The Internet never forgets. I have managed to track down 18 games from Rybka 1.5.32 (March/April 2004?), so one can compare it to Rybka 1.0 Beta in some vague sense. About all I can say from the data is that, given that Rybka 1.5.32 hits a higher "depth" in the middlegame against most opponents, perhaps this was not obfuscated at that point.
Similarly, the D, E, and F sections in Chess War V contain 11 games apiece from Rybka 1.5.32, to bring the total of games to 51. For reasons of length, I did not include these in the forum board-viewer with this post.
Note that Rybka 1.5.32 played two rook underpromotions, though by the time Rybka 1.0 Beta was released 20 months later, these had disappeared.
There also exist 9 games from January 31, 2004 from CCT-6, where no version number for Rybka is given, and I include those too. Given that Rybka 1.0 Beta of December 2005 still could not ponder, it would seem reasonable to expect that the Rybka version in this event could not either, which may have impaired performance.
2138 Partien von 2520 gespielt
Name des Turniers: Arena Turnier
Ort/ Land: ORION8-LATITUDE, Deutschland
Spielstufe: Turnier 40/20
Hardware: Intel(R) Core(TM)2 CPU T7600 @ 2.33GHz mit 2.048 MB Speicher
Betriebssystem: Microsoft Windows Vista Professional (Build 6000)
PGN-Datei: C:\schach\arena201\Arena.pgn
Internetseite:
E-Mail Adresse:
There are also 10 more games (+7-2=1) for Rybka 1.5.32 from the "intro" tournament to the third "System of Swiss" season. These appear in the last 10 (numbers 6-15) of the SdS3 files at http://americanfoot.free.fr/echecs/suisse/fichiers.htm
I will collate and post.
It can be noted that this Rybka version played a bishop underpromotion.
There was a knight underpromotion against JanWillem 1.09 in Round 2.2 of E32, and rook underpromotions against: Tinker 4.6.9 in Round 1.4 of C67, Nesik 0.6.4 in Round 2.7 of E32, Matacz 0.998 in Round 1.2 of A48, and Tinker 4.6.9 in Round 2.4 of ECO A87.
I still have all these pre-Beta 1.0 Rybka versions here, but Vasik Rajlich reqeusted me to keep them private.
Olivier
Would it be correct to assume this request is not of recent? In other words, the request happened years ago.
Those Rybka versions were meant to be private from the start. I have some private engines in my tournaments, and programmers know I stick to my promise not to let them leak.
In February 2010 Zach Wegner wanted to see them. I asked Vas for permission to send them to Zach, and he said no.
I still have all these pre-Beta 1.0 Rybka versions here, but Vasik Rajlich reqeusted me to keep them private.
Olivier
Would it be correct to assume this request is not of recent? In other words, the request happened years ago.
Those Rybka versions were meant to be private from the start. I have some private engines in my tournaments, and programmers know I stick to my promise not to let them leak.
In February 2010 Zach Wegner wanted to see them. I asked Vas for permission to send them to Zach, and he said no.
There was an off-line discussion about this, and the general agreement was that tournament directors like Olivier fully deserve to have their privacy concerns respected -- indeed, such privacy is one reason why they get so many engines submitted. In particular, Fabien was quite clear that he had no desire whatsoever to press this matter, and that any decision to allow early Rybkas to be inspected (by the ICGA forum or whomever) should be totally up to Rajlich. In a similar line, how many would have their source code compiled by Jim Ablett if they worried he would "cooperate" with a roguish ICGA request?
