Author: Roger D Davis
Date: 18:22:18 12/25/00
Go up one level in this thread
On December 25, 2000 at 18:41:53, Bruce Moreland wrote: >On December 25, 2000 at 15:21:34, Roger D Davis wrote: > >>Perhaps there should be a minimum and maximum number of games, then. We would >>have to decide what these parameters would be, but once the maximum was reached, >>if statistical significance was not found, there would either be no champion, or >>there would be co-champions. >> >>As for the fact that other sports do not require statistical significance, that >>is just baloney. It requires me to use other sports as a criteria for what chess >>should be, and I refuse to do that. Totally bogus. >> >>Roger > >Analogs are useful, as they allow you to predict how things will work. > >There aren't many people out there saying, hey, we need more statistical >significance in these world championship matches. The reason is that such an >effort is unnecessary, impractical, and devoid of entertainment value. > >bruce Ack! I think you're missing the point, Bruce. Change starts somewhere. The past should not always be a criterion for the future. There is no moral imperative that is inevitable given the laws of physics that says, "Thou shalt never have a "statistically significant champion." If you see such an effort as unnecessary and impractical, that is certainly your opinion and you have a right to it. I have a different opinion. By the way, since analogy is different from identity, analogy may be helpful, but its limitations need to be kept in mind. For me, I see the differences between other sports and chess to be too great for the analogy to be useful, since I am interested in reform. Since you have another opinion, you see the similarities to be greater than the differences. And that's just fine. :) I would find it interesting to look at the history of world chess championships, and determine which show statistical significance. Then we can say that, for example, "Player A was clearly better than Player B," and have clearly mean something. Moreover, we might find that there are very few times in the history of chess where this occurred. I think that's interesting, and I think that, whether or not the idea of a "statistically significant champion" ever catches on, the notion of statistical significance is useful in determining whether it can really be shown that one player was better than another in a given match. Roger
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.