Author: Mark Young
Date: 03:05:18 11/07/98
Go up one level in this thread
On November 07, 1998 at 05:30:46, Enrique Irazoqui wrote: >On November 06, 1998 at 23:56:57, Mark Young wrote: > >>On November 06, 1998 at 20:22:44, Enrique Irazoqui wrote: >> >>>Latest results: >>> >>>PII-400 256MB, 40/40 >>> >>> H7b J5 R10 F5 TOTAL % Played >>>Hiarcs 7b --- 4 7 8 19 63.3 30 >>>Junior 5 6 --- 5 4.5 15.5 51.7 30 >>>Rebel 10 3 5 --- 5.5 13.5 45.0 30 >>>Fritz 5 2 5.5 4.5 --- 12 40.0 30 >>> >>>Early next week Mchess 8 will start playing. I need a few days to do other kinds >>>of testing on the beta of Hiarcs 7. >>> >>>Quick notes about Hiarcs 7b (it's 2 am and I feel lazy): >>> >>>- Hiarcs 7 defeated Rebel 10 in the endings, the weaker side of Rebel and one of >>>the strongest of Hiarcs. The middlegame was even. >> >>I wonder just how weak Rebel 10 is in endgames, I heard the same thing about >>Fritz 5, but its endgames I have found are very strong. What positions and types >>of endgames do you base this on. Can you show us how Rebel 10 is weak in the end >>games over other programs. > >If you reread my comment, you wil notice I said that endings is the "weaker" >side of Rebel I know thats what you said, And I'm saying how many games did it take for you to lable Rebel 10 as "weaker" then Hiarcs 7 and Genius 5 in the endgame. I just want to know how much data you based this jugdement on. I am not saying your wrong.. I don't know. I just want to know how sure your judgement is. , not "weak". Hiarcs and Genius are the only two programs I know >that have stronger endings than Rebel. In my opinion, Rebel excels in the >middlegame, not in the endgame. > >> I don't have enough games to come to that conclusion >>about Rebel 10. Would Hiarc 7 being so strong in the endgame win playing the >>other side of the end game in its games with Rebel 10? > >I didn't try, but I am sure it would. Some of these endings were a dead draw, >even with symmetrical pawns. Today I'll send the games to Steve and you can >check this yourself. > >>>- Fritz 5 simply didn't have a chance and was outplayed by H7b in all phases of >>>the game. Fritz drew 4 games, lost 6 and won none. >> >>This is a good sign for Hiarcs 7 to do this, but Fritz 5 out classed Junior 5 >>too, for the first 8 games in my match. > >Sure. The result might be reversed in 10 more games, at least in theory. >Watching the games gives you a very different impression, though. > >>>- Junior 5 was the tough cookie for H7b. Until the very last game, H7b didn't >>>win one single game from J5 and I didn't think it could... >> >> I am finding Junior 5 very strong, > >Very! > >> it seems stronger then Fritz 5, but only a >>bit stronger.(after 21 games Junior 5 is only +1 over Fritz 5.} > >I don't think you can estimate strength after one single match. I'm not sure what you mean here. I played and posted many matches with Fritz 5. I know how strong this program is. I just don't play all the programs before posting the results as you did here. And I see other Junior 5 results on here and in other places, So I got a good feel how well Junior 5 plays. > >> I find it hard >>to think that Hiarcs 7 can out class Fritz 5 so badly, and itself be so out >>classed by Junior 5. > >Happens ALL the time. A beats B, B beats C, C beats A. This is one of the >reasons that makes so important to play as many different opponents as possible. > >> It can be very hard to get a good read on chess programs >>from a small sample. > >I agree. Particularly if you take into consideration only a match between two >programs. That's why I intend to include in my tournament all the new versions >that are coming now. We'll see what happens in the end. Still, >60% after 30 >games against the very best is at least an indication that H7 seems to be very >strong, don't you think? > >Enrique > >>>In my opinion, the weakest point of H7b is king safety. I know Mark is still >>>working on it, so the final version of Hiarcs 7 might be better in this regard. >>>Endgames are the best I have seen in a micro. H7 seems more selective and >>>quicker than H6, more solid and more knowledgeable. Overall, quite a bit >>>stronger than H6, which is already a great program, and in my opinion a very >>>firm candidate to #1. >>> >>>Tomorrow I will send the games to Steve and he will post them here. >>> >>>Enrique
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.