Author: Don Dailey
Date: 20:27:45 01/17/98
Go up one level in this thread
Bob, Ok, let's go with 10 seconds. I want to be able to do "quick" runs too and still be able to solve a few problems! I agree that for this kind of purely tactical test we do not want multiple solutions. Anything that creates ambiguity we will dismiss. I'm not worried about too many problems. I'll bet people discover multiple solutions and such and we end up dropping a lot of them. If the set is too big we can partition them or something. I would rather have too many than not enough. Now, let me see if I can find this set somewhere. - Don On January 17, 1998 at 22:57:13, Robert Hyatt wrote: >On January 17, 1998 at 22:39:08, Don Dailey wrote: > >> >>This sounds like a good idea to me. We take 3 programs including >>Crafty, >>and throw out any problem that all 3 get in less than n seconds. I will >>suggest 20 seconds but am open to any value you might choose (as editor >>of the set!) Any problem proven to be troublesome (multiple solutions) >>is >>also chucked. >> >>I will volunteer to be 1 tester but am not sure my program is good for >>this. I use very few extensions and may not solve problems quickly >>so Cilkchess may not be appropriate as an "easyness" filter. >> >>- Don > >I don't consider Crafty a great "finder" either. I don't spend a lot >of time trying to go deep. However, one other issue: I want positions >that don't have multiple solutions. And I don't want obtuse positions >where you can either win the queen or mate. I want my program to find >winning moves. I'd claim that if I win your queen, I'm probably going >to >win the game just as surely as if I find a mate. I found a few ECM >positions that fit that. ie one with a +16 eval but it's wrong. I'm >not wanting this to be a deep mate suite, but simply one where there is >one obviously best move... > >your 20 second limit sounds high... although it might be ok... IE that >could leave us with a suite that takes a while to run. I ran the entire >thing at 20 seconds and got something like 569 right I believe. Plus >another >10-15 I'd argue about. But that would eliminate a big chunk. But would >leave us with a suite of 300 that would produce a 0 for most programs at >times < 20 seconds. Not bad, but 5 hours to run 'em at 60 secs...
This page took 0.02 seconds to execute
Last modified: Thu, 07 Jul 11 08:48:38 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.