Thursday, July 15, 2010
The value of situational hitting
Posted by Jeff SackmannOn the occasion of Sean Forman adding Clutch to baseball-reference.com, Tango notes some of the best and worst career numbers. It's clear from the leaderboard that contact hitters tend to do well (Nellie Fox is at the top of the list!) and power hitters generally don't.
(With a stat named "Clutch," things can get confusing. From here on, "Clutch" with a capital "c" refers to the b-r stat; "clutch" with a lowercase "c" refers to the general idea of performance at important times.)
The comments at Tango's blog are insightful as usual, focusing on why the hitters at the top are there (switch hitters? contact hitters tougher to shut down with relief specialists?) and how to include clutch value in measures like WAR (does this really erase one-third of Sammy Sosa's value?). There are a lot of avenues for discussion here, that's for sure.
What interests me is something different. As I've noted, it's clear that contact hitters do well. Let's set aside the issue of why they do so. The fact that they do suggests that when we ascribe +10 clutch wins to Tony Gwynn, we're not necessarily saying that Gwynn was clutch due to his sterling character under pressure, or on the flip side, due to luck.
Gwynn was one of the great contact hitters of all time—he put the bat on the ball in more than 90 percent of plate appearances. If we knew absolutely nothing else about him, the fact that he was a great contact hitter would suggest that he would be worth more than average in the clutch.
As it turns out, if we take the 200 hitters with the most plate appearances in the expansion era, the correlation between their Clutch value and their contact rate is 0.54. Clearly contact hitting doesn't explain everything expressed by Clutch, but it explains some of it.
Innumerable previous studies have suggested that "clutch" (however you try to measure it) is either not a repeatable skill, or if it is, the "repeatable" part is very minor. But contact rate is based on strikeout and walk rates, and those are two of the most repeatable skills there are.
If contact rate is repeatable, contact rate is related to Clutch, and Clutch can be measured in runs or wins, then contact rate can be measured in runs or wins. Translating that insight to WAR stats is more complicated than that, but on the face of it, part of the enigma that is "clutch" is really about something else. And maybe Ichiro is worth a bit more than we think.
Jeff Sackmann is the creator of MinorLeagueSplits.com. With Kent Bonham, he founded CollegeSplits.com. Jeff and Kent blog about college baseball and the draft, and you can follow them on Twitter for bite-sized snacks of minor league and college stats. Jeff also has an email address.
If you have a man on third with 2 out in the bottom of the 9th and you need the run you would want Nellie Fox at the plate. My question would be, man on 1st, everything else the same, what power hitter do you want, assuming it is a power hitter that you want? Obviously contact hitters and power hitters have different functions on a team and measuring their relative value would have to be specific to their functions. You could also add a third party, the powerful contact hitter (20 to 30 hr range, higher contact rate than the Sosa’s)who might be preferable in the second scenario. Perhaps Sammy might be better measured in the third scenario where no one is aboard, everything else the same. I doubt if he would show very well there though either. Anyway, perhaps the way to measure these categories would be by slugging pct levels?