit         -63      22     500   41.1%   48.2%
   8 Booot 6.4 64-bit              -76      23     500   39.3%   41.8%
   9 Defenchess 2.2 64-bit         -87      24     500   37.7%   39.8%
  10 Laser 1.7 64-bit              -94      24     500   36.8%   41.2%
  11 Fizbo 2 64-bit               -118      24     500   33.7%   41.4%
  12 RubiChess 1.7.3 64-bit       -122      23     500   33.1%   42.6%
  13 Andscacs 0.95 64-bit         -130      25     500   32.1%   35.8%
  14 Gull 3 64-bit                -151      26     500   29.6%   32.8%
  15 Pedone 2.0 64-bit            -186      27     500   25.5%   30.6%
  16 Strelka 5.5 64-bit           -266      33     500   17.8%   18.8%

Regression run against top engines in tc=all/20+0.2:

Rank Name                          Elo     +/-   Games   Score   Draws
   0 Igel 2.8.0 64 BMI2 AVX2      -179       7    6000   26.3%   38.1%
   1 Stockfish 12 64-bit           458      30    1000   93.3%   13.0%
   2 Komodo 14 64-bit              208      17    1000   76.8%   39.7%
   3 Ethereal 12.75 64-bit         173      16    1000   73.0%   42.7%
   4 Fire 7.1 64-bit               130      17    1000   67.8%   41.1%
   5 Xiphos 0.6 64-bit             121      16    1000   66.7%   47.6%
   6 Komodo 14 MCTS 64-bit         103      16    1000   64.3%   44.3%
Assets 5
  • 2.7.0
  • 05f8eaa
  • Compare
    Choose a tag to compare
    Search for a tag
  • 2.7.0
  • 05f8eaa
  • Compare
    Choose a tag to compare
    Search for a tag

@vshcherbyna vshcherbyna released this Aug 19, 2020

What's new:

  • Switch to NNUE as a main evaluation function
  • Use Dietrich Kappe's Night Nurse NNUE net as a default net for Igel 2.7.0 release

Regression run against Igel 2.6.0 at tc=10+0.1:

Score of Igel 2.7.0 64 BMI2 AVX2 vs Igel 2.6.0 64 BMI2: 981 - 292 - 562  [0.688] 1835
Elo difference: 137.16 +/- 13.75

Regression run against top engines in tc=all/60+0.6:

Rank Name                          Elo     +/-   Games   Score   Draws
   0 Igel 2.7.0 64 BMI2 AVX2         3       6    8000   50.4%   42.8%
   1 Stockfish 11 64-bit           287      28     500   83.9%   30.2%
   2 Komodo 14 64-bit              157      23     500   71.2%   44.8%
   3 Ethereal 12.25 64-bit          80      21     500   61.3%   51.8%
   4 Fire 7.1 64-bit                59      22     500   58.4%   46.4%
   5 Komodo 14 MCTS 64-bit          58      22     500   58.3%   49.8%
   6 Xiphos 0.6 64-bit              41      22     500   55.9%   49.8%
   7 RofChade 2.202 64-bit          -3      22     500   49.6%   45.6%
   8 Laser 1.7 64-bit              -24      22     500   46.5%   47.4%
   9 Defenchess 2.2 64-bit         -36      23     500   44.8%   42.0%
  10 Booot 6.4 64-bit              -50      23     500   42.8%   44.0%
  11 RubiChess 1.7.3 64-bit        -55      23     500   42.2%   43.2%
  12 Fizbo 2 64-bit                -64      22     500   40.9%   45.8%
  13 Andscacs 0.95 64-bit          -88      24     500   37.6%   38.4%
  14 Gull 3 64-bit                 -97      25     500   36.4%   34.8%
  15 Pedone 2.0 64-bit            -115      24     500   34.0%   38.8%
  16 Strelka 5.5 64-bit           -153      26     500   29.3%   31.4%
Assets 5

@vshcherbyna vshcherbyna released this Aug 1, 2020

What's new:

  • Implement piece threats evaluation
  • Simplify connected pass pawn evaluation
  • Implement backward pawn evaluation
  • Apply tempo after interpolation and speed up null move evaluation
  • Implement verification qsearch for probcut
  • Remove L1 normalisation from tuner and retune all params
  • Implement trapped rook evaluation
  • Implement razoring in search
  • Implement king danger evaluation
  • Implement hanging piece evaluation
  • Implement weak pawn evaluation
  • Implement restricted piece evaluation
  • Implement lazy evaluation with a threshould of 300cp
  • Implement multicut search
  • Implement threat by a protected pawn evaluation
  • Improve safe check evaluations
  • Bench 7538060

Regression run against Igel 2.5.0 at tc=60+0.6:

ELO   | 55.64 +- 4.19 (95%)
SPRT  | 60.0+0.6s Threads=1 Hash=64MB
LLR   | 0.00 (-2.94, 2.94) [0.00, 0.00]
Games | N: 12028 W: 3703 L: 1793 D: 6532

Regression run against top engines in tc=all/60+0.6:

Rank Name                          Elo     +/-   Games   Score   Draws
   0 Igel 2.6.0 64 BMI2           -152       6    8000   29.5%   36.6%
   1 Stockfish 11 64-bit           455      42     500   93.2%   13.2%
   2 Komodo 14 64-bit              370      35     500   89.4%   19.6%
   3 Komodo 14 MCTS 64-bit         265      28     500   82.1%   29.8%
   4 Ethereal 12.25 64-bit         238      27     500   79.7%   33.0%
   5 Fire 7.1 64-bit               229      26     500   78.9%   34.6%
   6 Xiphos 0.6 64-bit             224      25     500   78.4%   36.0%
   7 Laser 1.7 64-bit              165      24     500   72.1%   41.4%
   8 RofChade 2.202 64-bit         141      22     500   69.3%   46.2%
   9 Booot 6.4 64-bit              141      25     500   69.2%   37.6%
  10 Defenchess 2.2 64-bit         119      22     500   66.5%   46.2%
  11 Fizbo 2 64-bit                102      25     500   64.3%   33.0%
  12 RubiChess 1.7.3 64-bit         94      22     500   63.2%   49.2%
  13 Andscacs 0.95 64-bit           84      24     500   61.9%   40.6%
  14 Pedone 2.0 64-bit              50      23     500   57.2%   44.0%
  15 Gull 3 64-bit                  37      23     500   55.3%   43.8%
  16 Strelka 5.5 64-bit            -16      24     500   47.7%   37.0%

Regression run against top engines in tc=all/10+0.1:

Rank Name                          Elo     +/-   Games   Score   Draws
   0 Igel 2.6.0 64 BMI2           -184       5   16000   25.7%   28.2%
   1 Stockfish 11 64-bit           501      37    1000   94.7%    8.6%
   2 Komodo 14 64-bit              396      27    1000   90.7%   15.6%
   3 Ethereal 12.25 64-bit         329      23    1000   86.9%   21.8%
   4 Xiphos 0.6 64-bit             265      21    1000   82.1%   27.4%
   5 Fire 7.1 64-bit               263      22    1000   82.0%   22.8%
   6 Komodo 14 MCTS 64-bit         226      21    1000   78.6%   24.4%
   7 RofChade 2.202 64-bit         187      18    1000   74.6%   33.5%
   8 Laser 1.7 64-bit              184      18    1000   74.2%   34.6%
   9 Defenchess 2.2 64-bit         178      19    1000   73.6%   32.3%
  10 Fizbo 2 64-bit                155      20    1000   71.0%   24.9%
  11 RubiChess 1.7.3 64-bit        142      18    1000   69.4%   35.2%
  12 Andscacs 0.95 64-bit           97      18    1000   63.6%   33.5%
  13 Strelka 5.5 64-bit             94      18    1000   63.2%   32.5%
  14 Gull 3 64-bit                  89      18    1000   62.6%   35.2%
  15 Booot 6.4 64-bit               82      17    1000   61.6%   35.9%
  16 Pedone 2.0 64-bit              68      18    1000   59.7%   32.3%
Assets 3
  • 2.5.0
  • d57fa46
  • Compare
    Choose a tag to compare
    Search for a tag
  • 2.5.0
  • d57fa46
  • Compare
    Choose a tag to compare
    Search for a tag

@vshcherbyna vshcherbyna released this Jun 15, 2020

What's new:

  • Implement OpenBench support
  • Implement BMI2 instruction support for faster search
  • Implement prefetch instruction support for faster search
  • Implement ponder support
  • Implement improved sudden death time control management
  • Implement use of hugepages in Linux OS
  • Implement more efficient tt management and re-write tt replacement scheme to 'always replace'
  • Implement proper handling of chopper moves
  • Implement improved king safety evaluation
  • Implement improved weights management
  • Implement symmetric evaluation function
  • Set tt score in qsearch
  • Set bestmove as a hashmove by default
  • Bugfix for knight outpost
  • Bugfix for depth skipping in smp mode
  • Implement pst based evaluation for connected passers
  • Bugfix for probcut in check
  • Factor in 'improving' even when in checks
  • Adjust bestscore from tt score
  • Bugfix shuffling of moves at root in smp mode

Regression run against top 50 engines in tc=all/60+0.6:

Rank Name                          Elo     +/-   Games   Score   Draws
   0 Igel 2.5.0 64 BMI2            -65       6   10000   40.8%   35.0%
   1 Stockfish 10 64 POPCNT        478      46     500   94.0%   11.2%
   2 Ethereal 11.75 (PEXT)         293      30     500   84.4%   26.4%
   3 Fire 7.1 x64 popcnt           260      28     500   81.7%   29.8%
   4 Xiphos 0.6 BMI2               256      27     500   81.4%   31.6%
   5 rofChade 2.202 BMI            181      23     500   73.9%   41.8%
   6 Defenchess 2.2 x64            180      26     500   73.8%   34.4%
   7 Laser 1.7                     174      25     500   73.1%   35.8%
   8 Andscacs 0.95                 138      24     500   68.9%   39.0%
   9 Pedone 2.0                     89      23     500   62.5%   41.8%
  10 RubiChess 1.6                  84      22     500   61.9%   47.4%
  11 Strelka 5.5 x64                73      24     500   60.3%   38.6%
  12 Arasan 22.0                    58      23     500   58.3%   43.0%
  13 Nemorino                       18      24     500   52.6%   38.4%
  14 Deep iCE 4.0.853 x64/popcnt    16      23     500   52.3%   44.2%
  15 Texel 1.07                     -1      23     500   49.9%   44.2%
  16 Vajolet2 2.8.0                 -3      24     500   49.6%   40.4%
  17 Protector 1.9.0               -29      23     500   45.8%   41.2%
  18 Winter 0.7 BMI2              -151      27     500   29.6%   30.4%
  19 zurichess neuchatel          -203      27     500   23.7%   32.2%
  20 GreKo 2018.08                -478      53     500    6.0%    8.0%

Regression run against top engines in tc=all/10+0.1:

Rank Name                          Elo     +/-   Games   Score   Draws
   0 Igel 2.5.0 64 BMI2           -187       5   18000   25.4%   26.7%
   1 Stockfish 11 64 BMI2          554      40    1000   96.0%    7.3%
   2 Komodo 14 64-bit              449      32    1000   93.0%   11.0%
   3 Ethereal 12.00 (PEXT)         399      29    1000   90.8%   13.7%
   4 Xiphos 0.6 BMI2               339      24    1000   87.5%   20.9%
   5 Fire 7.1 x64 popcnt           329      24    1000   86.9%   19.2%
   6 Laser 1.7                     253      21    1000   81.1%   25.6%
   7 Defenchess 2.2 x64            241      20    1000   80.0%   28.8%
   8 rofChade 2.202 BMI            223      20    1000   78.3%   29.4%
   9 RubiChess 1.7.2 Popcnt        182      19    1000   74.0%   28.8%
  10 Andscacs 0.95                 156      19    1000   71.1%   31.0%
  11 Pedone 2.0                    147      19    1000   70.0%   31.2%
  12 Strelka 5.5 x64               147      19    1000   70.0%   30.2%
  13 Booot 6.4_x64_popcnt          136      18    1000   68.7%   33.1%
  14 Deep iCE 4.0.853 x64/popcnt   111      18    1000   65.5%   32.1%
  15 Texel 1.07                     78      17    1000   61.0%   35.4%
  16 Arasan 22.0                    76      18    1000   60.8%   33.3%
  17 Nemorino                       50      18    1000   57.1%   32.7%
  18 Vajolet2 2.8.0                  1      17    1000   50.2%   36.0%

Regression run against top engines in tc=all/10+0.1:

Rank Name                          Elo     +/-   Games   Score   Draws
   0 Igel 2.5.0 64 BMI2           -235       5   16000   20.5%   24.8%
   1 Stockfish 11 64-bit           550      40    1000   96.0%    7.5%
   2 Komodo 14 64-bit              449      31    1000   93.0%   12.2%
   3 Ethereal 12.25 64-bit         361      25    1000   88.8%   18.7%
   4 Fire 7.1 64-bit               318      24    1000   86.2%   19.8%
   5 Xiphos 0.6 64-bit             298      22    1000   84.8%   24.3%
   6 Laser 1.7 64-bit              266      21    1000   82.2%   25.6%
   7 Komodo 14 MCTS 64-bit         256      22    1000   81.3%   23.3%
   8 RofChade 2.202 64-bit         242      20    1000   80.1%   27.8%
   9 Defenchess 2.2 64-bit         211      20    1000   77.1%   27.3%
  10 RubiChess 1.7.3 64-bit        205      19    1000   76.4%   29.7%
  11 Fizbo 2 64-bit                184      20    1000   74.3%   24.1%
  12 Andscacs 0.95 64-bit          183      19    1000   74.2%   28.9%
  13 Strelka 5.5 64-bit            153      19    1000   70.8%   29.5%
  14 Gull 3 64-bit                 152      19    1000   70.6%   31.8%
  15 Booot 6.4 64-bit              137      18    1000   68.7%   31.4%
  16 Pedone 2.0 64-bit             125      18    1000   67.3%   34.7%
Assets 3
  • 2.4.0
  • 90bf916
  • Compare
    Choose a tag to compare
    Search for a tag
  • 2.4.0
  • 90bf916
  • Compare
    Choose a tag to compare
    Search for a tag

@vshcherbyna vshcherbyna released this Mar 22, 2020

What's new:

  • Implement singlular extensions
  • Detect drop of score and using time more aggressively
  • Remove IID and simplify search
  • Do not tt cutoff on pv moves
  • Simplification for nullmove pruning
  • Simplification for connected pawn evalution
  • Basic endgame knowledge to detect draw in eval/search
  • Retune everything + decrease number of tunnable parameters from 2600 to 1500
  • Apply LMR when at least one legal move has been played
  • Decrease reductions in LMR for pv moves

Regression run against top 50 engines in tc=all/60+0.6:

Rank Name                          Elo     +/-   Games   Score   Draws
   0 Igel 2.4.0 64 POPCNT          -80       6   10000   38.7%   32.1%
   1 Stockfish 10 64 POPCNT        501      48     500   94.7%   10.6%
   2 Ethereal 11.75 (PEXT)         324      31     500   86.6%   24.4%
   3 Fire 7.1 x64 popcnt           311      31     500   85.7%   23.8%
   4 Xiphos 0.6 BMI2               299      30     500   84.8%   26.4%
   5 rofChade 2.202 BMI            217      25     500   77.7%   35.8%
   6 Laser 1.7                     215      27     500   77.5%   33.0%
   7 Defenchess 2.2 x64            192      27     500   75.1%   32.6%
   8 Andscacs 0.95                 161      26     500   71.7%   33.0%
   9 Pedone 2.0                    109      24     500   65.2%   38.8%
  10 RubiChess 1.6                  81      23     500   61.5%   44.2%
  11 Arasan 22.0                    66      24     500   59.4%   39.6%
  12 Strelka 5.5 x64                62      24     500   58.8%   36.8%
  13 Deep iCE 4.0.853 x64/popcnt    48      24     500   56.8%   36.8%
  14 Texel 1.07                     41      22     500   55.9%   47.0%
  15 Vajolet2 2.8.0                 39      24     500   55.6%   37.6%
  16 Nemorino                       27      24     500   53.9%   37.4%
  17 Protector 1.9.0               -13      24     500   48.1%   36.2%
  18 Winter 0.7 BMI2              -136      27     500   31.4%   30.0%
  19 zurichess neuchatel          -234      27     500   20.6%   31.2%
  20 GreKo 2018.08                -487      54     500    5.7%    7.8%

Regression run against top engines in tc=all/10+0.1:

Rank Name                          Elo     +/-   Games   Score   Draws
   0 Igel 2.4.0 64 POPCNT         -219       5   18000   22.1%   23.7%
   1 Stockfish 11 64 BMI2          574      43    1000   96.5%    6.5%
   2 Komodo 14 64-bit              497      34    1000   94.6%   10.2%
   3 Ethereal 12.00 (PEXT)         407      28    1000   91.3%   14.5%
   4 Fire 7.1 x64 popcnt           341      24    1000   87.7%   19.2%
   5 Xiphos 0.6 BMI2               334      24    1000   87.3%   19.5%
   6 Laser 1.7                     307      23    1000   85.4%   20.4%
   7 rofChade 2.202 BMI            287      22    1000   83.9%   23.0%
   8 Defenchess 2.2 x64            274      23    1000   82.9%   21.4%
   9 RubiChess 1.7.2 Popcnt        221      20    1000   78.1%   29.2%
  10 Pedone 2.0                    196      20    1000   75.6%   25.8%
  11 Andscacs 0.95                 191      20    1000   75.0%   24.9%
  12 Strelka 5.5 x64               186      20    1000   74.5%   24.3%
  13 Booot 6.4_x64_popcnt          160      19    1000   71.5%   28.4%
  14 Deep iCE 4.0.853 x64/popcnt   134      18    1000   68.3%   31.3%
  15 Arasan 22.0                   109      18    1000   65.1%   31.5%
  16 Texel 1.07                    101      18    1000   64.2%   32.6%
  17 Nemorino                       95      18    1000   63.3%   30.1%
  18 Vajolet2 2.8.0                 50      18    1000   57.1%   34.4%
Assets 3
  • 2.3.1
  • 4d2a583
  • Compare
    Choose a tag to compare
    Search for a tag
  • 2.3.1
  • 4d2a583
  • Compare
    Choose a tag to compare
    Search for a tag

@vshcherbyna vshcherbyna released this Jan 14, 2020

What's new:

  • Re-write the root search routine
  • Re-write aspiration loop with a smaller aspiration window
  • Do not cutoff qsearch on pv moves in qsearch
  • Handle repetitions at the qsearch entance
  • More strict pruning conditions for quiet moves
  • Proper way to pass position to child threads in SMP mode to handle repetitions
  • Unified condition to killer moves as non-tactical moves
  • Perft tests for six different positions

Regression run against top 50 engines in tc=all/60+0.6:

Rank Name                          Elo     +/-   Games   Score   Draws
   0 Igel 2.3.1 64 POPCNT         -106       6   10000   35.2%   31.0%
   1 Stockfish 10 64 POPCNT        527      55     500   95.4%    8.0%
   2 Fire 7.1 x64 popcnt           346      35     500   88.0%   19.6%
   3 Ethereal 11.75 (PEXT)         340      31     500   87.6%   23.6%
   4 Xiphos 0.6 BMI2               318      31     500   86.2%   24.4%
   5 Laser 1.7                     259      28     500   81.6%   29.6%
   6 rofChade 2.202 BMI            220      27     500   78.0%   32.8%
   7 Defenchess 2.2 x64            206      26     500   76.6%   34.0%
   8 Andscacs 0.95                 172      27     500   72.9%   30.2%
   9 RubiChess 1.6                 161      24     500   71.7%   41.0%
  10 Pedone 2.0                    143      25     500   69.5%   34.6%
  11 Strelka 5.5 x64               119      25     500   66.5%   34.6%
  12 Deep iCE 4.0.853 x64/popcnt    98      24     500   63.7%   39.0%
  13 Arasan 22.0                    95      25     500   63.4%   35.6%
  14 Texel 1.07                     89      24     500   62.6%   38.8%
  15 Vajolet2 2.8.0                 65      23     500   59.2%   44.4%
  16 Nemorino                       44      25     500   56.3%   34.2%
  17 Protector 1.9.0                 2      23     500   50.3%   40.6%
  18 Winter 0.7 BMI2              -111      26     500   34.5%   29.8%
  19 zurichess neuchatel          -215      26     500   22.5%   33.8%
  20 GreKo 2018.08                -411      44     500    8.6%   11.2%
Assets 3
  • 2.3.0
  • 3f6bed0
  • Compare
    Choose a tag to compare
    Search for a tag
  • 2.3.0
  • 3f6bed0
  • Compare
    Choose a tag to compare
    Search for a tag

@vshcherbyna vshcherbyna released this Jan 1, 2020

What's new:

  • Fix issue when reductions in LMR could lead to directly qsearch
  • Rename uci option 'Level' into 'Skill Level'
  • Do not apply razoring when in check
  • Do not apply 'improving' factor when in check
  • Do not prune quiets at root
  • Fetch history only when a quiet move is detected
  • Improve branching factor and apply LMR more aggressively
  • Fix compilation issue with clang/gcc9 compiler

Relative strength compared to previous versions of Igel in CCRL 40/4 time control:

os=linux
hash=256
tc=40/102 (CCRL 40/4)
Score of Igel 2.3.0 64 POPCNT vs Igel 2.2.2 64 POPCNT: 331 - 195 - 805  [0.551] 1331
Elo difference: 35.62 +/- 11.69

Regression run against top 50 engines in tc=all/60+0.6:

Rank Name                          Elo     +/-   Games   Score   Draws
   0 Igel 2.3.0 64 POPCNT         -178      14    2000   26.5%   24.9%
   1 Stockfish 10 64 POPCNT        531     132     100   95.5%    9.0%
   2 Xiphos 0.6 BMI2               494     115     100   94.5%   11.0%
   3 Ethereal 11.75 (PEXT)         478     121     100   94.0%   10.0%
   4 Fire 7.1 x64 popcnt           436     103     100   92.5%   13.0%
   5 Laser 1.7                     330      68     100   87.0%   26.0%
   6 rofChade 2.202 BMI            323      74     100   86.5%   23.0%
   7 Defenchess 2.2 x64            295      73     100   84.5%   23.0%
   8 Andscacs 0.95                 263      72     100   82.0%   22.0%
   9 Pedone 2.0                    230      65     100   79.0%   28.0%
  10 RubiChess 1.6                 230      61     100   79.0%   32.0%
  11 Strelka 5.5 x64               220      71     100   78.0%   20.0%
  12 Arasan 22.0                   186      61     100   74.5%   31.0%
  13 Texel 1.07                    168      63     100   72.5%   27.0%
  14 Deep iCE 4.0.853 x64/popcnt   143      54     100   69.5%   41.0%
  15 Nemorino                      143      61     100   69.5%   29.0%
  16 Vajolet2 2.8.0                119      58     100   66.5%   33.0%
  17 Protector 1.9.0               100      56     100   64.0%   36.0%
  18 Winter 0.7 BMI2               -31      58     100   45.5%   29.0%
  19 zurichess neuchatel           -63      53     100   41.0%   40.0%
  20 GreKo 2018.08                -295      83     100   15.5%   15.0%
Assets 3