Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Playing style of top engines...

Author: chandler yergin

Date: 18:30:32 10/18/05

Go up one level in this thread


On October 18, 2005 at 21:16:36, chandler yergin wrote:

>On October 18, 2005 at 18:50:19, Dann Corbit wrote:
>
>>On October 18, 2005 at 18:30:21, chandler yergin wrote:
>>
>>>On October 18, 2005 at 17:47:51, Uri Blass wrote:
>>>
>>>>On October 18, 2005 at 17:12:05, Robert Hollay wrote:
>>>>
>>>>> I apologize if this question sounds odd.
>>>>> But if it's true that chess game has a SOLUTION - the perfect chess game -
>>>>>then it should be logical that the top chess engines tends to that perfection.
>>>>>In other words, they will more and more often find THE BEST move in any
>>>>>position.
>>>>
>>>>There is often more than one best move.
>
>I said it sounds like one ;) (lighten up)
>
>>>
>>>Sounds like an Oxymoron to me..
>>
>>[D]3b1qk1/r5rR/6Q1/pppPp1p1/4P1N1/1PP3P1/1P3PK1/7R w - - acd 4; acn 2; acs 0; bm
>>Nh6# Rh8#; ce 32766; dm 1; id "C.A.P. 811930"; pv Nh6#;
>
>New game,
>[D]3b1qk1/r5rR/6Q1/pppPp1p1/4P1N1/1PP3P1/1P3PK1/7R w - - 0 1
>
>Analysis by Shredder 8:
>
>1. +- (#1): 1.Rh8#
>2. +- (#1): 1.Nh6#
>3. +- (#2): 1.Nf6+ Qxf6 2.Rh8#
>
>Weapon of chioce
>
>(,  18.10.2005)
>
>>
>>>>> If so, then their playing styles become more and more monotonous and are
>>>>>gradually moving to a uniformity.
>>>>
>>>>No because there is often more than one best move.
>>>
>>>Another oxymoron
>>
>>See above.  It is not uncommon at all.
>>These are from the 1924 NY Chess Club match.  See also:
>>[D]8/8/8/8/1K3k2/1p4r1/1N6/8 b

New game,
[D]8/8/8/8/1K3k2/1p4r1/1N6/8 b - - 0 1

Analysis by Shredder 8:

1. = (0.00): 1...Ke4
2. = (0.00): 1...Ke3
3. = (0.00): 1...Ke5
4. = (0.00): 1...Kf3
5. = (0.00): 1...Kf5
6. = (0.00): 1...Kg4
7. = (0.00): 1...Kg5
8. = (0.00): 1...Rf3
9. = (0.00): 1...Re3
10. = (0.00): 1...Rd3
11. = (0.00): 1...Rc3
12. = (0.00): 1...Rg2
13. = (0.00): 1...Rg1
14. = (0.00): 1...Rh3
15. = (0.00): 1...Rg4
16. = (0.00): 1...Rg5
17. = (0.00): 1...Rg6
18. = (0.00): 1...Rg7
19. = (0.00): 1...Rg8

(,  18.10.2005)

- - acd 19; ce 401; pv Ke4 Kc4 Re3 Na4 Kf4 Nb2
>>Rf3 Na4 Ke5 Nb2 Ke4 Na4 Kf4 Nb2 Re3 Kd5 Rc3 Kd6 Rf3; pm Ke4; bm Ke4; id
>>"ECE3.0447";
>>[D]3b1q1k/r4r1p/7R/pppPpQp1/4P1N1/1PP3P1/1P3PK1/7R w

New game,
3b1q1k/r4r1p/7R/pppPpQp1/4P1N1/1PP3P1/1P3PK1/7R w - - 0 1

Analysis by Shredder 8:

1.Rxh7+ Kg8 2.Rh8+
  +-  (#3)   Depth: 5/5   00:00:00
1.Rxh7+ Kg8 2.Rh8+
  +-  (#3)   Depth: 5/5   00:00:00
1.Rxh7+ Kg8 2.Rh8+
  +-  (#3)   Depth: 5/5   00:00:00
1.Rxh7+ Kg8 2.Rh8+
  +-  (#3)   Depth: 5/5   00:00:00
1.Rxh7+ Kg8 2.Rh8+
  +-  (#3)   Depth: 5/5   00:00:00
1.Rxh7+ Kg8 2.Rh8+
  +-  (#3)   Depth: 5/5   00:00:00

(,  18.10.2005)


- - acd 4; acn 740; acs 0;
>>bm Rxh7+; ce 32762; dm 3; id "C.A.P. 811926"; pv Rxh7+ Kg8 Rh8+ Kg7 R1h7#;
>>[D]1k6/1P6/K7/8/P7/2P5/8/8 w - - acd 9; acn 1034866; acs 25; bm a5 c4; ce 32748;
>>dm 10; id "C.A.P. 805301"; pv c4 Kc7 Ka7 Kd7 c5 Ke7 c6 Kf7 c7 Kg7 c8=Q Kf7 Qd7+
>>Kf6 b8=Q Kg6 Qg8+ Kh6 Qdh7#;
>>
>>>>> Weak engines have very different styles, often very aggressive, so they are
>>>>>great for weak players to play against. But newer - and stronger - versions of
>>>>>these same engines are playing in a much more cautious style (see Golem).
>>>>>Similarly, strong engines have "personalities" to make their playing style more
>>>>>interesting, but the default personality is usually the strongest.
>>>>
>>>>No
>>>>
>>>>The default version is usually not the strongest because programmers cannot tune
>>>>every parameter to be the best because they may need thousands of games for
>>>>that.
>>>>
>>>>The pawn value in the default personality may be 100
>>>>It is possible to increase the default value to 101 or to reduce it to 99
>>>>
>>>>In most cases one of this options is going to be at least 0.1 elo stronger than
>>>>the default personality because programmers have not enough time to play more
>>>>than 1,000,000 of games to detect difference of 0.1 elo.
>>>>
>>>>Same for other default values.
>>>>
>>>>Uri



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.