Monday, March 07, 2005

Can Choi Keep Up his .819 OBP against Southpaws?

Before I continue, a typo that had me laughing too hard to ignore it... From the latest Gurnick mailbag:

According to general manager Paul DePodesta, if Choi winds up with statistics similar to his season-ending totals of last year -- .251 batting average, .819 on-base percentage, 15 home runs and 46 RBIs -- that would be sufficient.

Did he really have a .819 on-base percentage? That would suffice in my book. A .251/.819/.251 season would be MVP caliber.

In any event, I wanted to take a look at a question which has come to th fore a bit more recently. Should Hee Seop Choi start on days when the Dodgers face a southpaw?

We can assume that Jose Valentin is not going to start on days when a left-handed pitcher is throwing, so that means that one of the APerez/Nakamura/Saenz trio would be starting anyway. That means the potential platoon mate would be the second-best of that group.

To get a quick and dirty idea of what kind of impact we're talking about here, I'm going to use the average projection from PECOTA, ZiPS, and Marcel that I found a few weeks back. In the quick and dirty spirit, I'll use each player's Gross Production Average. Choi's is .281, Saenz' is .265, Perez' is .248, and Nakamura's is .250.

To project each player's performance against southpaws, we can use the universal platoon ratio, which, given the sample size for each individual player, is normally a better predictor of future platoon ratio than past platoon ratio. Left-handed batters hit about 1.17 times better against right-handed pitchers and right-handed batters hit about 1.09 times better against left-handed pitchers. We can reverse engineer our general projections based on how many plate appearances against each type of pitcher they had in the past, since each player's projection is based on seasons where they had a certain number of PA against LHP. I don't have the actual numbers for Nakamura or Perez, but I'll assume that since they've been everyday players they've had the normal ratio of about 28% PA against LHP, although ideally I would find the actual data for Japan and AAA... Anyway, here's the table, with plate appearances being from 2002-2004:

RHP PA LHP PA %LHP ProjGPA PlatoonRatio ProjLHP ProjRHP
Choi 647 71 10 .281 0.85 .242 .285
Saenz 160 146 48 .265 1.09 .277 .254
Perez 28 .248 1.09 .264 .242
Nakamura 28 .250 1.09 .266 .244
Using this method, Saenz comes out on top, followed by a virtual tie between Perez and Nakamura. However, Saenz is the one least likely to play third base, so we can figure that it will be one of the others playing in Valentin's stead. So let's compare Saenz to Choi. Saenz' has a .035 edge in GPA using these numbers. A rough run estimator for GPA is Runs = (GPA_1 - GPA_2)*1.2*(plate appearances). 698 starts were made in the NL last season, or 27% of the 2590 games played. The average starter lasts 27 batters, so we're talking about three plate appearances a game in 27% of the Dodgers 162 games. 3 PA times 44 games equals 132 plate appearances at stake. So 132 PA * 1.2 * .035 GPA difference equals five and a half runs, or half of one win.

So, the difference between Choi and Saenz starting, offensively, is half a win. I don't think there's too big a spread in their defensive talent, so let's assume it's simply a question of offense.

Now, there's another question at stake here: will playing regularly against left-handed pitching improve Choi's long-term ability, both in general and against left-handed pitching? The answer is almost certainly yes, but the magnitude of the impact is a different question. On the one hand, it's not a big deal to carry around a right-handed backup and platoon partner at first base. That's perhaps the most easily replaceable role in baseball. On the other hand, having to often remove your first baseman in the sixth or seventh inning for a pinch-hitter creates a number of strategic limitations, so it's certainly preferable to have your premier left-handed hitters have better than average platoon ratios. And while keeping a decent right-handed first baseman on the roster is easy and fairly inevitable, it's certainly better to not have to depend on one and to not have to spend a little extra for a very good one. And some of the value added from Choi getting experience against southpaws will be this season when he has to face left-handed relievers and it's not strategic or possible to remove Choi from the game.

I would argue that Choi should probably be the default starters against left-handed pitchers but that Saenz should get the nod in high-leverage games. That is, when the Dodgers are playing a clear rival for a playoff spot or are in a tight spot late in the season - in other words, when wins are at a premium - Saenz should face the southpaw.

With regards to pinch-hitting, the same philosophy should apply. Eighth or ninth inning of a close game and a lefty's on the mound? Bring in Saenz. Three-run lead in the seventh? Let Choi hit. Also, the possibility of the next time through the order has to be factored in. Saenz has a slight edge over Choi if the task is to one PA against a southpaw and one against a righty. However, if there's reason to believe that there will be another trip through the order and that Choi's spot in the lineup will occasion a higher leverage plate appearance the next time, then it makes more sense to leave Choi in the game.

I'm going to use your method of combining then averaging the PECOTA, ZIPS, AND MARCEL systems to create cards for 2005 rookies (including new guys from Japan) and other uncarded players (e.g., Aaron Boone: sat out 2004)in order to play out a 2005 season in Strat-O-Matic with up-to-date 2005 rosters. I've also looked at the Roto Times projections (what do you think of them?), but they don't provide as many numbers as the other systems, so I wonder if I should stick to only the three. In any event, I'm glad you posted the universal platoon formula and the average lefty/righty differences in this post -- as lefty/righty splits play a significant part in the card ratings in Strat-O-Matic.
I don't have statistical proof, but I think there might be a bigger difference between the quality of Choi and Saenz's defense.

Olmedo looks pretty bad anywhere on the field, except the batter's box. I'd feel more comfortable with Choi over there most of the time.
Hi there Blogger, a real useful blog.Keep with the good work.
If you have a moment, please visit my average salary of physical therapist site.
I send you warm regards and wishes of continued success.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?