THE BOOK--Playing The Percentages In Baseball

What about LF/RF for the Red Sox? Isn’t Crawford the better fielder than Drew?

StL plays a weak hitting OF at 2B ... and Freese is actually pretty decent in the field (and at the plate).

We had this discussion at FG a little bit ago regarding 3B and 2B.

3B that can hit well seem to get moved to another position rather quickly. Either a corner OF spot or 1B.

Good hitting 2B’s seem to stay put. Right now 2B is a big difference between the elite teams and everyone else. Pedroia, Kinsler, Cano, Utley, etc.

Be interesting to go back and look at the deep 3B pool from years back and see what they’re doing now.

Miguel Cabrera and Albert Pujols moved across the diamond. Throw in the demise of Eric Chavez and that’s a pretty big hit for the position to take.

Tom/1: I said “strongly prefer”.  If it wasn’t a strong preference, then I called it a draw.

3B that can hit well seem to get moved to another position rather quickly. Either a corner OF spot or 1B.

Zimmerman, Langoria, Rodriguez, Sandoval, Beltre, Youkilis, Wright, Chipper?

I took him to mean: disproportionately more than 2B.

Though, I don’t know if that’s true (Soriano, Prada, etc).

These are all just off-the-cuff theories that require evidence…

"Never do what others have done in the past, and that is to “zero out” the stats such that the average LF (offense + defense) is considered equal to the average CF or average RF.  This is clearly the wrong thing to do.”

So does this mean that B-R is calculating their WAR incorrectly? Does anyone have a/the link that explains B-R’s positional adjustment in more depth?

http://www.baseball-reference.com/blog/archives/8225#comment-48965

When Sean says that that explanation is a crude simplification, I think he’s just trying to give the general concept behind position adjustments in a quick and easily-digestible way rather than explaining how they are actually calculated.

Rally talked about position adjustments on FanGraphs a couple years ago:  http://www.fangraphs.com/blogs/index.php/historical-position-adjustments/

I would assume B-R is still using those types of adjustments since they are using Rally’s WAR.

I would say that through 2009, which is Rally’s last update, Rally follows the spirit of the positional adjustments as I’ve described them in the past.  That is, not to rely solely on the offensive performance of that season.

What Forman has done since then I have not yet confirmed.  I did bring up to B-R.com that the presentation for relievers is obviously differently than the way Rally presents it (even if they both end up at the same answer).

I will also presume that Rally had done some manual insights into the positional adjustments, so that he doesn’t rely on just some algorithm 100%.  (You’ll always get some head scratcher that requires baseball intelligence to tweak.)

Whether Forman also does that is an unknown to me, and I’m not going to speculate without evidence.

Checking really quickly for 2011, using Play-Index.  For the 63 players that come back as LF, it’s a total of 19008 PA, and 47 WAR.

For the 52 CF, it’s a total of 19208 PA and 80 WAR.

So, that’s great evidence that the positional adjustments are not being zeroed-out.

Is 33 WAR about right?  Let’s see…

The gap in hitting (in 2011) is about 0.1 wins higher for CF than LF.

The gap in fielding is set at about 1 win higher for CF than LF.

That’s about 1.1 wins higher per CF over LF per team.  Times 30 teams and it’s 33 WAR.

Bingo!  Looks like Forman’s got the calculation in-line with expectations.

Might as well check BPro.

The positional adjustment for LF is set at around -4.5 to -5.0 per 700 PA.

The positional adjustment for RF is set at around -7.0 per 700 PA.

I don’t agree with this at all.  My standard adjustment is to make it -7.5 per 700 PA for both LF and RF.  And, if you want to tweak it annually, go ahead.  In 2011, it seems to me that your better fielders are in RF not LF.  So, if anything, you’d want to tweak the LF going the other way over to -10 or something.

As for CF, it’s set at +1.7 per 700 PA.

To have the gap between CF and LF on the fielding side to be only 6.5 runs per 700 PA is just much too tight for my liking.

Having it 8.7 runs between CF and RF is ok.

***

3B is set at +2.6, and 2B at -0.9.

As I said, I don’t think that’s the right thing to do.  I think they should be swapped.

1B is set at -12.1, which is good.

C is set at +11.4 which is good.

SS is at +8.7 which is good.

***

So, in my opinion, BPro should look harder at the positional adjustments for 2011, and recognize that, probably, the fielders at 2B are better than at 3B and the fielders in RF are better than in LF.

But, then again, they don’t really do the adjustments exactly like I do it.  So, this issue may be for more than just 2011.