Author: Bruce Moreland
Date: 18:55:01 04/20/00
Go up one level in this thread
On April 20, 2000 at 08:34:05, James T. Walker wrote: >On April 19, 2000 at 23:53:31, Robert Hyatt wrote: > >>As I had mentioned a while back, I have a sack full of quad xeon 550 machines >>in a beowulf cluster. While waiting on a few final pieces to arrive, I decided >>to do what I thought was an interesting test: >> >>two identical machines, and I mean _identical_. Quad xeon 550's, 27 gigs of >>SCSI disks in a raid-0 (striping) configuration, 512mb of ram, etc. IE >>everything is identical, with all the 3-4-5 piece compressed tablebases, >>same opening books, etc. >> >>The only difference was that 'crafty' plays computers and humans, while scrappy >>only plays humans. Several of us had postulated over the years that if you only >>play humans, you can drive your rating through the roof. Using the same >>formulas (5 3 blitz or faster, 60 60 standard or faster, or most any bullet) >>I have been watching the two programs for a month now. And they seem to >>hover at the point scrappy == crafty+100, roughly. Standard has crafty >>actually higher, but that is because crafty is playing standard against >>computers, while scrappy is playing very little standard as humans seem to be >>avoiding that for the most part... and those that do play standard play crafty >>as it is better known. >> >>100 points was a surprise... as I thought it would be more. At present crafty >>is at 31126 and scrappy is at 3219 blitz (which is the most stable ratings since >>most games are blitz). >> >>It seems that not playing computers is _not_ a way to grossly inflate your >>rating, unless you consider 100 as inflated. Note that a rating of 3200 is >>very high, considering that there are not a lot of GM players that are rated >>even 3000. I watched scrappy play a 16 game match earlier this week, it won >>8 games, lost one, then one 7 more, for a 15-1 result (5 3 blitz). It lost 32 >>rating points for the effort. :) >> >>I am going to continue the experiment until I get the rest of the beowulf >>hardware (another quad box and a fast ethernet switch to complement the >>giganet switch). If you watch the ratings, you will get a feel for the >>difference in playing only humans and humans + computers... > >Hello Bob, >Interesting as it is, I think that to get a real feel for Crafty's strength on >ICC you need to change Crafty's formula to "Formula=rated". As long as Crafty >and most other computer accounts use 4 lines of formula to protect their ratings >against various attacks the ratings will forever be inflated. Just one opinion >of course. I believe that an "anything goes" attitude is the only way to get a >true rating. >Jim Walker A less restrictive formula is more fun for the opponents and will result in more losses for the computer. If you know you are good at 5 12 and bad at 3 0, you will play 5 12. If the computer is best at 3 0 and is bad at 5 12, it will have rating problems if it plays 5 12. This makes not a bit of difference to you if you are running a computer account if you aren't trying to do one of two things: 1) Maximize rating, either average rating or "highest" rating. 2) Maximize consistency by reducing effects caused by time control differences. I think that you'll still have rating spikes no matter what time control you select. If I run one version and it's playing well and it has a huge rating, people will congratulate me and ask to buy my program. I can run the same version the next day and it gets cratered, and people ask me if I've done something bad to my program. Sometimes the same people (hello Lonnie). The same thing is true at chess tournaments. You are only as good or bad as your last result, even if you haven't changed your program in a year. bruce
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.