Sunday, May 15, 2005

Whoa Nelly? Statistical Information from Conor's Corner

Keywords: Red Sox, Orioles, White Sox, Yankees, statistics

My son, Conor, is exceptionally gifted with respect to mathematics and has degrees in both computer science and economics. He also loves baseball and growing up in Maryland, he is naturally an Orioles fan first, a Yankee hater second, and he tolerates the Red Sox (previously not much of a threat).

Here is an email he sent to me today...

I made my first foray into the world of baseball statistical extremism. Not content to read the articles where people write about how April records translate into full-year records, I wrote a perl script that grabbed the day by day standings for all teams in the 162-game era from www.baseball-reference.com. I then allowed it to input partial season records and had it spit back the year-end records of those teams sorted into buckets.

Here are the Orioles, Red Sox, Yankees, and White Sox through Saturday's games (note that only records where teams ended upplaying at least 160 games are counted, which throws out strike seasons).

Orioles (ie, distributions of teams that win 21-23 of their first 35 games):
<65 wins: 0
66-70 wins: 1
71-75 wins: 3
76-80 wins: 12
81-85 wins: 27
86-90 wins: 36
91-95 wins: 32
>95 wins: 36
Sample size: 147 teams

Red Sox (20-22 wins in their first 35):
<65 wins: 0
66-70 wins: 3
71-75 wins: 9
76-80 wins: 21
81-85 wins: 34
86-90 wins: 46
91-95 wins: 41
<95 wins: 41
Sample size: 195 teams

Yankees (17-19 wins in their first 37):
<65 wins: 7
66-70 wins: 28
71-75 wins: 45
76-80 wins: 64
81-85 wins: 55
86-90 wins: 61
91-95 wins: 42
> 95 wins: 17
Sample size: 319 teams

White Sox (26-28 wins in their first 36):
< 65 wins: 0
66-70 wins: 0
71-75 wins: 0
76-80 wins: 0
81-85 wins: 3
86-90 wins: 1
91-95 wins: 3
> 95 wins: 11
Sample size: 18 teams

The Yankees are still up against long odds if they expect to win more than 95 games, which only about 5% of teams have done given their current standing after 37 games.

A) Yes, he is employed
B) Yes, he has coauthored a book How Markets Really Work
C) No, I do not fully understand the wagering implications of this
D) No, I cannot assess the validity of the programming, data importation, and statistics

No comments: