Author: J. C. Boco
Date: 12:31:26 06/30/02
Go up one level in this thread
In the above thread (which I only perused) there is the complaint of the small sample of games played. I agree of course. I just have a question about the statistics used. First of all, I bet that most people on these boards has taken at least a basic statistics course. So have I. Sample sizes were always large when theory was being taught, often N was 500 or more. In most fields of endeavor, this is enough. When an undergrad in biochemisty, I had to take a particularly leathal class called "Biostatistics". This is a mathamatically rigorous discipline. The purspose of this branch of statistics is to draw meaningful conclusions based on small sample sizes inherent in live animal experiments (after all, the upkeep and experimentation on rats is expensive--let along "higher" animal models. This stuff was nasty. Sample sizes of less than 25 were the norm. There was awful, nested levels of analysis which was used to combine several different environmental conditions without having to replicate experiments in each of the different conditons. I took this class over 7 years ago and have had no need for it since, so I really cannot be more lucid in my description than this, and most certainly could not give meaningful suggestions on how to perform a nested-analysis of variance for ssdf data (other than saying that a round-robin type of tournament would probably avail itself into a nested biostatistics analysis.) I have typed too much already. All that text above was pretty much to set up the simple question: Is anyone at the SSDF aware of biostistics. If so is there a reason you don't use it? (Since it has been many many years since I've studied it, perhaps there is a trivial reason for this I don't remember). It's my perception that very few people who study the life science really get into computer chess, and therefore the branch of biostatistics is unknown to the SSDF.
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.