August 6, 2004
Can He Be Pitched To?
In the wake of last week's Dodgers/Marlins trade, there was a flood of Hee Seop Choi bashing. Critics, when not remarking upon the time Paul Lo Duca rushed into a burning orphanage and saved 14 kids and a three-legged dog, pounded Choi for his alleged weaknesses, completely missing his status as one of the NL's better first basemen and a player on the good side of his improvement curve.
Choi's unimpressive BA (.270) and high strikeout totals are unimportant in comparison to his power and patience. His .304 EqA puts him among the top first basemen in baseball, and while he's still learning to hit lefties, it's fair to say that he hasn't been given much of a chance against them yet: just 53 at-bats in parts of three seasons. He plays good defense at first base, too.
So why don't baseball people like him? The most common refrain, generally echoed by those who used to have jobs in the game and the space fillers who love to quote them, was that Choi can be pitched to. This is one of those beautiful baseball phrases that looks good set off by quotation marks and is difficult to contradict. What does it mean, anyway? I know that a guy with a 650 OPS can be pitched to; what does it say when a guy batting .270/.386/.495 has this particular trait?
When the criticism moved past banalities, it generally centered on two flaws. The first is that Choi can't hit good pitching, the second that, if the situation warranted it, he could be pitched to in a manner--hard stuff up and in, to be specific--that would render him useless.
I can't do much about the latter; I'm sure the folks at Tendu or teams doing proprietary pitch-by-pitch analyses could break it down a bit better, although I'd be reluctant to write off the likelihood that a player who is just now reaching a season's worth of playing time in his career will develop into a threat against all types of pitching.
The former claim, about not being able to hit good pitching, is one I can take a run at. Thanks to Keith Woolner, who generated Choi's performance against every pitcher he's ever faced, I was able to see if Choi has any kind of notable break in his record against good and bad pitchers. As it turns out, this is one of those questions that's a bit more complicated to answer than it seems, but I want to throw a the information out there, because no one is doing it.
Through Wednesday, Choi had 650 major-league plate appearances, in which he'd hit .243/.363/.452. Breaking that down by the career ERA (through Wednesday) of the pitchers he'd faced, we find:
Career ERA PA AVG OBP SLG <3.00 23 .095 .174 .238 3.00-3.99 194 .230 .355 .412 4.00-4.99 313 .258 .377 .465 >4.99 120 .255 .408 .532There's nothing much unusual about this pattern. Choi is noticeably helpless against excellent pitchers, although that dataset is largely very small samples against closers or other good relievers. Just for fun, I looked up Barry Bonds' performance against the 11 pitchers in the under 3.00 group: .254/.374/.451, which is largely based on 18 years of beating the crap out of Greg Maddux. Against the other 10 guys, all relievers, he's .201/.337/.324.
As you'd expect, Choi's performance gets better as he faces worse pitching. However, there's no break in this data that would suggest that he's hopeless against good ones and just building numbers against the league's dregs.
Choi, being a player wth just a few partial big-league seasons, has an awful lot of pitchers against whom he has three or fewer plate appearances. What if we just look at the guys he's faced at least four times, which would indicate either seeing a starter four times in one game or a reliever on four occasions? Same breakdown, but with a minimum of four PAs against a pitcher for inclusion:
Career ERA PA AVG OBP SLG <3.00 4 .000 .250 .000 3.00-3.99 120 .208 .325 .366 4.00-4.99 196 .309 .418 .568 >4.99 47 .250 .340 .500The first group is just Maddux. Man, he's good.
There does seem to be a prediliction for Choi to struggle against good pitchers. There's a big break in his record once he's facing guys with a career ERA of 4.00 or higher, and he seems to have a lot of trouble with hitting for average against better pitching. Just eyeballing the lists...actually, let me run one:
Pitcher PA AVG OBP SLG OPS PERA Brandon Webb 6 1.000 1.000 1.500 2500 3.17 Tomo Ohka 5 .250 .400 1.000 1400 3.89 Tom Glavine 8 .500 .500 .750 1250 3.41 Danny Graves 13 .300 .462 .700 1162 3.83 Russ Ortiz 9 .444 .444 .444 889 3.88 Vicente Padilla 7 .143 .143 .571 714 3.65 Wade Miller 12 .200 .333 .300 633 3.87 Roy Oswalt 6 .200 .333 .200 533 3.13 Matt Morris 10 .200 .200 .300 500 3.47 Jake Peavy 6 .167 .167 .333 500 3.80 John Reidling 4 .000 .500 .000 500 3.89 Al Levine 4 .000 .250 .000 250 3.83 Kevin Millwood 9 .000 .222 .000 222 3.88 Brian Lawrence 6 .000 .167 .000 167 3.81 Jason Schmidt 7 .000 .143 .000 143 3.89 Roger Clemens 4 .000 .000 .000 000 3.18 Ricky Stone 4 .000 .000 .000 000 3.85
The 3.00 ERA and 4.00 groups look a bit like this, where Choi's better performances tend to come against the less accomplished members of the group, or those with notable problems retiring lefties (Brandon Webb, Russ Ortiz). The Tom Glavine thing is a little weird, although Glavine has never had much of a platoon split.
Career ERA might not be the best metric here. After all, Choi never faced the good version of Jeff Fassero, and the fact that Shane Reynolds was good in 1996 doesn't inform Choi's at-bats against him in 2003 all that well. How has Choi done against the best pitchers over the past three seasons? Just pulling the 10 best pitchers against whom he has at least four PAs, we get a line of .250/.354/.412. That doesn't seem out of the ordinary, although the fact that all the pitchers in the pool are right-handed means that the gap from his normal performance is greater than you think.
(If you're curious, Choi's line against all pitchers he's seen three or fewer times is .214/.346/.406.)
I have to be honest: I came into this expecting to find no patterns that were out of the ordinary. We would expect a certain trend for players to hit inferior pitching better than good pitching. Choi hasn't played long enough for his splits to have tremendous meaning yet, and I still think his core skills are more important than everything I presented above.
That said, I think you can see in the data at least some evidence for the argument that Choi can be pitched to. Conceding the sample-size argument and the crudeness of this study--we probably need more context for Choi's performance against each subset; what is the shape of a typical hitter's performance?--I think this information supports the observation that Choi hasn't proven his mettle against good pitching.
That doesn't make it a bad trade for the Dodgers, of course, and it doesn't mean that Choi is a hopeless case. It just means that, in this case, his critics might have a point.