Author: Komputer Korner
Date: 06:01:28 07/31/98
Go up one level in this thread
On July 30, 1998 at 12:55:26, Shaun Graham wrote: > >> >>While your way of (not calculating an interim rating and waiting until all the >>games were put in at once and then doing an all at once calculation is more >>accurate) it is not practical because only one program's rating could be >>calculated this way. > >It is very practical if your goal is to get the most accurate rating, further >because the calculations don't have to really be done by hand, because there are >so many programs, and database programs(excel type)where the formulas can be >entered it really wouldn't take too much extra effort. > > All the other ones would have to have interim ratings in >>order for your method to work. > >Not exactly correct i think. In the calculation that i did, i just hypothesized >all computer opponents to be 2400, now of course many of the programs are >stronger than this, but almost everyone including people with extreme computer >biases such as Bob Hyatt accept that strong programs are at least 2400. So by >setting the rating at 2400 some slight accuracy is lost, but no one will say >that your rating calculation is inflated. If i didn't set all opponents rating >at 2400 there would be much wrangling against the 2535 USCF rating that i >calculated for chessmaster based on 53 40/2 games. > > > > So you see the present system is the only one >>possible. > >Ah ah ah, please be carefull when saying "the only". > >There are different ways of doing incremental calculations. The system >>in practice is to take lump sum tournament date starting and endpoints and the >>other way would be to do it after each round. The increased mathematical >>accuracy does not justify the increased cost however. > >The cost is only high when you work harder instead of trying to work smarter. Well if you would assign a rating of 2400 to all of your opponents, why wouldn't you do this to your program? I admit that when the rating system was first started, the ratings had to be assigned for at least 1 program to start it off, but now that we have established ratings (SSDF for example) why would we need to assign ratings? Because programs only learn to avoid certain lines, they really don't learn like humans anyway so no rating system will make their ratings like human ratings. Besides the SSDF list is only good for comparative purposes. You are attaching too much importance to the isolated rating number. Ratings abhor a vacuum. You need lots of competitors to have a good system and the SSDF is a closed shop. As for other systems, ICC is pretty wide open but it's problem is that it doesn't allow more than 1 time control period. -- Komputer Korner
This page took 0.01 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.