THE BOOK--Playing The Percentages In Baseball

In the Primer thread, John says that he doesn’t do park adjustments (yet).  I think you’d want such an adjustment, ideally, though I don’t know how much a difference it would make for non-extreme parks.  Does MGL’s arm rating have a park adjustment?

That’s right: there is no park adjustment, although I realize that I need to investigate that (I, too, was surprised by Cameron’s low ranking.)

Does anybody have any input on Cameron’s positioning?  I know he plays in a big park, but my impression (very possibly incorrect) was that Cameron plays fairly shallow.  Am I wrong about that?

I got an email from a reader who made another good point about parks: artificial turf.  Rios and Cuddyer, the top-ranked RF in Hold+, both play their home games on turf.  It’s natural to postulate that runners might be less aggressive when the ball gets to the OF quicker due to the turf.

John comes back with this year’s arms:
http://www.hardballtimes.com/main/article/best-outfield-arms-of-2007/

John provides more data:

http://www.hardballtimes.com/main/blog_article/more-outfield-arms-results/

Here’s a snippet:

Best and Worst Overall, 2004-2007
---------------------------------
RF Francoeur 23.7
Rios 21.5
Cuddyer 17.9

Those three guys happen to be #3, #5, and #4 according to the Fans Scouting Report.  I’ve got them averaging 5 runs a piece a year (or 20 runs over 4 years).  Huge bingo here.

The top 2 are Delmon Young (might have questionable Tampa data) and Shane Victorino.

Dye -8.3
Giles,B -13.5
Green,S -19.7

Fans have Green as one of the worst, but the other two are around average.

CF Edmonds 16.3
Jones,A 11.7
Payton 11.5

I’ve got Edmonds and Jones at +3 runs per year (bingo) but Payton as below average.

Williams,B -10.2
Damon -15.4
Pierre -19.1

Bernie didn’t play in 07, but he always ranked near the bottom.  The bottom 2 on 07 are Pierre and Damon averaging 4.5 runs below average each.  Bingo again.

LF Soriano 20.5
Cabrera,M 10.4
Ibanez 9.1

I’ve got Cabrera and Soriano as #3, #4 arm, behind Chavez and Jenkins.  Bingo there, but I don’t think anyone anticipated Soriano’s big showing.

Ibanez’s arm is hated by Mariners fans. 

White, R -11.4
Bay -13.3
Gonzalez,L -14.9

Bay is third-worst in the Fans, Gonzalez 6th worst (all very close to the #2 guy, but behind the worst that is Podsednik).  Rondell didn’t play, but he always had one of the worst arm ratings around.

I’ve got to say that the Fans have done a superb job in evaluating Arms here.

We are lucky that Arm ratings are:
1. one of the easiest for Fans to evaluate
2. one of the easiest for play-by-play metrics to evaluate

That we have a strong agreement here shows that both sides know what they are talking about.

I’ve been looking at the agreement between my OF arms analysis and the Fans Scouting Report. 

One thing I noticed is that there is quite a bit of correlation among the various defensive categories. Obviously, some will be really be correlated, but it looks to me like there is some fan bias creeping in.

E.g., why should arm strength have a high (positive) correlation with arm accuracy? Unless the scout is giving high scores in all arm categories based on his general impression of a player having a good arm.

The high correlation among strength, accuracy and release make it impossible to use the categories separately to determine which aspect counts most towards having a good arm.

In any case, if you click on my name, you’ll see the pair-wise correlation plots of the 7 scouting categories for outfielders only.

I think there should be some correlation.  I have no doubt that there is some fan bias.  For example, if a guy has a “5” for an arm strength (on a 1-5 scale), you would expect the arm accuracy to be 3-5, not 1-5.  It’s kinda hard to be in MLB with a fantastic arm, and just leave it to the gods on arm accuracy.  It stands to reason that someone with great arm strength will work on his arm accuracy to some extent such that he’s at least average in that regard.  If he has horrible accuracy, they will simply tell him not to throw the ball so hard (arm strength 4), and that will improve his accuracy (2 or 3).

Those are really cool charts. 

I just focused on arm strength and speed, likely the two easiest things for a fan to evaluate.  It seems to support my general opinion that if you are good in one thing, you are good in something else (or fan bias to that regard).

I see that the 3 arm categories are tightly coupled, as are speed/first step and hands/instincts.