Author: Serge Desmarais
Date: 02:52:52 09/06/98
Go up one level in this thread
On September 05, 1998 at 21:25:31, Robert Henry Durrett wrote: >On September 05, 1998 at 16:53:39, Serge Desmarais wrote: > >><snip> > >> My opinion is that Genius is easier to play against for amateurs than other >>programs, because it has a quiet style of play. Note that I said EASIER in a >>sense that meant that the games would be longer before the balance would tilt in >>Genius favor ; I never meant WEAKER. Anyway, this to say that if I made other >>draws with Genius 3, I was never able to draw against good masters in tournament >>games (maybe a psychological blocking? hehehe)... >> >> >>Serge Desmarais > >This "easier to play against for amateurs than other programs, because it has a >quiet style of play" idea seems to me to be worth pursuing. I wonder if other >programs, such as Fritz, could be adjusted to play the same way. If so, it >would be possible maybe to determine if this style is effective independent of >engine design. > >Incidentally, I wonder how strong the humans would have to be if the Genius were >in it's latest version, playing on a very fast computer? Fritz is just the opposite of Genius! With Fritz the board is "on fire" quite soon and for me I have more problems playing Fritz (either 3 or 4) than I have playing Genius, despite the fact that Genius is rated higher than Fritz on the SSDF list. Also, note that Genius 5 is not sensibly better than Genius 3 or 4, just slightly (still referring to the list of the SSDF. Fritz takes risks and could make speculative sacrifices (pawns most of the time) to open the position, while Genius plays it "safe", avoiding situations that could get "out of control". So Fritz, playing other computers, could get in trouble after a pawn sacrifice (due to the horizon effect) and lose a long and painful ending (with one or 2 pawns less after the attack stopped). Did you look at the difference in the results of the matches played by the SSDF? You could see that Genius would lose less games than others against stronger programs, but its winning percentage would be lower too. (To win, one has to take risks and Genius is "anti-risks".). In the last SSDF list, the match between Genius 5 and Fritz 5 was tied 9 to 9. Then, Fritz exploded to take the lead 29.5 to 18.5. That could be caused by Fritz book learning (which avoids drawing and losing lines VERY fast) which is not the case with Genius 5, which has independent learning files whose size never change apparently. Anyway, just to say that even against Fritz 3, I am usually in a losing position by move 30 (attested by SEVERAL games I have here), while (playing at a friend's home since I do not own Genius 5) I was still not losing after 80 moves (wasn't it for a stupid mistake, it was a draw)! Also, my personnal temper (and for fun!) makes me enter the dance of sacrifices and counter sacrifices. Maybe later I'll post a game against the Fritz 3 engine in which I made a very complicated sacrifice and got a WINNING position, despite the fact that Fritz, at first, thinks my move was a bad mistake! But later, I went wrong and got lost in the complications and lost the game (by move 30 again!). Serge Desmarais
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.