-=> Terry May took a called strike three for saying to Paige Miller <=- 
 TM> On 15 Dec 96, Mark Wohlers K'd Paige Miller for saying:
 
 PM> I agree that it will be a small sample size per team, 15 games per team,
 PM> but that adds up, doesn't it? Multiply that times the number of teams,
 PM> and all of a sudden you have a large sample size. Statistically, it
 PM> doesn't really matter that this is less than 10% of all games played.
 PM> And it will be much better information about how the leagues compare
 PM> than we have now.
 TM> I still think it's too small of a sample size to draw any meaningful
 TM> conclusions.  Perhaps over a period of years you can draw a
 TM> conclusion, but not from one season. 
I would think that a bigger problem is that it is not a random sample of
the teams. In other words, you don't have equal chance of each AL team
playing each NL team. But you will never get "random sampling" of
baseball games in this manner, it isn't going to happen. Other than that,
10% of the total number of games seems fine for me to draw conclusions.
By the way, most types of random sampling (such as election polling) rely
on much less than 10% of the total votes cast to yield conclusions.
... You tell 'em, Bald Head, You're smooth.
___ Blue Wave/DOS v2.30
---
 * KMail 3.10o Knight Moves
--- QScan/PCB v1.17b / 01-0406
---------------