It wasn't very long ago that Graham and Matthew launched StatCorner and, in so doing, made tRA available to the public. tRA, as we should all know by now, is a pitching metric designed to evaluate pitchers based on the run values of the pitch outcomes that they generate on the mound. It should be better than FIP and xFIP because it adjusts for things like grounders and line drives. It should be better than ERC because it doesn't fool around with hits and constants. It should be better than ERA because it doesn't suck. One could make a fairly convincing argument that, at least as far as the non-PITCHf/x realm is concerned, tRA (or, if you prefer, tRA*) is about as good of a pitching metric as anyone could design.
This leads to an obvious follow-up question: if tRA is so good, then shouldn't we be evaluating hitters the same way?
The answer is, yes, we should. But all too often people don't even try, because it's pretty freaking hard.
It's not a trivial thing. You can try to apply the same principles to hitters as you do for pitchers, but you'll find that you run into trouble in a hurry. It's easy to figure out the average run value of a line drive allowed by, say, John Lackey or Aaron Harang. That's because we can make the safe assumption that, over a long enough period of time, both Lackey and Harang face a ~representative sample of hitters throughout the league, so the run value of a LD will simply be (or approximate) the league average run value of a LD. Pitchers face everybody. And when you have everybody blended together, you approach the league average. Obviously there will be little variations if you have a guy in, say, a really strong or really weak offensive division, but this generally isn't a big deal. Run values of outcomes for pitchers are easy to determine because the batters that pitchers face even out over time.
It's not so for hitters. Hitters are individual. Each one will put a unique spin on every ball he puts in play. While the average line drive allowed by John Lackey will be worth ~the same as the average line drive allowed by Aaron Harang, the average line drive hit by Albert Pujols is not worth the same as the average line drive hit by Miguel Cairo. Pujols is stronger. He hits the ball harder. So his line drives will rather obviously be better than Cairo's. You can see this reflected in their career splits; Pujols' career BA on line drives is .816 with a bunch of home runs, against .725 and two for Cairo. I think this is pretty intuitive.
So you can't apply the same principles to hitters as you can for pitchers, because the whole averaging-out phenomenon doesn't take place. You can't treat a groundball hit by David Ortiz the same way you treat a groundball hit by Ichiro. It wouldn't make sense. They're clearly two different types of balls in play, and they should therefore be treated as such.
This is where we begin to understand why nothing like tRA has ever really been attempted with hitters. It hasn't been attempted because to do so, you need to (A) calculate run values on a player-by-player basis, rather than applying the average to everyone, and (B) accumulate enough of a sample size to be able to calculate those run values in the first place. Sound fun? You're a nerd. And a dreamer. Such a project would be unfathomably complicated. PrOPS tried, sort of, and because it's better than raw OPS it's fun to look at from time to time, but it leaves a lot to be desired. A lot that I'm not sure is possible to obtain.
So we're stuck. Tango's wOBA is the perfect (and I mean literally perfect) measure of what a hitter has already done in terms of results, but why should we care that much about the results? If we never talk about a pitcher's BAA, why should we have to talk about a hitter's BA? Okay, yeah, so a hitter's batting average is more meaningful than that of a pitcher, but I think the point should remain. If we're so ready to accept that a pitcher should be evaluated on the immediate results of his pitches, why aren't we the same way with hitters?
In an ideal world, hitters would be evaluated not on their BA/OBP/SLG slash lines, but rather on the balls they put in play (and, of course, the balls they don't). This is the more relevant information, right? Not all hits deserve to be hits. Not all outs deserve to be outs. The ball is out of a hitter's control the instant it leaves his bat. After all, you can't aim your line drives. So shouldn't we be judging hitters on how well they do their jobs, rather than on some combination of that, defense, and luck?
Just because we can't have the same level of accuracy as we can with tRA doesn't mean we shouldn't at least try to look at things this way. Educated estimates are better than nothing. Adrian Beltre's current batting average is .247. That sucks. But he's also hit just .622 on his line drives, against a career average of .748. Which seems more likely - that Beltre's suddenly gotten weaker than ever before in his age-29 season, to the point at which he's turning fewer line drives into base hits than Willie Ballgame, or that he's simply the victim of unsustainable bad luck? If you regress Beltre's BA on line drives to his career rate, his 2008 average jumps from .247 to .270. Right there you're talking about an OPS difference of 50-70 points. By the traditional measures, Beltre hasn't been a very good hitter so far this season, but in reality, he's actually done his job fairly well. It just hasn't worked out like he and the rest of us have hoped.
Isn't that important information? And forget about Beltre; isn't this the sort of approach we should be taking with everyone? Luck doesn't only happen to pitchers. It happens to hitters, too, and it can often have a significant effect on their results. If we're going to try and eliminate it when evaluating one, we should do the same for the other. Ultimately, what we're after is an accurate measure of how well a guy has performed, and accepting that a hitter's slash line automatically reflects how good he's been just doesn't strike me as being good enough.
HITf/x is going to be a godsend. Where currently we have to deal with the limitations of human observation when it comes to the quality of balls put in play, down the road we'll be able to look at a hit with a given trajectory and assign it a 40% chance of dropping in for a double, then look at another hit with another given trajectory and assign that one a 90% chance of being caught. It's going to open so many doors that, over time, it may bring about the death of the slash line. At least among us dorks. There are going to be real stats - BA, OBP, and SLG - and there are going to be theoretical stats - tBA, tOBP, and tSLG - that tell a more accurate story. This will represent the pinnacle of hitter analysis. Once you can assign a ball in play an accurate probability of turning into any given event, there is no further room for growth.
But just because we're not there yet doesn't mean that, given present constraints, we can't try to be as accurate as possible. Question the slash line. Ask yourself whether a guy's BA truly reflects how well he's performed at the plate. Dig deeper. Because to not do so is to do yourself a disservice, and there is no more noble endeavor than the pursuit of new knowledge.