Author: Robert Hyatt
Date: 10:33:01 04/19/04
Go up one level in this thread
On April 19, 2004 at 02:50:06, Ed Schröder wrote: >On April 18, 2004 at 20:36:56, Steve Maughan wrote: > >>I'm looking to implement the new ELO rating support in UCI 2. In the past I >>haven't given much thought as to how one could make an engine predictably >>weaker. Does anyone have any suggestions? After the little thought I have >>given the subject here's what I've come up with. >> >>1. Search to a shallower depth / shorter time. This seems somewhat >>unsatisfactory as it will ignore the other setting. >> >>2. Add random elements to the evaluation term - the bigger the term the weaker >>the play. >> >>3. Round off at the root. So to weaken the engine slightly you would have the >>root round off to 0.1 pawns, a further weakening would be in units of 0.5 pawns >>etc. This means that if a value came back as +1.23 the engine would regard it >>as +1.20. I like this idea since it is relatively simple to implement :-). >>However I'm not sure how effective it would be at actually weakening the engine >>(it may speed the engine up since there may be more cuttoffs!?!). >> >>4. Within the tree you could ignore some moves e.g. all losing capture after x >>depth. This is also interesting. It may closely match how humans think and >>err. Of course it is more complex to implement. >> >>Any other ideas. This could be an interesting area for discussion as it has >>virtually no commercial value (IMO nobody is going to buy an engine that is >>weak). I'd be interested to know of other idea. > >One idea. Have a look at Rebel's [Playing Srength = 100] parameter. When you >are decreasing this "100" value you will see Rebel's NPS drop. How is it done? >By just calling some expensive IO stuff such as kbhit(), clock() multiple times >somewhere in the tree, make_move(), eval() or wherever. > >Works like a charm, 10 minutes to program, a few hours testing (tuning) so that >when the value < 10 your NPS has dropped to 500 nodes (or so). > >Ed I have done that in the past, and it has never produced a weak player that feels like a weak human player. Fiddling with the speed only tones down the tactics. But how many 1200 players understand pawn structure, majorities, passed pawns, candidate pawns, weak pawns, pawn levers, weak squares, etc? IE someone reported that sd=1 Crafty still plays over 1500 on ICC. It's losses are tactical losses only, not "positional losses" which is not how a 1300 player (or 1500 player) actually loses games... Mike's personalities in Crafty are a bit better in that they also turn down some of the positional scoring as well as turning down the tactics so that it feels somewhat more like a weak player...
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.