THE BOOK--Playing The Percentages In Baseball

This is no surprise.  I think 2B is the least paid of all the positions as well.

Last year I found that offensive replacement level at 2nd and 3rd is about equal, even though the average hitter at 3rd is better.

This is no surprise to people who don’t try to force a round ball into a square peg.

It’s a shock to the statheads (basically, almost all of them) whose metrics force the average player at each position to be equivalent to one another.

Our job is to model the reality of baseball, and not make one up because it’s easier.

That argument assumes that any player can play any position without loss of fielding skill, correct? 

Because, if not, you can’t say things like “the average SS is a better fielder than the average 2B”.

That argument assumes that any player can play any position without loss of fielding skill, correct?

No, it assumes that any shortstop can play second base without any loss of fielding skill.  And since they require the same skillset, but the better athletes get selected to play shortstop earlier in their careers, this is pretty obviously true in this example.

Tango also spoke about 3B and 2B, so it must also assume that fielding skill is transferable there, too.

I’m more inclined to believe that you can easily move an SS to 2B (or vice-versa) than the 3B case.

The larger point is: to what extent do the skills transfer?  100%?  95%?  90%?  It makes a difference, obviously.

I’ve talked about this in the past.  The transferability of skills is based on pools, so you have 5 pools of nonpitchers:
pool1: C
pool2: 2b/ss/3b plus pool1
pool3: lf/cf/rf, plus pool2
pool4: 1b, plus pool3
pool5: dh, plus pool4

There is a familiarity factor within each class (if, of), and a larger familiarity factor between classes.  But that factor gets reduced the more you play at the other position.  Something like 4 runs for in-class, and 8 runs for out-of-class.

A 2b and 3b are a subset of SS, and lf/rf are a subset of CF.  (i.e., ss,cf can play any of the other positions with a smaller familiarity factor)

Phil:  We had an interesting (if long) thread on this a while back.  A real Tango/MGL cage match:  http://www.insidethebook.com/ee/index.php/site/comments/how_should_teams_make_personnel_dicisions/#comments

* *

I’m not sure which metrics Tango is criticizing.  I do think it’s valid to measure players against a position-specific definition of replacment level.  And the average value of each position is the difference between the average player and replacement at that position—by definition.

If you use Woolner’s old data, that indicated replacement was 80% at most positions, but 75% at 1B and 85% at catcher (IIRC).  So that would mean 1B had the most value by a good amount, catcher the least, and the other positions would vary according to their mean run production.  However, those multi-year averages certainly don’t have to mean the 80% relationship applies to every position every season.  You might have a very strong group at a position—such as 3B right now—but there’s no reason to think 3B replacement level has changed.  Same for SSs.  So I have no problem accepting the idea that the current 3B and SS crops are strong.  And that would likely mean 2B is the weakest position, since replacment offense at 2B is higher than at SS (or C). 

However, I would expect a rough equilibrium to be reestablished at some point.  SS won’t remain offensively equal to 2B indefinitely, if they are better fielders.  Instead, some SSs now sitting on the bench will become starting 2B.  And some backup 3B as well, if Tango is right about their ability to convert to 2B.

Indeed, 46 pages of cage match!  Thanks, I’ll get myself up to date.