Things The Stats Miss

Nate Silver wrote a post a few days ago about how a certain basketball player's value cannot be seen just by looking at his own stats, but must be judged by looking at his effect on the performance of his teammates. There are a variety of ways in which I think this phenomenon might be happening in baseball, too, that there just aren't any statistics to measure it with.

First of all, catchers. A catcher has three functions: they are supposed to hit, they are supposed to field, and they are supposed to work with the pitcher in crafting a winning game plan to mow down the opposing team's lineup. And I think it is generally admitted that, just as some catchers are better hitters than others and some catchers are better fielders than others, some catchers are also better at calling a game than others. Best I can tell, there are no statistics to measure this effect. And I can't see how in the world one would do so. You could, of course, take a given catcher and look at the average ERA of all pitchers pitching to him against those pitchers' average career ERAs when they weren't pitching to him, but there are soooooo many confounding variables. Most teams have one primary catcher in any given season, and a pitcher will spend most of their time pitching to that catcher. Thus, if they happen to be having a good year in a given year, that would show up in that catcher's stats. When they then get injured and have a few years of struggle, that would also show up in that catcher's stats. Some pitchers even have their own "personal catchers," in which case one catcher would have up to 100% of that pitcher's starts in a given year. Perhaps if you saw that a long-time catcher had compiled a significantly lower ERA on average with his pitchers than those pitchers had without him, you might be able to conclude that he had something to do with it, but boy it's messy. One person who I think would benefit from having this kind of stat would be Mike Piazza, a dreadful defensive catcher but from what I've heard perfectly competent at calling a game.

Second, lineup "protection." This one, I think, would be somewhat easier to measure. The idea is that a player who is batting immediately before a highly dangerous hitter in the lineup will face easier pitches to hit, because the pitcher doesn't want to walk them and give the slugger in the on-deck circle an extra guy to drive in. You could do something similar here as what I suggested for catchers, only I suspect it would work better. For hitter X, you would take the average OPS, or whatever total hitting stat we're enamored of, of all plate appearances one ahead of his, and compare it to the average OPS of those same players when hitter X wasn't batting right behind them. I suspect there would be less contamination here, although it would still be possible to say, you know, "hitters X and Y hit 3rd and 4th all year this year, and hitter X happened to have a career year. Was that because of hitter Y's protection, or something else entirely?"

Finally, speedsters. If you watch any given Mets game, say circa 2007, you would observe that pitchers really, really, really didn't like it when Jose Reyes got on base. Hell, he's so good he's made multiple pitchers balk him home when he's on third base. An elite base-stealer, especially one who is aware of this power and uses it deliberately, can distract the hell out of pitchers. You'd expect that to have a Carmelo Anthony-like effect of making the guy at the plate's numbers better. I think the methodology here would be the same: you would define situation X, perhaps plate appearances with a given runner on base with an open base in front of him, and you'd look at the OPS or whatever of hitters during those plate appearances against their average OPS when not in situation X. My intuition is that the level of difference required for it to feel significant would be lower than for the other two. After all, sure David Wright had a great year in 2007, and Jose was running like crazy that same year, but not every David Wright plate appearance was one in which Reyes was on base with a running opportunity. If it turned out that Wright did even better in '08 when Reyes was a threat, it would probably be significant, and we also might think that part of Wright's monstrous numbers that year was the monstrous number of stolen bases posted by Reyes.