As many of you know John Dewan publishes a "Stat of the Week" on the ACTA publishing site. These are often inciteful but I'm a little perplexed by the stat for last Friday January 11th which will appear in the upcoming book from ACTA Sports, The Bill James Gold Mine, available in February 2008..
In that nugget he notes that...
"Scott Podsednik has proven quite effective as a leadoff man. Until last year. Every year prior to 2007 Pods' teams have scored significantly more runs when he led off an inning than when others have led off."
This is accompanied by a table that illustrates how in 2007 the team scored -.01 runs fewer per inning when Podsednik led off than when his teammates did so.
While I haven't done exhaustive research on this I do know that Podsednik went from being a leadoff hitter in previous years to batting further down in the order in 2007. In fact, here are his plate appearances by lineup position since 2003.
Year/Pos 1 2 3 4 5 6 7 8 9
2003 279 291 2 27 11 2 1 15
2004 703 8 2
2005 564 2 2
2006 574 2 1 11 3
2007 89 19 3 2 27 95
Clearly a large part of the difference is due to the fact that the two and three hitters in the order are guaraneteed to bat when Podsednik leads off an inning from the leadoff spot in the order as opposed to when he's in the 6th or 7th hole. A little follow-up would be to see whether on average when a leadoff hitter leads off an inning it typically raises scoring by at least +0.25 runs as in the case of Podsednik in previous seasons or whether his totals actually indicate that he was a hinderance in the leadoff spot (as might be guessed by his career .338 OBP). It's also the case that lineup balance will play a role since it could be the case that the two, three, and four hitters for the White Sox were relatviely more potent than the rest of the lineup when compared with other teams.
In any case this Stat of the Week, while interesting, doesn't really show what it purports.
125x125_10off+copy.jpg)
1 comment:
Dan,
I've got access to Bill James online that shows this stat among others in all its glory. You are right to question it. All the stat does is show hows Pods does when leading off an inning, followed by non-random next 3 hitters, compared to other guys leading off innings followed by other non-random next 3 hitters.
There is no control group here, there is no "one thing" that is different to which you can attribute the difference.
At some point, everyone will realize that all you want is Linear Weights by the 24 base/out state (or by the Game State).
Tom
Post a Comment