Author: Heinz van Kempen
Date: 13:11:07 10/16/05
Go up one level in this thread
>I did not know that you already started testing and already have results. > >Can I understand from your post that history threshold=50 is the only change in >the Uri setting?(note that it is the only change that I suggested) > >Result seems good so far and I hope that fruit can keep 50 elo advantage >relative to Fritz9. > >I read in another forum the following words(original words are in hebrew and >I did translation to english in this post): > >"In my opinion for most people there is no justification to spend 35$ for an >engine with no interface even if it is 30 elo stronger than Fritz9. > >It is better to buy Fritz9 for the same money because it comes with a nice >interface and a lot of possibilities to analyze" > > >Maybe if Fabien makes it more than 50 elo better then it will convince more >people to buy Fruit(remember that we can hope that the update with tablebases is >going to be stronger than Fruit Uri setting). > >Uri Hi Uri, I started last night and the other CEGT testers this afternoon. The only change is the history threshold to 50. King Safety is unchanged in 100 and nothing else is modified. We also test with Fruit WCCC'05. This is identical to Fruit 2.2 in playing strength but is not prone to the parameters changing bug. Some also test in Arena and Shredder Classic GUI. Please note that EloStat overates engine versions with few games. Christian calculated the rating with bayeselo and "translated" the value to 2812 after 51 games instead of 2823. Ratinglist will currently be updated frequently. Check again on Wednesday evening for a more significant amount of games. Best Regards Heinz http://www.husvankempen.de/nunn/
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.