In the June 26 Nats-Cubs broadcast, Washington announcers Bob Carpenter and F.P. Santangelo had a conversation about how managers use statistics, and in particular how Matt Williams managed a 16-inning epic against Milwaukee.
“He was relying on batting average with runners in scoring position,” Santangelo said, “and to me that’s the best stat going.” They added that Cardinals manager Mike Matheny uses it, and Matheny told them that “it was a lot of managers’ favourite stat.”
Go ahead and freak out a little. But I’m curious. If you were judging hitters based on batting average with RISP, how different would your judgments be than if you judged them based on AVG, wOBA, or wRC+?
I can’t figure out how to insert a table, so here is a handy and also dandy chart for your surveyal, with helpful, pretty colors! You shall behold the 2013 leaders – minimum of 100 plate appearances, to include pinch-hitters who might pop up in a 16-inning game – for average, wOBA, wRC+, and BARISP. Hitters who appear in all 4 columns are colored peach, and hitters who appear in 3 of the 4 columns are colored blue. (Note: Hanley Ramirez should have been in the fourth column, too. Somehow the leaderboard I pulled left his name out.)
What do you notice? Well, yes, there is a lot more overlap between the first three columns than the last one. I might be counting wrong, but it looks like over half of the BARISP leaderboard does not appear in a single other column. (Many of them are Cardinals.) And yet, the truth is, almost all of the top 25 BARISP leaders were, in fact, good hitters in 2013. The three worst hitters on the list, by wRC+, are Michael Brantley (104), Manny Machado (101), and Brandon Phillips (91). That’s not a terrible bench. (On the other hand, Pete Kozma looms.)
The truth is, good hitters are good hitters. A manager relying on BARISP would not suddenly disregard Josh Donaldson, Miguel Cabrera, or Paul Goldschmidt.
Who would lose the most from a reliance on BARISP instead of advanced stats? Arguably, the guys who appear in the wOBA and wRC+ columns, but not the BARISP one. There are 15 of those players, favored by advanced numbers but not by “the managers’ favourite”. Of those 15, 8 still have BARISPs above .280. Here are the bottom five:
5. Khris Davis, .250 (43 PA)
4. Shin-Soo Choo, .240 (144 PA)
3. Joe Mauer, .239 (113 PA)
2. Yasiel Puig, .234 (99 PA)
1. Jeff Baker, .162 (44 PA)
On my custom BARISP Snub Leaderboard, there are only a handful of players a real manager might pass over. (Who would bet against 2013 Joe Mauer?)
In other words, even though a true stathead might yelp in terror at the thought of his team’s manager using BARISP to select a hitter, the process does not actually yield many bad results. Good hitters will be good hitters, even if your measure is slightly faulty. Your coach might bench Yasiel Puig for Brandon Phillips, which obviously would be bad. It’s also unlikely. More likely might be benching Khris Davis for Michael Brantley, and would you truly be that offended?
On the other hand, Pete Kozma looms.
Print This Post