Can the Astros get the most out of platoon players?

he Unofficial Scorer at the Houston Chronicle, Zachary Levine, shows us how the Astros' plan to platoon Boone and Blum at third base can boost the expected offense.

Here are his results in terms of OPS:

All Blum .705

All Boone .683

Platoon .722

Levine points out: "We've taken a player who scored a .705 and by 'averaging' him with a player who scored 22 points lower, we've still raised the figure 17 points."

Unfortunately, as noted by Levine, even with this "boost" from platooning, the OPS is about 50 points lower than the positional average.  However, the result is 8 points higher than last year (despite the loss of a key player at 3d base).

This article at fangraphs.com got me to thinking about a possible lefty / right platoon at CF.  First, I will take a diversion to talk about the article.

This is one in a series of articles which attempts to demonstrate "replacement level" players at each position.  And the Astros' own Reggie Abercrombie made the list.  The article is like a backhanded compliment for Abercrombie:

This group is -12 offensively, and considering they’re mostly above average defenders, we might have our first deviation from the norm. With a +2.5 position adjustment for CF, these guys would have to be something like -10 defenders in center field to be two wins below average. But there’s no way to make that argument - Patterson, Abercrombie, Duffy, Ellison - these guys are good defenders. Using their historical UZRs, we’d conclude that they’re probably +5 with the gloves.

-12 offense, +5 defense, +2.5 position adjustment = -4.5 runs.

Half a win. The CHONE projections suggest that you could get a center fielder who is half a win worse than league average for free.

This raises an interesting question: can we get a boost by platooning Bourn and Abercrombie in CF?  Individually, Bourn and Abercrombie each is at least one half win or more below league average.  But both players have distinct platoon splits.

Based on their career OPS splits, I estimated that Bourn and Abercrombie are 6.2% and 5.6% better (than overall career stats),  when they bat versus RHP and LHP, respectively.  I used Levine's work to estimate that Bourn gets 62% of the plate appearances, with the rest to Abercrombie.   I averaged the  CHONE and  Oliver projections  (Oliver is a  more sophisitcated version of  Marcel)  to arrive at the expected non-platoon  wOBA : .302.5 for Bourn and  .303.5 for Abercrombie. I assumed that Cooper can maintain a strict platoon discipline, with each player batting no more than 5% of the time against the wrong hand. Now, let's estimate the platoon results:

wOBA= .318

This platoon results in offense -6 runs below average. Adding the +2.5 runs positional adjustment stated in the article, the net is -3.5 runs. However, assuming the same 62% / 38%  playing time split between Bourn and Abercrombie, based on UZR data, the defense produced by the tandem is +3.42 runs above average.

So, that means the total offense and defense result from the platoon is virtually equal to average production (-.07 runs below average).  I have to say, this would be quite an improvement over last year's production out of CF.  In fact, given my expectations for CF next season, i would be quite happy if you told me that CF production would be average.