General remarks: This rating list is made for my private enjoyment! I am happy to provide it and hope you like it too. In case you don't like it, I am sorry and hope you find something that fits your needs. Nonetheless I will listen to any good argument about what could be improved or might be wrong.
Playing conditions: Ponder : ON CPU : Phenom2 + Piledriver Operating system : Windows 7 Cores : 1 core each engine Hash : 256 MB Bases : 4 pieces Cache for Tbs : 64 MB Time control : 5 min/game + 3 sec/move (mean game length ~ 16 Minutes) Openings : 110 different opening positions (50 pos. before 2012.01.28, 75 pos before 2014.01.01) Positions : The Open/Half Open/Closed opening distribution of the 110 openings is 21%/30%/49% Changing colours = 220 games engine vs engine Learning : OFF Ordo 1.0 : ordo-win64.exe -Q -N 0 -D -n 4 -J -j Head2head.dat -a 2993 -A "Critter 1.6a" -W -p games.pgn -U "0,1,2,3,4,5,6,7,8,9,10,11" -s1000 -V -o ORDO_FULL.dat All following rules are subject of change. They show the intention to make a proper clean list without to much entries of doubtful origin. However there are points were it is impossible to ignore facts. I try to keep the IPON as clean as reasonably possible! 1. In general, only the best program of one kind/programmer is being tested. This means if there is a successor for an engine the weaker version is no longer tested. If two version are VERY close some additional games might be played. 2. Engines families which are to closely related will be represented by their best member. To define what is to close a widely accepted method called "common sense" is used. 3. No anonymous engines or engines with maliciously 'false' programmer names are accepted! 4. All new entries will play 3520 games vs 16 best single engines. New releases of the 16 best will play 3300 vs the 'other' 15 best single engines. 5. Engines with a high failure rate will be eliminated by new entries prior to the last entry. 6. The games are played with the current Shredder Classic GUI. 7. Engines from the same author will only be added in reasonable time intervals (aprox. 3 months) as this list should not be abused for beta tests. 8. The range of tested engines is the top 16 from top down. To verify if a new engine is in that range a short internal test will be made. 9. Engines are tested in 64 bit and SSE4.2 (POPCNT) if available. 32 bit Engines are marked with '32b'. 10. Reference point for the lists is 2993 for Critter 1.6a. All other engines are derived from that number. 11. A maximum of 1% of time losses is allowed for a new entry. 12. No download of games. 13. Each test run with 3300 games uses about 35kWh of energy. If you are willing to pay for your favorite engine PLEASE send me an email! We will find an arangement :-)
© 2017 by Ingo Bauer