Author: Melvin S. Schwartz
Date: 07:21:38 06/10/99
Go up one level in this thread
On June 10, 1999 at 05:10:46, Shep wrote: >On June 09, 1999 at 17:42:03, Melvin S. Schwartz wrote: > > >>>I will not make a final decision about these issues (Tiger 11.5 in/out, Rebel >>>10a/10d, CM6000/MCP8) until the starting date anyway (this will be sometime >>>during the week starting on June 21st). >>> >> >>I personally think you should include CM6000 regardless of what your match >>between MCP8 and CM6000 reveal. In fact, they probably both belong in your >>tournament. I also feel it isn't right to have the two Chess Tiger programs >>because that isn't fair to the others - except for Chessmaster because your > >I want to have the strongest programs in, regardless of how "fair" this seems in >absolute numbers. If the top 5 programs were Rebel 6 to 10, I would have all of >them in. Remember this is not W(M)CCC where - technically - not the programs are >the participants but their programmers, and each programmer may only bring one >program. I've had many tournaments where up to 4 versions of one program >participated. > >>version is not comercially available and is your own modification of an older >>CM5500. Chessmaster 6000 is rated quite highly by the SSDF. Is their rating > >Actually, "CM 5555" _is_ commercially available, you just have to buy CM 5500 >and adjust the personality settings as described on my site. >There are no further modifications beside that. > >>system so bad and CM6000 that bad as well? I think a lot of us would like to >>know how CM6000 would do in your tournament regardless if MCP8 wins the >>qualifier. Put them both in and take out one Tiger. You intend to have Rebel 10 >>and both Tigers. Isn't Ed Schroeder associated with all of them? Is it right to >>have 3 programs from Ed? I believe the credibility of your tournament will be > >"Associated" is a very loose term. In this sense, Chessbase is "associated" with >Junior, Hiarcs, Nimzo, ... >Tiger is the work of Christophe Theron; the fact that he cooperates with Ed and >that Tiger will be some sort of "Rebel add-on" or "engine" does not mean it's >Ed's program now. :) > >>hurt some if you don't include CM6000. By the way, you should know I don't own > >Let's put it that way: if CM loses the qualifier series against MChess, it has >accumulated enough below-average results that it would not be fair if I still >included it. Because then for the same reason you could ask why Zarkov and >WChess are not in. That's my idea of "fairness": If I have a principle of not >including programs which had suboptimal results in my tests, I make no >exceptions. >The qualifier against MChess is already a courtesy act out of respect for the >opinion of those of you here on CCC who have pressed me to give CM 6000 a second >chance. >If it fails there, there will be no third. > >BTW, first game was a draw yesterday. Your analogy between CM6000 and Zarkov is really not fair at all. CM6000 is rated just 1 point below Fritz. Where is Zarkov rated?? Furthermore, I have seen a tournament where MCP8 was 6.5 - 3.5 over Hiarcs 7. Now, suppose one program such as MCP8 does better than CM6000 in your qualifier - so what? The fact that CM6000 was rated number 1 by the SSDF, and is now number 2 by a mere single point, would dictate that it belongs in any tournament until proven unworthy of competing against other top programs. There is much controversy regarding CM6000. You can do us all a great service by including it in your tournament so that we will know if it truly is unworthy of inclusion in your future tournaments. I admire the way you run your tournament on identical computers. It is for that reason I look at your results with a great amount of interest. I believe that running program against program on identical hardware presents the most feasible way to conduct a tournament. You are to be congratulated for your effort - I think it's great! I also believe that tournaments at a time control of 40/2 is the absolute most meaningful in terms of evaluating a programs strength - just as 40/2 is the most meaningful in determining human strength. Even if you're put off by the idea of including CM6000 in your tournament, you could present an extremely important result by putting it in your tournament and let us all know how it does againt the top programs. Regards, Mel > >--- >Shep
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.