5/28/10

Ranking Twins hitters based on offensive productivity

Today I am doing a slightly different exercise: I am ranking the Twins' hitters based on offensive productivity. What do I mean with "offensive productivity"? I take the old Runs Scored plus Runs Batted In measure (R+RBI) and divide it by Plate Appearances. This way hitters do get credit for both scoring and driving in runs (which are the 2 things a hitter is supposed to do) normalized by the opportunities (PAs) they had to do that. This numbers is pretty nice, because it comes to a familiar format, close to the Batting Average format that is very familiar to everyone who watches baseball. Here are the rankings (for the 2010 season, including last night's game) :

Justin Morneau: .333
Joe Mauer: .285
Jim Thome: .265
Delmon Young: .262
Jason Kubel: .250
Michael Cuddyer: .239
Denard Span: .232
Orlando Hudson: .230
Alexi Casilla: .191
J.J. Hardy: .185
Nick Punto: .173
Brendan Harris: .144
Drew Butera: .103

probably a few surprises there, mainly Span and Hardy who would appear more productive than what their numbers indicate.

6 comments:

Dez Tobin said...

What doesn't surprise me but probably surprises a lot of other people is that Delmon is ahead of Cuddy. Cuddy is the biggest guess hitter and the majors and is killing the team.

Bill said...

Beyond just punishing players for bad luck (and rewarding them for good luck) as all R and RBI-based stats do, this measure also double-counts HR (a solo HR = 1 run but gets counted as 2 here) and treats all PA as the same "opportunity." In reality, Morneau and Cuddyer regularly come up with one or two guys in scoring position, and Span almost always comes up with the bases empty (or maybe a guy on first). Of COURSE Span is going to look less productive than we'd think, because this artificially makes him look that way.

thrylos98 said...

dez: I think that it is not surprising that Delmon Young has been more productive than Michael Cuddyer this season to most people.

bill: Agreed up to a point. Yes R and RBI depend a lot on other players as well, but they are the 2 most fundamental offensive goals in baseball: score and drive in runs. Yes one could normalize more by dividing runs by OBP and RBI by total runners on base in all PAs (and that would be an interesting exercise; and, of course further normalize based on PAs). The one thing that would penalize is hitters who get on base more, like Span. No particular measurement can really stand by itself. Just another piece on the puzzle and if you look at yesterday's post (about RISP) you can probably add them together and get a better idea.

Bill said...

Hadn't seen that post yet...but Nick is right on in the comments. It's luck, and nothing you write there suggests otherwise. No hitter, given enough chances, produces RISP numbers that are consistently better or consistently worse than his overall numbers. It just fluctuates randomly.

A hitter's job is to help create runs, but all he can control to that end is his own hitting (and he doesn't even have perfect control over that -- BABIP is less luck-driven for hitters than for pitchers, but it's still got a TON to do with luck).

Once you bring runs and RBI into it, you're not looking only at how well he does his job, but also (and primarily) (a) how well the hitters *around* him do their jobs, and (b) luck.

Figuring runs adjusted by OBP tells you almost nothing about the player scoring the runs. I don't think the RBI thing tells you much either, but if you're interested, BPro already does that (here's the Twins' report if that link works -- note where Span comes in, and how few PA he's had with runners on).

thrylos98 said...

Thanks. Did not realize that BP calculated that already, but I still do not understand not counting driving yourself in as an RBI...

There was a lot of work done about hitters BABIP. The assumption is that similar to pitchers' BABIP (which should regress to some league average number from .290 to .305, depending who you are talking to), hitters' BABIP should regress to a particular number that is individual to a hitter. There was an interesting article last year (I think Hardball Times) that try to correlate BABIP with other measures to figure out what that particular magic number might be. The closer they came to (with .46 correlation factor, mind you, which is not that great) is something calculated with using about 30 different variables. Close to that (.43) was last seasons' BABIP. Hitters' skills change, some get better in a particular season due to maturation, getting close to prime etc (e.g. Span), and some get worse due to loss of skill (e.g. Thome). So I do not really believe in luck :) I do believe in pressing when things are not working for you and usually with negative results (Harris is in a funk right now, for example)...

Bill said...

It's fine to count driving yourself in as an RBI (if you think there's any reason to look at RBI at all, which I don't other than they're just kind of fun sometimes in a totally frivolous way), but your system in this post counts both runs AND RBI together. If a guy hits a HR, he gets credit for both an R and RBI for the same one run, so it makes it look like he "produced" twice as much as he did. Which is why the typical formulation of "run production" (which I obviously don't support in any way :)) is R+RBI-HR.

Of course hitters' "true" BABIP can change some from year to year. Just not from game situation to game situation. If you see differences like the ones Nick pointed out, that's luck. Cuddyer has a career .306 BABIP, and a career .321 BABIP with RISP, but a .226 so far this season. There just has to be a lot of bad luck in that.

I realize by now that you can never really convince people who are determined to believe in things like clutch performance and "pressing" that they don't exist in the way they think of them, even though no evidence of them actually exists, but let me just say this anyway: every MLB player has tens of thousands of people, at least, watching and scrutinizing his every move. If he can't handle a ton of pressure without falling apart, he doesn't stay in the league very long, assuming he makes it there at all. And I do believe that sometimes players "press" and their performance suffers -- but when/if that happens, they press ALL the time, becuase they're under intense pressure ALL the time, not just in "clutch" situations. Harris is actually a great example, because he's been terrible in all situations and every facet of the game, and maybe it IS because he's pressing (though there are any number of other, equally likely causes, and it's probably a combination of several of them). If he's 2-for-13 instead of 1-for-13 with RISP, his RISP numbers fall almost exactly in line with his terrible overall numbers.