THE BOOK--Playing The Percentages In Baseball

How do Fans see these players?

http://www.tangotiger.net/scouting/scoutResults2005.html

Among the top UZR fielders, Erstad, Rowand, Beltran, Sizemore, Jenkins, get high marks overall.

Logan, Crisp, Matos, DeJesus, get high marks for at least two of speed, first step, instincts.

So, no surprises.

Among the bottom of the barrel:
Ward, Pena, Ramirez, Mondesi, Giambi, Buchanan, get low marks overall according to the Fans.

Wilson gets low marks for speed, first step.

Mackowiak is average according to fans.

And then there’s Griffey.  Fans are probably in severe denial.

Out of the 18 players listed (sans Albert Belle), Fans were surprised with two of the results (Mackowiak, and definitely Junior). 

I like the Bill James rule that says that any stat that gives you 85% of what you expected is a sign of a good stat.

Tango,

Thanks for the post.  Has MGL done a comparison with The Fielding Bible results for 2003-2005? That would be very interesting.

I don’t think he has, but perhaps I will.

This is great.

1.  Is that rating for Placido Polanco for 2B, 3B, or both?

2.  Do we have any earlier data that says Roberto Alomar was good in the field?  If he’s worth -1.5 wins per season on the field, that takes a huge bite out of his HOF case.

It includes all three infield position, scaled so that the average SS is +2, while the average 2B,3B are -1.

As for Alomar, that’s all MGL did for UZR, but STATS used to do it as well, and Alomar never looked good by the numbers, in direct contrast with the visuals.  Same for Junior.

Tangtiger:

These recent posts have been great. Thanks.

I’m wondering if the data sheds any light on the rate of players’ defensive decline. Is there an age where a prolonged period of decline begins to fall off the cliff?

On the flip side, are improvements made in steady increments or can a player make a sudden jump from defensive mediocrity to excellence?

I apologize in advance if such questions have already been addressed elsewhere, but finding a central location for defensive discussions on the web has so far eluded me.

Thanks again.

Thanks Kent.

I did do age adjustments for fielding when MGL did the 99-03 UZR.  Now that he has redone his UZR, for 00-05, I have to redo the aging patterns for fielding. 

In a nutshell, we’ve already been able to establish with SB and 3B that speed declines very fast.  CF and SS, which rely on speed the most, had shown an earlier peak than 1B. 

Once I redo the aging study for fielding, I expect to find that peak for fielding will top off at around age 26.

Thanks for the response.

That all makes perfect intuitive sense.

Have you, in general, found a player’s decline from his defensive peak to be gradual, or does it frequently collapse?

For example, if a player’s three-year UZR tracks as 15/13/11, does that suggest that his next year’s might settle in at 9ish, or is there a point where a player’s defensive abilities have eroded so consistently for so long, that at year 4 (or whatever) you’re just as likely to see his UZR drop off to 3 (or something).

Thanks again. Sorry about hogging the comments. This is just all very new and interesting to me.

Click my name for some great home/road breakdowns on Manny.

In a thread on Manny (which usually brings out all the trolls) on BBTF, I gabe Manny’s road numbers the last few years.  Here they are, per 150 games:

07 -15
06 -28
05 -11
04 -15

That is still a REALLY bad left fielder folks.  Like worst outfielder in baseball, not named Ken Griffey, bad. Not as bad as his UZR numbers in Fenway.  I am willing to admit that I don’t trust the LF UZR numbers in Fenway all that much.

That is about 500 chances, which gets regressed about 35% toward the mean.  The mean of a low, lazy, overweight 30 something OF’er is probably close to -10.  So you have -17 regressed toward -10, or around -14 to -15.  That is really bad and about what I have been calling him for years (a -15 LF’er).

From now on, when I quote Manny’s UZR, I am only going to give his road numbers.  I am so tired of people telling me that his poor UZR numbers are only because of left field in Fenway.  That otherwise he is not that bad.

Thanks, MGL.  He’s a really bad outfielder, but by those numbers he’s not much different than a lot of leftfielders, like Carlos Lee, Josh Willingham, Pat Burrell, Raul Ibanez, and Adam Dunn.  I’ll bet all of them are worse than -10 but not as bad as -20.

When people see things like a -50 in front of Manny’s name, I don’t think they are acting as trolls if they stop and question that a bit.  Well, some of them are trolls I’m sure, but its a reasonable reaction to question an outlying number like that instead of accepting it at face value.

I think it’s fair for people to criticize the fielding numbers, since all they are given is an explanation of a previous version of UZR, and the results.

In my opinion, the best way to make the presentation is by giving enough of the underlying data, similar to when UZR was first presented and similar to what The Fielding Bible does and similar to what ESPN/STATS does.

To wit:
1. Tell me how many BIP per 27 outs are being presented to all the fielders (say it’s 25 for the Yanks when Jeter is on the field)

2. Tell me the league average outs per BIP for each position times the BIP for that player (say it’s .12*25=3.0 outs an average SS would get giving the BIP counts of the Yanks)

3. Tell me the adjustments for Jeter’s context for his parks, his ball distribution, his pitchers tendency to GB, the runners on base, lefty/righty batters

So, a league average SS might get
-.10 for Jeter’s parks,
+.05 for having “easier” GB hit to SS,
-.02 for having “fewer” GB balls hit to SS,
-.03 for having few GB pitchers,
+.01 for having more favorable base/out situations, and
+.04 for having favorable batter handedness,
for a total adjustment of -.05 outs per game.

(All numbers for illustration only.)

The number of outs per 27 outs Jeter had would be trivial, say 2.85.

Then, you can say:
Jeter, 2.85, 3.00, -.05

Which means that Jeter had 2.85 outs per 27 outs, the league average SS (without adjustments) would have had 3.00 outs per 27 outs, but is adjusted downward by .05 when you consider the Jeter context (with that .05 being given the breakdown as noted above).

You can add whatever additional adjustments you want, like ball hogging, line drives, bunts, etc.

To me, this would be the best presentation and would move the discussion forward.  Until then, these discussions simply tread water.

I probably make it confusing by saying “per 27 outs”, when we might as well just do “per xx BIP”, be it 25, 27, or 30 or whatever constant you want to apply.

I agree with you both.  I was being a little hyperbolic about the “trolls” and such, plus when you are posting in a thread you just throw out the numbers with little explanation.  If I were writing an article about Manny’s or someone else’s defense, I agree that as much of the underlying data as possible helps to understand exactly what is going on.