Moving Manny

Though I had some harsh words for White Sox management earlier this week over the Jim Thome/Manny Ramirez situation (said words seem to be mostly aimed at Ozzie Guillen, the brainchild of the rotating DH, but I still hold the GM ultimately responsible for giving such deference to his manager), I nonetheless noted that ” [w]ithout question, the Manny move is an upgrade for the White Sox.”

The question I seek to answer today is how much value Manny should prospectively provide the Pale Hose. Dave Cameron saw the difference between Manny and Mark Kotsay as worth about 8.0 runs per about 100 plate appearances. What I am going to do is take Manny’s numbers at Dodgers Stadium and attempt to translate them over into the AL for the White Sox.

As a preemptive disclaimer (given past experiences), these translations tend to be more theoretical than actual, forming a baseline around which statistical noise and random luck oscillate around. Over a meager 30 game sample, anything is possible.

For the Dodgers, Manny hit .312/.407/.508 (.915 OPS) with 15 doubles, no triples and 8 homers over 196 AB/232 PA. His NL BABIP of .348 is .028 points ahead of what his expected BABIP (xBABIP) is, a mark of .320, given his batted ball profile playing for Los Angeles. This xBABIP-BABIP split represents a difference of four hits for Manny. If we adjust Manny’s 2010 triple slash line for the Dodgers, optimistically assuming all four subtracted hits would have only been of the singles variety, Manny’s theoretical luck-neutral line for the Dodgers “falls” to .291/.390/.490 (.880 OPS). That line would represent 57.0 hits over 196 AB, distributed as 34.0 singles, 15.0 doubles and 8.0 home runs.

Now that we have a neutralized Dodgers line to work with, we need to translate Manny’s triple-slash components from the Dodgers to the White Sox. First, we need to translate the BABIP, again assuming that all hits added/subtracted would have only been of the singles variety. Next, we will need to translate the walk and hit-by-pitch
rate between parks (case studies, such as this one, shows that certain factors, such as temperature and humidity, affect control rates between ballparks). Finally, we will need to translate the hits-type rates between ballparks.

According to THT’s “top secret” park factor data (henceforth referred to generically as “park factor data”), Dodgers Stadium has a four-year BABIP index of 1.00265724. Likewise, U.S. Cellular Field has a BABIP index of 0.97111976. That may seem strange at first glance, given that Dodgers Stadium is more of a pitcher’s park (-10% effect on runs production, per Baseball-Reference) and U.S. Cellular Field is more of a hitter’s park (+8.0% effect on runs production, per Baseball-Reference). However, a glance at the park factor data indicates that the dimensions of U.S. Cellular Field only tend to exaggerate home run (+21.3% HR/FB% index) production at the expense of all other hit outcomes (-0.7% 1B index, -5.6% 2B index, -18.3% 3B index). U.S. Cellular Field’s flyball oriented park dimensions come paired with higher-than-average flyball and pop-up rates, two ball-in-play (BIP) types with low expected hit outcomes (xH). Dodgers Stadium, meanwhile tend to yield more groundballs (.237 xH) and less pop-ups (.008 xH)).

Taking Manny’s luck-neutral xBABIP of .320 for the Dodgers and multiplying it by one-half of the U.S. Cellular Field BABIP index and dividing it by one-half of the Dodgers Stadium index reveals a translated xBABIP of .315. This additional .005 point drop in BABIP would result in the loss of another hit from Manny’s season (again, keeping the PA rate constant). Optimistically assuming this subtracted hit would have only been a single, Manny’s expected triple-slash line for the White Sox only takes a slight tumble: .287/.384/.486 (.870 OPS).

Park factors also reveal that a move from Dodgers Stadium to U.S. Cellular Field would have a favorable impact on Manny’s already elite on-base percentage. According to the park factor data, the half-country move would result in a +2.51% impact on a hitter’s walk total while simultaneously making that hitter 12.45% more likely to be hit by a pitch. Applying these numbers to Manny’s walk totals (32) and hit-by-pitch totals (1) with the Dodgers gives us a new, translated total of 32.8 and 1.1. Applying these numbers to Manny’s AB total and OBP, we get a new adjusted triple-slash line of .287/.387/.486 (.873 OPS) — not too much of a difference with on-base factors considered, but a slight OBP bump nonetheless.

Finally, in a translation from Los Angeles to Chicago, we need to adjust Manny’s prospective power output to account for park factor differences. For the Dodgers, Manny posted a HR/FB% of 14.0% (57 FB) over 236 PA. Technically, the statistically significant HR/FB rate threshold for hitters is 300 PA, but for the sake of simplicity, let’s pretend that the 14.0% mark represents what Manny is capable of on the season. Before we translate the home run numbers, we must consider weighted play time effects. A hitter only plays one-half of his games at home, assuming he plays all 162 games his team plays in a season. Manny was no exception in Los Angeles, accruing 50.5% of his total at-bats at home. For the sake of simplicity, I am going to weight the park factors for both parks by a 1/2 step.

A move from Dodgers Stadium to U.S. Cellular Field (PF_USCF/PF_DS) would first have the following impacts on Manny’s numbers:
*No impact on Manny’s flyball rate
*A 0.978029562 singles index (0.989014781 impact)
*A 0.964803037 doubles index (0.982401519 impact)
*A 1.031384918 triples index (1.015692459 impact)
*A 1.19796759 HR/FB% index (1.098983795 impact)

Applying the half-step of these indicies, as indicated in parenthesis above, to Manny’s adjusted numbers, we find that over a 232 PA/195 AB (recall, the AB total had to be modified slightly to reflect the minor change in walk rates and HBP rates between parks) Manny would be expected to produce 32.8 singles, 14.7 doubles, 0.0 triples, and 8.8 home runs.

All of the above considered, Manny would have an expected triple-slash line of .289/.389/.499 (.888 OPS) playing for the White Sox.

Those numbers are quite good, albeit a down-step from Manny’s career OPS of exactly 1.000, and represent a substantial upgrade over Mark Kotsay‘s line of .237/.311/.385 (.695 OPS) for the White Sox. Manny should post around 100 or so PA for the White Sox in his brief tenure on the South Side of Chicago, so cutting Manny’s total expected numbers, above, in half should yield a reasonable forecast of what White Sox fans should expect in September (in terms of absolutes): about 4-5 home runs, 16-17 walks, 56-57 hits and plenty of Manny-being-Manny moments.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: Pitchers nowadays are on speed
Next: The good face, the halo, and projectability »


  1. MikeS said...

    How can a park affect HBP numbers?  Even BB are a stretch but at least that is a more common event and easier to evaluate.

    If you have a pitching coach that really pushes pitching inside or you have a guy who gets hit a lot *cough* CarlosQuentin *Cough* that’s going to skew the numbers of the park they play in.  I just don’t see how the park can influence these events in any kind of way.

    It’s not enough to say that “more batters get hit at stadium x than anywhere else” if you don’t know why.  Otherwise it could just be a statistical outlier.

  2. Jeffrey Gross said...


    I tended to agree with you before I started working at THT, but I’ve been sent a litany of reading material since showing how humidity, for example, affects the movement and location of pitches. Hence, temperature impacts accuracy and BB/HBP rates. The numbers overall are largely negligible.

    Quentin probably skews the numbers for US Cellular Field some, but the parks with the most BB/HBP increasing are The Ballpark at Arlington, US Cellular Field, Minute Maid park, Pet Co and Turner field.

    On the opposite end is Fenway.

    Four year BB indicies range from .970 (the Metrodome, now defunct) to 1.027 (Miller park), so those numbers are pretty stable. It’s HBP that tends to vary more: from 0.865 (Miller park) to

    I honestly tend to agree that HBP is probably more noise than not, but I might as well incorporate them until proven otherwise. Besides, the impact of even a 10-15% swing on what are always only a few HBP (quentin, one of the most extreme examples, has only about 15-20 per year) is negligible, I suppose. Manny’s HBP prediction increased by 0.1. I’d have to say that’s not enough of an impact for me to be “take it safe” and ignore park/weather effects on movement as it applies to hit batsmen.

    Again, agree with you that HBP Park Factors seem bogus at first glance, but there is data to back up that weather does have some impact.

  3. nathaniel dawson said...

    Does this factor in the switch from the NL to the AL? You’re jiggering his numbers using a lot of things that impart a really small effect, but switching between leagues could actually have a relatively large effect, no?

  4. kds said...


    I like what you did with component park factors; to often authors just look at run factors, as if Coors affected Juan Pierre the same as Dante Bichette.  Handedness park factors may be relevant here.

    However, I see no adjusted for league differences, which may be of larger magnitude than some of the park factors.

    Also, there should be regression, especially with the small sample size.  (But the mean toward which to regress would be Manny’s estimated current ability.)

  5. Jeffrey Gross said...


    Agree that handedness should play a factor here, but I do not have detailed handedness PF data at my disposal.

    With respect to the AL>NL change, I only view that relevant as it pertains to pitchers. Yankees excepted, there is really no reason to expect the quality of the pitchers to vary league to league other in that they face no pitcher in the AL. From a hitter’s perspective, league changes have much less of an impact on the bottom line. Whether, for example, Jake Westbrook pitches in the AL or NL will affect Jake Westbrook’s overall numbers, but if he’s facing Russell Branyan in the AL or Russell Branyan in the NL really has no impact on Russell Branyan.

    For pitchers, though, you need to at least make a 0.4 adjustment to the K/9, though.

  6. Jeff said...

    All I know is he hasn’t hit one ball hard yet that I have seen.  He has been getting on base, which is good, but I thought we wanted him for his power, which appears to be non-existent up to now.  I’m still waiting Manny.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>