Fielding Win Shares

[Specs]

According to Bill James' Win Shares system, the top 20 fielders in the majors last year were:

1. MON C Brian Schneider (10.60)
2. TOR 2B Orlando Hudson (9.84)
3. LAD C Paul LoDuca (9.75)
4. LAD SS Cesar Izturis (9.52)
5. HOU C Brad Ausmus (9.39)
6. FLA SS Alex Gonzalez (9.29)
7. CHC SS Alex Gonzalez (9.11)
8. OAK 2B Mark Ellis (8.66)
9. CHW SS Jose Valentin (8.08)
10. SEA 2B Bret Boone (8.00)
11. OAK C Ramon Hernandez (7.97)
12. PHI SS Jimmy Rollins (7.86)
13. ANA C Benjie Molina (7.86)
14. KCR C Brent Mayne (7.74)
15. SEA OF Mike Cameron (7.73)

As much as I want to love this system, I'm finding a hard time believing that Brent Mayne saved more runs than Andruw Jones, even if the margin is three. Also, the only three shortstops in the majors who had less than 2 win shares / 1000 innings were Derek Jeter, Erick Almonte, and Enrique Wilson... even a Red Sox fan has to see something fishy there. Does anybody know why catchers seem to be so valued, and why players in pitcher's parks seem to be so valued (every Seattle player is among the top at his position--they are a good defensive team, but they can't be that good)?

[Hammer755]

Fielding win shares is a crock. The example that I always cite is that Craig Biggio was one of the best defensive CF in the majors last year according to DWS, while Adam Everett was one of the worst SS. That, my friend, just ain't right.

[JDOldSchool]

I'm pretty sure that James himself has said his win shares system isn't very good.

The best defensive stat I've seen is Ultimate Zone Rating (UZR). I have absolutely no idea how to calculate this stat, but it is the most objective of all the attempts to quantify good defense.

Edit: As for Andruw Jones - does anybody else find it interesting that he's a great fielder when he has Glavine, Maddux, Smoltz, etc. for years and years. This year he has lesser pitching in front of him and he's less of a fielder. Things that make you go, "Hmmm..."

[mlyons]

So, if I understand win shares correctly, does this mean that 10% of Montreal's wins are solely attributable to the defense of their catcher? He must be awfully good; does he pick off everyone who reaches base?

[matskralc]

Originally Posted by mlyons

So, if I understand win shares correctly, does this mean that 10% of Montreal's wins are solely attributable to the defense of their catcher? He must be awfully good; does he pick off everyone who reaches base?

No. 10.60 win shares is about 3 and a half wins. Montreal won 83 games last year. That's about 4.25%.

[matskralc]

Originally Posted by Specs

Does anybody know why catchers seem to be so valued?

Originally Posted by Page 67 of Win Shares

To convert the claim percentages into claim points, we do two things:

1. Subtract .200 from each Claim Percentage, and

2. Multiply the Claim Percentage (-.200) for each position by the intrinsic weight assigned to that defensive position.

The intrinsic weights in modern baseball are:

Catcher		38 points	19%
First Base	12 points	 6%
Second Base	32 points	16%
Third Base	24 points	12%
Shortstop	36 points	18%
Outfield	58 points	29%

Skip a bit...

Originally Posted by page 68 of Win Shares

2. You may wonder why the intrinsic weight given to catchers is greater than the intrinsic weight given to shortstops. The answer is, I tried the other way, not once but repeatedly, and it simply doesn't work.

Catchers don't play nearly as many innings as shortstops. For this reason, even with more weight being given to catchers than shortstops, the most valuable defensive player on most teams is the regular shortstop. If you give as much weight to shortstop as you do to catcher, it becomes extremely rare for the regular catcher to receive as many defensive Win Shares as the regular shortstop.

Skip a bunch more...

Originally Posted by page 182 of Win Shares

The nine defensive positions may be arrayed on a spectrum, which traditionally goes, from left to right: first base, left field, right field, thrid base, center field, second base, shortstop, catcher, pitcher. As you move rightward on the spectrum (away from first basemen, toward pitchers) players are evaluated more on their defensive contribution, and less on their offense. First basemen hit the most; pitchers hit the least--but pitchers are expected to make the largest contribution toward preventing the other team from scoring.

"Pitcher" and "catcher" are really so far off to the right that they're usually not considered part of the defensive spectrum.

So, basically, James just plain set up the system so that A) catchers are considered more important than shortstops, and B) catchers are mostly evaluated on their contributions to preventing runs.

[mlyons]

Originally Posted by matskralc

No. 10.60 win shares is about 3 and a half wins. Montreal won 83 games last year. That's about 4.25%.

Woops, my mistake. I thought it was for this year. Still, 4.25% seems like an awfully high proprotion to be assigned to one player's defense.

[Specs]

2003 Win Shares Gold Gloves

American League
C: Ramon Hernandez, OAK
1B: John Olerud, SEA
2B: Orlando Hudson, TOR
SS: Jose Valentin, CWS
3B: Eric Chavez, OAK
LF: Randy Winn, SEA
CF: Mike Cameron, SEA
RF: Ichiro Suzuki, SEA

National League
C: Brian Schneider, MON
1B: Derrek Lee, FLA
2B: Alex Cora, LAD
SS: Cesar Izturis, LAD
3B: Adrian Beltre, LAD
LF: Luis Gonzalez, ARI
CF: Andruw Jones, ATL
RF: Jose Cruz, Jr., SFG

Real Gold Gloves

American League

C: Ben Molina, ANA (WS: 2nd, probably would have been James' vote, with almost as many WS in far fewer innings played)
1B: Olerud
2B: Bret Boone, SEA (WS: 3rd)
SS: Alex Rodriguez, TEX (WS: 4th)
3B: Chavez
OF: Cameron
OF: Suzuki
OF: Hunter (WS: 3rd, Winn would probably have won a left field gold glove)

National League

C: Mike Matheny, STL (WS: 5th)
1B: Lee
2B: Luis Castillo, FLA (WS: 2nd)
SS: Edgar Renteria, STL (WS: 9th)
3B: Scott Rolen, STL (WS: 3rd, but he has a stranglehold)
OF: Cruz
OF: Jones
OF: Jim Edmonds, STL (WS: 7th, but another stranglehold)

The win shares system was right on seven gold gloves last year. Not too bad.

[Jason Moyer]

Originally Posted by Hammer755

Fielding win shares is a crock. The example that I always cite is that Craig Biggio was one of the best defensive CF in the majors last year according to DWS, while Adam Everett was one of the worst SS. That, my friend, just ain't right.

This is going to be a daft question maybe, but what isn't right about that?

[Jason Moyer]

Originally Posted by matskralc

So, basically, James just plain set up the system so that A) catchers are considered more important than shortstops, and B) catchers are mostly evaluated on their contributions to preventing runs.

Without pulling out my spreadsheet and just waxing from memory, I thought most of a catcher's defensive win shares came from independent putouts (on a team level) and RTO%.

[Jason Moyer]

Originally Posted by mlyons

So, if I understand win shares correctly, does this mean that 10% of Montreal's wins are solely attributable to the defense of their catcher? He must be awfully good; does he pick off everyone who reaches base?

Have you looked at his defensive stats from last year?

One of the best defensive seasons for a catcher, ever, and that's just looking at raw numbers. 43.5% RTO% is nuts, and it wasnt' a fluke. He's throwing out 51.4% this year.

[mlyons]

Originally Posted by Jason Moyer

Have you looked at his defensive stats from last year?

One of the best defensive seasons for a catcher, ever, and that's just looking at raw numbers. 43.5% RTO% is nuts, and it wasnt' a fluke. He's throwing out 51.4% this year.

I did look, and, while it's certainly an impressive RTO%, that's still only 27 outs over the course of the season. I understand that they're leveraged outs, but I still don't understand how they can be weighted quite so heavily. His other stats were good, but I still don't see how they account for more than 3 wins in and of themselves -- unless he's being given credit for some share of the pitching as well (which would be a crock).

[dougaiton]

Originally Posted by Jason Moyer

This is going to be a daft question maybe, but what isn't right about that?

Skipaway answered this better than I could, but Houston outfielders' are being rewarded for the defensive efficiency of their infield.

EDIT: But I can't find the darn post.

[Skipaway]

Checking the numbers behind win shares, apparently Expos catchers got 10 win shares purely by the caught stealing percentage.

http://www.baseballgraphs.com/winshares/teamtotals.xls

The formula for caught stealing claim points is : 25 + [(A-B)*150] with a 50 cap.
A: Caught stealing percentage
B: League caught stealing percentage

[Skipaway]

Originally Posted by Jason Moyer

This is going to be a daft question maybe, but what isn't right about that?

Bill James added team park-adjusted DER into the formula of outfielder claim points. Reasons:

First, we are starved for data with which to evaluate outfielders. We need a lot more than we have. Second, it has long been my observation - informal, not supported by research - that when a team has a fast outfield, they tend to have a good Defensive Efficiency Record. Third, it is not illogical to argue that certain markers of defensive excellence may be tied to certain positions. If a team allows few stolen bases, we assume that they have good defensive catchers. If a team turns double plays, we credit this to the shortstops and second basemen. It is equally reasonable, when a team has a high DER, to associate this more with one defensive position than with another.

None of the reasons looked convincing.

The Astro's got the highest park-adjusted DER last season, so that already gave the whole defense an extra edge when dividing win shares between pitchers and fielders (rightfully). And that same high team DER also gave the whole outfield lots of extra credit they don't deserve. The whole Astro outfield got high defensive win shares because of that.

[Jason Moyer]

Originally Posted by dougaiton

Skipaway answered this better than I could, but Houston outfielders' are being rewarded for the defensive efficiency of their infield.

EDIT: But I can't find the darn post.

I agree with that, what I meant was more like "What's wrong with Biggio being evaluated as a good defensive centerfielder?"