Friday, November 09, 2007
Ridiculous “Science”
Posted by David GasskoAn article on the Newsweek website highlights a new psychological study (and I use that word lightly) that supposedly proves that players with the letter "K" in their initials are more likely to strikeout than other players. I was pretty skeptical of that conclusion, so I decided to run the numbers myself.
I used a somewhat different dataset than the authors, looking at all players between 1954-2006, but the results should be about the same. Indeed I found that players with the letter "K" in their initials struck out a little more often than average, 15.5% of the time versus a 15% average for all players. Due to the sample size (about 465,000 plate appearances), the result was quite significant, about 8.5 standard deviations away.
But what happens if we look at other letters? Suddenly, "K" doesn't look so special. In fact, there are 8 other initials with higher strikeout rates; most impressively, "J," which happens to be 27.5 standard deviations away from what we would expect. Due to huge sample sizes, only a few initials are insignificantly different from the overall average. Why is that? I have a few theories, but frankly, I don't care enough to type them out. This whole idea is ridiculous, and I don't want to give it any more time than I already have. If anyone is interested, here is the data for all 26 letters:
Letter PA SO SO/PA SD A 562486 83689 0.149 -3.26 B 1469882 216353 0.147 -10.67 C 1032493 155140 0.150 -0.22 D 1031912 162318 0.157 19.79 E 335659 50852 0.151 1.89 F 403298 56323 0.140 -18.98 G 825656 121436 0.147 -8.28 H 621357 92994 0.150 -1.48 I 50991 7904 0.155 2.95 J 1260983 200621 0.159 27.53 K 464664 71920 0.155 8.47 L 624412 91771 0.147 -7.44 M 1286492 193153 0.150 -0.62 N 156653 25547 0.163 14.11 O 261779 33667 0.129 -31.10 P 551803 85030 0.154 7.81 Q 16914 2635 0.156 1.98 R 1101643 160904 0.146 -12.56 S 873970 136209 0.156 14.43 T 750565 111151 0.148 -5.44 U 27699 3861 0.139 -5.10 V 201911 29294 0.145 -6.60 W 576619 86152 0.149 -1.97 X 1400 259 0.185 3.63 Y 66924 8995 0.134 -11.53 Z 33737 5510 0.163 6.67
David Gassko is a former consultant to a major league team. He welcomes comments via e-mail.