A's rating of Karros a bit too high

The A's have sound reasoning in their signing of Eric Karros, but they may be disappointed with him in the end.

Originally Published: February 6, 2004
By Rob Neyer | ESPN.com

It makes all the sense in the world, the A's signing Eric Karros for what counts as chump change in the fantasy world that is professional sports. After all, look what Karros and his new teammate Scott Hatteberg did last season, and over the last three seasons, against left-handed pitchers:

OPS vs LHP   Hatteberg Karros
      2003      713     986
 2001-2003      687     904

First thing's first: the problem with Scott Hatteberg isn't that he can't hit left-handed pitching; the problem with Hatteberg is that he can't hit. Even against right-handed pitchers, Hatteberg's one- (729) and three-year (762) OPS's aren't anything thrilling for a first baseman.

But let's assume that Hatteberg's a 762 OPSer against right-handed pitching. If 1) he's as good with the glove as Oakland's management claims, and 2) his willingness to run deep pitch counts is as valuable as Oakland's management says it is ... well, I suppose you can live with him. All that remains is to find him a platoon partner, and that's just what the A's did when they picked up Karros, who has "mashed" left-handers, especially last season. Signing Karros makes so much sense, so much obvious sense, that everybody is saying and writing the same things ...

One of our weaknesses, going in offensively, was that we were starting to get a little left-hand heavy again. And one of the things Eric has been very good at recently is hitting against left-handed pitching. At a minimum, I see Eric playing quite a bit against left-handed pitching. It gives us another weapon.

    -- Billy Beane in the Sacramento Bee

The A's are looking forward to Karros' right-handed bat balancing out their lineup. Karros' numbers against lefties the last three years -- .316 average, .389 on-base and .515 slugging -- should make the A's less vulnerable in that area.

    -- Josh Suchon in the Oakland Tribune

We were starting to get left-handed heavy again, and ... Eric has been very, very good against lefties. Historically, that's something we've struggled with.

    -- Billy Beane in the Contra Costa Times

While in a first-base platoon last year in Chicago, (Karros) hit .366 against lefties, with a .441 on-base percentage. Beane would not declare first base a platoon situation with the arrival of Karros, who also could share time at DH with lefty-hitting Erubiel Durazo ... In fact, Karros might challenge Hatteberg for the everyday job, after a season in which Hatteberg, playing injured much of the time, hit .253 overall.

    -- Susan Slusser in the San Francisco Chronicle

I'm presenting all those quotes, from both Beane and a pair of knowledgeable local writers, because I want to make it clear just how Conventional this Wisdom is, this Wisdom being that Eric Karros rakes southpaws. After all -- and kudos to Josh Suchon for citing three years rather than just one -- Karros has been potent against lefties for a while. Three seasons ... that can't be just a fluke, can it? Karros may not be a great hitter, but one has to admit that he's been very good against left-handed pitchers. And it follows, doesn't it, that he will continue to be very good against left-handed pitchers?

Eric Karros
AP Photo/M. Spencer GreenEric Karros, right, batted .366 in 112 at-bats against left-handed pitchers last season.

All of this rests on the reasonable assumption that players innately have different platoon splits; that is, some players have small platoon splits, some players have large platoon splits, and these varying degrees are the result of ability. Like I said, that's a reasonable assumption, and Conventional Wisdom to boot. I certainly believed it.

That assumption, however, is probably incorrect.

As Bill James pointed out in a seminal study of left/right splits, back in 1988, and as frequent Baseball Primer contributor MGL pointed out earlier this week in this thread, there's a lot more to these numbers than what happened last season, or even what happened over X number of seasons. MGL gets into things like "correlation coefficient" and "y-t-y correlation," neither of which I would explain in this space even if I perfectly understood them, but let me make a small start toward explaining the issue ...

Let us suppose we've got a player -- Nathan, we'll call him (because we need a Nathan in the majors, don't you think?) -- who, last season, batted .250 in 450 at-bats against right-handed pitchers, and .350 in 100 bats against left-handed pitchers. Now, would you assume that Nathan will again bat .350 against lefties next season?

No, you probably would not. You probably would say, "Yes, .350 is impressive. But that was only 100 at-bats, so we also need to consider that he hit only .250 the rest of the time. It's not immediately apparent how he'll do against lefties next season, but I'm pretty sure he won't bat .350 against them again."

You'd be right ... and it's exactly the same with Karros. Well, not exactly the same. Instead of 100 at-bats, we're talking about 307 at-bats for Karros against lefties, over the last three seasons. But in those same three seasons, his OPS against right-handed pitchers, over the course of nearly 1,000 at-bats, is a miserable (for a first baseman) 672.

That 672, we're supposed to just ignore? No, because that 672 also represents Karros' abilities, and not just against right-handed pitching. His 904 OPS against lefties and his 672 against righties are both pieces of the same puzzle, and if you separate one from the other, you'll wind up with an unfinished puzzle. Those 307 at-bats are not, and the 112 at-bats last season are certainly not, enough at-bats to prove that Karros is a special sort of creature with a crazy-big platoon split.

In fact, if every player played enough games -- thousands and thousands of games, I mean -- eventually all of them would have roughly the same platoon split. There is some evidence that some types of hitters will have slightly larger platoon splits than others, but essentially they're all the same. I know, it sounds crazy. But everyone who's looked at this with any degree of sophistication has come up with the same answer. As James wrote in 1988, "It's innate. You can't get away from it."

The A's are better with Karros than without him. But that's a testament more to Hatteberg's ineffectiveness than to Karros' effectiveness. Karros did mash lefties last year, but that doesn't mean he's going to mash lefties this year. What's most likely is that he'll revert to numbers against lefties that represent his true ability. And that means an OPS against southpaws in the 750-800 range. Worth a million dollars? Perhaps. But not much more.

Senior writer Rob Neyer writes three columns per week during baseball's offseason. This spring, Fireside will publish Rob's next book, "The Neyer/James Guide to Pitchers" (co-authored with Bill James); for more information, visit Rob's Web site. Also, click here to send a question for possible use on ESPNEWS.

ALSO SEE