27.10.05

Small n

After updating my post-season analysis with the results of last night's game, so that we now have a World Series victor for 2005, it appears that:

1) the wild card team is still not significantly more likely to make it to the World Series or to win it (chi-squares: .727 and .866).

2) the team with the best record in its respective league is significantly more likely to make it to the World Series (chi-square: 3.96, significant at the 5% level) . . .

3) . . . but is not significantly more likely to win the World Series (chi-square: .866).

4) wild card teams are not significantly more likely to win a division series, but the team with the best record in its league is (chi-squares: .97 and 3.88).

Of course, none of this tell us how much more likely these outcomes are. That would need a degree of analysis that I haven't the time to undertake at the moment.

The punchline: n=80 is not a very big number at all, and n=88 isn't either. We'll need several more years of the modern post-season before we can actually draw any really important conclusions about the wild card's effect on baseball's seasonal results.

[Posted with ecto]

No comments: