The "K" study for real (post deleted)
I deleted this post because it was wrong -- I said the authors considered only players whose initials were both K, but that was wrong. It was players who had EITHER initial K.
Thanks to commenters below who pointed this out.
LMK if you want to read the deleted post -- I'll e-mail it to you. I've left the comments below that told me I was wrong.
Labels: baseball, critiquing a study based on secondhand sources
posted by Phil Birnbaum @ 11/26/2007 10:26:00 AM
6 comments
6 Comments:
I believe that the study was done on players with _either_ first-initial K _or_ a last-initial K, not just both. Page 4 of the study states, "Accordingly, we predicted that players with first- or last-initial K would show an increased tendency to strike out."
Also, on page 5, their conclusions were that, "K was both the first and last initial associated with the highest strikeout rate," not that players with both initials as K.
But you'd certainly be right about the author's conclusions, if it was "KK vs non-KK".
Oh, wait, maybe you're right. Let me check again.
OK, I think they did use both initials. Reason: there wre 6397 players total, and 6395 players in the non-K group. This leaves two players in the K group, which exactly corresponds to the two KK players (Keltner and Kevin Koslofski) with 100+ PA in their career.
By coincidence, the total number of PA, 6424, is close to the number of players (6397). I will correct the post.
Phil - I love your blog and am generally impressed by your analyses. But I don't think you're reading the paper accurately. It's not both initials (it's first OR last name starting initial), and the analysis is multivariate (including year and country of origin).
Maybe I'm misinterpreting again. Sigh. Will be back.
OK, corrected in new post. Thanks, drmarcey and joe. Sorry for screwing this up.
Post a Comment
<< Home