Chessmaster 10 - Page 2
| Forum Index > General Games |
|
Silvanel
Poland4736 Posts
| ||
|
BigBalls
United States5354 Posts
| ||
|
BigBalls
United States5354 Posts
| ||
|
pokerforums
Canada52 Posts
| ||
|
pheered.user
United States2603 Posts
| ||
|
PlayJunior
Armenia833 Posts
Guys! Computers play endings bad. They play certain endings awfuly. No strategy, no sacrifices for activity, no plan, no micro improvements (don't hurry principle), no two weaknesses, and so much pushing! The computer program problem is that endings need much more planning/knowledge compared with tactics than the midgame. Humans see tendentions in endings, while the computers don't. They just calculate... The endgame tablebases help a computer very much, but, unless it's a technicaly hard ending with a very small number of pieces left(Rook+f and h pawns vs Rook, try to win this against Fritz with 6-men tablebases installed )), humans are superior.Also, it doesn't matter what level you are, play humans, because if you play computers, by the time you will be afraid of all kind of tactics and complications(because the computer will always beat you). This is very bad tendention. | ||
|
pheered.user
United States2603 Posts
| ||
|
pokerforums
Canada52 Posts
![]() | ||
|
pheered.user
United States2603 Posts
| ||
|
soundwave
United States363 Posts
| ||
|
Soun
Poland373 Posts
| ||
|
PlayJunior
Armenia833 Posts
![]() About endgame: Nalimoff tablebases help a lot, but it doesn't change the overall picture. Just watch those damn games Kramnik vs Deep Fritz, games 1, 2, 3... Fritz was running on an 8-way Compaq Server with some Gigabytes of RAM and, but had some very big problems playing the endings. The same thing was in Smirin vs Computers match, where various programs used to push vs Smirin in endings and get weaknesses. It's possible to win an absolutely drawn ending to a computer, a one that you couldn't even dream to win against a descent 2600 GM. You don't need to have a won endgame, just one where there are some hard strategic decisions to be made(exchanges are an example where the comp sucks hard ), there is little tactics and there are not too many or too little pieces left on board. As a proof, just watch those games of Kramnik vs Fritz. | ||
|
Luhh
Sweden2974 Posts
Variation and complexity seem a lot higher in this game since it's a balance between capture/battle/territory etc. | ||
|
BigBalls
United States5354 Posts
| ||
|
PlayJunior
Armenia833 Posts
I suggest beginning from Rook endings. Agree? | ||
|
SurG
Russian Federation798 Posts
On January 07 2005 21:11 PlayJunior wrote: Bigballs, it seems we have to run some tests. We can set up some typical endgame positions that have very precise analysys and evaluation and let the comp play them. I suggest beginning from Rook endings. Agree? Let's not exploit the fact that computers don't understand fortress concept in rook endings =) How about some queen endings, doubt humans can play that with computer. | ||
|
iD.Surv
Belgium827 Posts
| ||
|
USMCgamer
Korea (South)255 Posts
As far as CM 10 goes I noticed the same thing about the computer opponents. They make very good and very bad moves... I prefer human players. Anyone know what the best free chess site is right now? Sometimes I get frustrated trying to find a game that I want on Yahoo... Edit: Oh yeah, I forgot about my biggest gripe. I have a dual monitor setup and it doesn't open properly with dual monitors. Most games do this but since it runs in a window the way it does this irks me. O well | ||
|
Naib
Hungary4843 Posts
I play chess on tournaments for 11 years by now![]() | ||
| ||
)), humans are superior.