Old timer makes comeback
Entrepreneur Chris Whittington author of Chess System Tal who retired from the commerce
in 2000 rediscovered chess programming a few years back with the focus on the new and
ruling NNUE technology.
Chris mailed me his current version with a 54Mb [!] neural net. I decided to run
Chess-System-Tal-NNUE-1.26 for the Gambit Rating List and it moved right into the top-5 with an incredible elo of 3507. Welcome back Chris into the Arena.
The 6502 running at 100 Mhz ?
It does, my old MM5, Polgar, Milano etc. are put in a $50 FPGA card and currently tested again in the German nostalgic forum (Google translation is your friend) where computer enthusiasts are reliving the 80's.
do you remember?
It's back! At zero cost. No expensive plugin card needed any longer. The MAME team created the drivers for the ChessMachine, they now have emulated the hardware miracle of the early 90's. Download. More info.
Can an engine write opening theory ?
Is it possible for nowadays strong engines that an engine can write its own opening book? We tried the almighty Stockfish 10. From 82,000 human games between 2600+ elo rated players we extracted from the first 20 moves 1.4 million unique positions and let Stockfish 10 analyze them at 1 second, 2 seconds and 5 seconds. From each we created an opening book.
Then we pitched the 3 books against the handcrafted ProDeo book created by expert Jeroen Noomen over a period of 15 years. None of the 3 Stockfish books was able to win one of the 3x5000 games matches, scores for Stockfish varied from 45-48%, the 45% notable from the 5 second book and so we gave up. Close, but no cigar.
The human still superior in this area?
Encouraged by the excellent results of Lc0 on OKE topping alll lists we tried Lc0 next. The same 1.4 million positions were analyzed at depth=1 (about 100-200ms) which is a factor 25-50 slower than Stockfish and surprise Lc0 was able to win the 5000 game match with a small margin (51.9%).
What does the future hold?
Or deeper analysis depths to begin with.
How strong is Stockfish 10 ?
What's the strength of Stockfish 10 when it is only given 0.1 second thinking time? The answer might surprise you. We played 3 handicap matches SF10 at 0.1 seconds vs known CCRL elo rated opponents at 40/4 level (thus a time handicap factor of 60!) emulating the CCRL 40/4 rating list, table left. We also measured the elo rating when SF10 was given one second vs the 6 second at 40/4 level, table right.
Time handicap 100ms vs 60.000ms (factor 60)
81 - 19
70 - 30
65 - 35
53 - 47
Time handicap 1s vs 6s (factor 6)
74 - 26
62 - 38
53 - 47
55 - 45
We can safely conclude that:
1. SF10 at 0.1 second already plays at 3000 elo.
2. SF10 at 1.0 second already plays at 3250 elo.