Tuesday, August 07, 2007

The Rangers' Live-Time Experiment;
When to Try Contrarian Platooning  

Baseball is much smarter about testing the talent at hand than any other line of work in North America. It's not that performance evaluation is more important in Baseball, just that in Baseball, they accept the Truth of how critical on-going performance evaluation is. And because they know it, they do it all the time, and because they do it all the time, they are more likely to master the subtlety of doing it in the right context.

Beyond baseball when staff assignments aren't working out the way they should, work isn't getting done right or at all, managers are likely to make one, two or three of these mistakes:

  • Holding on too long to the status quo,
  • Changing too quickly without giving the staffer coaching or enough chances to succeed, or
  • Doing enough analysis to reasonably gauge where the shortfall is and how to fix it.

Once you've assessed there is a problem and that it's worth addressing, you have act, and I always recommend starting with the resources at hand.

I did work not too many years ago at a real estate management operation. I was part of a team working on an electronic document management system (EDMS), a not-universal (but not rare either) technology. There were four of us on the team, each a lead for an area. I was to be the process analyst, escaping for the first time in a long time the need to be the technology point-human. The technology point-human was a woman on loan from the IT department. The fact that I'd implemented EDMS software before was a plus, but not to be a responsibility.

But the technology lead, as intelligent as she was, didn't like to learn new things and EDMS was new to her. So we lugged for months while she refused to dig into the new realm and her responsibilities idled. Eventually, I started to try to coach her. No luck. Management whinged but took no action. So I started picking up some of her her responsibilities, and months later other people did, too. The Phase I of the project came in months late and that contributed, I'm almost sure, to a decision to suspend the subsequent Phases, a decision that assured that the sunk investment in the system would not yield net gains.

They held on too long to the status quo, and they didn't do enough analysis of the problem to realize and act on the fact that they had resources at hand to address the limitations of the status quo. A pretty common American business failure with a strong foundation of willful ignorance of what each staff member can d that's not within her current job description. 

This almost never happens in Baseball.

Case in point: The Texas Rangers' twiddling of staff assignments in mid-campaign.

The Rangers have an excess of left-handed hitting outfielders. So, according to a New York Times three-dot last week:

The Rangers plan to platoon Frank Catalanotto and Brad Wilkerson in left field even though each is a left-handed hitter. Wilkerson would start against left-handed pitchers, against whom he is hitting .262 this season (.209 against right-handers) with 6 of his 15 home runs. In his career, Wilkerson has hit .266 against left-handers and .244 against right-handers.

In concept a clever ploy, worth an experiment. Now the traditional platoon involves a right-handed batter who hits against left-handed pitchers and a left-handed hitter who bats against right-handed pitchers, because in the general case batters produce better against their opposite-handed pitchers. It's not universal, but a general tendency. For example in 2007 so far, left-handed batters are hitting about 7% better against right-handed pitchers than right handed batters are, while right-handed batters are hitting about 11% better against left-handed pitchers than left-handed batters are. Some call batters who go against that norm, "contrarian".

Split           PA       H    2B   3B   HR  BB    IBB   SO      BA   OBP   SLG   OPS  tOPS+ Split
 vs RHP as RH   49605  11592 2354  163 1251 3485  197   8786   .259  .318  .403  .722   92 vs RHP as RH 
 vs RHP as LH   43107  10259 2116  287 1093 4283  492   6859   .270  .346  .427  .774  106 vs RHP as LH 
 vs LHP as RH   26764   6577 1403  109  743 2412  201   4128   .277  .347  .440  .787  109 vs LHP as RH 
 vs LHP as LH    9694   2126  413   45  204  844   28   1993   .249  .321  .379  .700   87 vs LHP as LH 
stats from Baseball-Reference.Com

Why was this a clever thing to experiment with? Several reasons.

The Rangers are in last place and are fairly-well out of the race for a playoff spot.

Team W L PCT GB E# L10
Detroit 62 49 .559 - - 2-8
New York 62 50 .554 0.5 51 7-3
Seattle 60 49 .550 1.0 52 6-4
Minnesota 57 54 .514 5.0 47 6-4
Toronto 56 55 .505 6.0 46 5-5
Oakland 54 59 .478 9.0 42 5-5
Baltimore 52 58 .473 9.5 43 6-4
Chicago 52 59 .468 10.0 42 7-3
Kansas City 48 62 .436 13.5 39 5-5
Texas 48 64 .429 14.5 37 2-8
Tampa Bay 42 69 .378 20.0 32 4-6

1. They have less to lose from an experiment that doesn't work out than a team that's in contention.

Baseball works with this model all the time -- beyond baseball though, management rarely learns this lesson. The real estate management endeavor had a project that slipped out of meeting deadline, but they still lashed themselves to the ineffectual technical lead and her lassitude like Ahab to the Whale. Trying to use others' talents, job description included or no, would have made perfect sense...and did when it finally, too late, happened.

2. Brad Wilkerson might be able to hit a predominant diet of left-handed pitchers better than right.

The evidence is inconclusive. In his peak years before his injury last season, Wilkerson contrarily had equal or better production against left-handed pitchers than right-handed ones (stats again from Baseball-Reference) as often as not. Catalanotto had already proven in his long career that he was unlikely to hit lefties -- in his six peak years, he never once got close to being good against left-handed pitchers.

The evidence is "good enough" to experiment with. Beyond baseball, an organization that took the trouble to engage in close assessment of the talent at hand would make a point of knowing who else on staff might be able to complement or fill in for a struggling team member. For the Rangers it meant looking closely at data and then overcoming a standard practice and the assumptions around it, and any organization can do that if it cares about success.

3. Wilkerson certainly isn't hitting right-handed pitchers, so if they're going to get any utility out of his bat, they have every reason to hope it could come against lefties.

It's close to impossible to justify Wilkerson's .213/.285/.427 output against the right-handed pitchers The Book says he should hit. In the real estate management outfit case, they should have taken the Rangers' approach and found something the woman holding the technology lead's role could have done to add to the team's progress. She was bright and promising -- they could have taken a chance on finding out what she could contribute. Instead, they just resented her lack of production at what they hoped she would achieve.

Baseball teams sometimes do what the real estate management outfit did, but it's a remote exception.

Teams are willing to try players in different positions, take multiple training approaches, bring in new coaches, twiddle with their mechanics or mental attitude or the pattern with which management applies them in their work.

Who do you have dragging you down who could use a redefined workload or a job swap or a platoon partner? ┬┐What are you waiting for?

NOTE: In the end, the Rangers didn't execute on this experiment, at least not yet. They made a trade at 1st base right after this story appeared, and faced one left-handed starter since...and Wilkerson was not in the line-up at all; a rookie, Jason "Gigantor" Botts got penciled in at LF.

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

free website counter