2000 Game Test Run

MikeB · Post by **MikeB** » Tue Jun 13, 2017 8:31 am

6/5/2017 Dev Stockfish vs Current Dev McBrain (based on 6/5/17 SF)

The focus is on "analysis" with the current version, not ELO ( ELO has never really been the focus anyway)

Just started , should run about 20 hours. Game is 100 sec/ with 1 sec increment - very long relative to SF Fishtest standards. I have no idea how this will turn out - just hoping McB can stay close. Anything within 20 ELO will be good here as the analysis would be much more in line with what many users desire.

Code: Select all

80 of 2000 games completed...
time control&#58; 10000+100
Date&#58; 06/13/17 &#58; 02&#58;21&#58;30

80 game&#40;s&#41; loaded
Rank Name                      Rating   &#916;     +    -     #     &#931;    &#931;%     W    L    D   W%    =%   OppR 
---------------------------------------------------------------------------------------------------------
   1 McBrain 2.5d 64 POPCNT     3100   0.0   15   15    80   40.0  50.0    8    8   64  10.0  80.0  3100 
   2 Stockfish 050617 64 POPC   3100   0.8   15   15    80   40.0  50.0    8    8   64  10.0  80.0  3100 
---------------------------------------------------------------------------------------------------------
80 game&#40;s&#41; loaded

Realtime Updates - about every 5 minutes or when 10 games are completed
results: https://www.dropbox.com/s/lfzfh7zp8famr ... a.txt?dl=1

pgn files: https://www.dropbox.com/s/o6k11fk6gvwfv2b/all.pgn?dl=1

Note: The opening book may look a little wacky, but it is exactly the same 40,000+ positions that Fishtest uses in their testing. Opening lines are chosen at random and they may occasionally be repeated. The draw rate with these opening position may be a little less than what a normal book may show - I believe that was one of the the intents with these positions.