Some Notes: -I receive oftenly mails/interests by different users,about testing their Houdini settings in SCCT *And due to i have no much free CPU time, SCCT has already a new rule: -Since today (28.07.2012) serious data-minimum 1.000 games is required *I need a such confirmation that those new settings should be least 20 Elo stronger than Houdini 2.0c default *Note:Game results,played with ultra-fast time controls will be not accepted as serious data
Attention: -For clarification minimum 1.000 games is needed -The modified Houdini's Elo performance should at least 20 Elo stronger than Houdini 2.0c default -The Houdini testings should be played at Blitz or Slow time controls (min 3min or at least 2min+2sec is required) -The chess engines should be tested with strong openings,maximum up to 10-12 moves *Note that this new rule is for people, who has interest to participate with Houdini settings in SCCT
One thing more, The opponents (in your testings) should not be based only on Houdini versions I mean you need to test your Houdini against various Top engines (e.g Critter,Stockfish,Rybka,Komodo,Ivanhoe...) Note also that 1.000 games per player only Houdini vs Houdini can lead to misunderstanding results
And Last, Without such serious data (games) and confirmation,SCCT will not accept/start testing any new Houdini settings...
Forum
CSS-Shop