Some analysis on WAR by position

[dominik]

I made a little analysis to see how WAR is disributed by position. I wanted to check the theory that some positions are out of the line. I thought that middle IF guys were overrated and catchers underrated. to account for the outliers I also calculated the median value.
1B:

top list:

http://www.baseball-reference.com/play-index/share.cgi?id=10epN

highest gehrig 108.5
15th beckley 57.1
average top15 72,8
median 67.8

2B:

top list

highest hornsby 124,6
15th Utley 53,3
average 76,3
median 67,3

3B:
top list

http://www.baseball-reference.com/play-index/share.cgi?id=zm3Ua

highest schmidt 103
15th evans 55.1
average top15 71,2
median 66.2

SS:
top list

http://www.baseball-reference.com/play-index/share.cgi?id=cFrpt

highest wagner 126
15th cronin 61,9
average top 15 77,4
median 70,5

RF:
top list (plus ruth who doesn't reach the 50% probably due to his pitching)

http://www.baseball-reference.com/play-index/share.cgi?id=9qxyz

highest ruth 159,2
15th bobby bonds 55,7
average top 15 81.5
median 69,9

CF
top list http://www.baseball-reference.com/pl...e.cgi?id=wlOIO

highest mays 150,8
15th wynn 53,1
average top15 81,3
median 63,1

LF
top list

http://www.baseball-reference.com/play-index/share.cgi?id=gdRLu

highest bonds 158
15th johnson 52,8
average 75,8
median 64,2

C
top list

http://www.baseball-reference.com/play-index/share.cgi?id=kOIMA

highest bench 72,3
15th freehan 41,3
average 52,6
median 50,7

[dominik]

the average of the average value was
73,6
the positions were compared to the average:
1B -0.8
2B 3.3
3B -2.4
SS 3.8
RF 7.9
LF 2,2
C -21

Medians
average 64.9

1B 2.9
2B 2.4
3B 1.3
SS 5.6
RF 5
CF 1.8
LF -0.7
C -14,2


So I could not really confirm that middle IF positions are overrepresented especially compared to RF and CF. SS is elevated but not more than RF. however there is a huge gap between catchers and the rest of the positions.

[Jackaroo Dave]

Interesting. By looking at the top 15, though, aren't you limiting yourself to outliers by definition?

I wonder if catchers would be rated so low if you calculated WAR/PA?

[dominik]

I used your lists to do that.

I could do more but I didn't want to type in 100 rows into excel.

[Jackaroo Dave]

Geeze, Dom, I didn't know you were typing them. Can't you sort them by position-- the last column, I think-- and copy and paste?

This is an interesting project, because for me anyway, the defensive aspect of WAR is very obscure.

[dominik]

I did sort by position but for some reason you can't copy and paste from bref (always marks the whole rows). so to to the percentages you need to get them to excel or so.

[GiambiJuice]

WAR is a counting stat and catchers generally play far fewer games than the rest of the positions. It makes sense that Catcher's would have lower overall.

Have you attempted to look at WAR per game or per 162? If the top catchers have a lower WAR per game or per 162 than the top players at other positions, I think you'd be onto something.

[willshad]

WAR is a counting stat and catchers generally play far fewer games than the rest of the positions. It makes sense that Catcher's would have lower overall.

Have you attempted to look at WAR per game or per 162? If the top catchers have a lower WAR per game or per 162 than the top players at other positions, I think you'd be onto something.

Bench actually played 2158 games, which is a decent amount. He has .9 more WAR than Lou Whitaker, in about a seasons worth less games played. They are almost identical, accoridng to WAR. Berra has about 15 LESS WAR than Whitaker, in about 200 less games played. Chase Utley has 53.3 WAR in 1192 games...making him about TWICE as valuable as Berra on a per game basis.

Considering these are the consensus two best catchers of all time, and Whitaker is not even a HOFer ( and utley probably won't be either), this does not bode well for the stat.

Buster Posey this season had what is probably one of the top 5 or 10 offensive seasons of all time for a catcher, with good defense, and his WAR total was 7.3. Ben Zobrist has already had WAR seasonal totals of 8.3 and 8.5...one of which he batted .269.

Somewhere along the line, catchers are getting cheated out, and/or good fielding infielders are getting overrated..and it has nothing to do with the amount of games played. You can pretty much add 1 WAR each season, and 2 WAR for each great season of a catcher's career in order to even things out.

[Honus Wagner Rules]

Bench actually played 2158 games, which is a decent amount. He has .9 more WAR than Lou Whitaker, in about a seasons worth less games played. They are almost identical, accoridng to WAR. Berra has about 15 LESS WAR than Whitaker, in about 200 less games played. Chase Utley has 53.3 WAR in 1192 games...making him about TWICE as valuable as Berra on a per game basis.

Considering these are the consensus two best catchers of all time, and Whitaker is not even a HOFer ( and utley probably won't be either), this does not bode well for the stat.

Buster Posey this season had what is probably one of the top 5 or 10 offensive seasons of all time for a catcher, with good defense, and his WAR total was 7.3. Ben Zobrist has already had WAR seasonal totals of 8.3 and 8.5...one of which he batted .269.

Somewhere along the line, catchers are getting cheated out, and/or good fielding infielders are getting overrated..and it has nothing to do with the amount of games played. You can pretty much add 1 WAR each season, and 2 WAR for each great season of a catcher's career in order to even things out.

Also Posey played a fair amount of first base this season as well (29 starts). From what I understand this affects Posey's defensive WAR negatively.

[dominik]

Also Posey played a fair amount of first base this season as well (29 starts). From what I understand this affects Posey's defensive WAR negatively.

Yes. if catchers get shifted they usually play 1B or DH. this doesn't affect their dWAR but their positional runs. the lack of PAs certainly plays a big role too. another thing might be that catchers usually rack up a lot of negative runs on the bases (or at least no positive values).

I don't actually think that middle IF guys are overrated (as their WAR is not much higher than 1B or OF guys) but compared to catchers they don't have those games were they lose a lot of positional runs and they usually create a of runs with their legs. a great runner can create up to 10% (in rare occasions even more) of his value with his legs while a catcher might even lose a few percent.

[Jackaroo Dave]

Bench actually played 2158 games, which is a decent amount. He has .9 more WAR than Lou Whitaker, in about a seasons worth less games played. They are almost identical, accoridng to WAR. Berra has about 15 LESS WAR than Whitaker, in about 200 less games played.

Somewhere along the line, catchers are getting cheated out, and/or good fielding infielders are getting overrated..and it has nothing to do with the amount of games played.

I was surprised to see this, but thinking it over, we do give catchers a huge subjective position bounce for reasons that are not captured in the stats: seasons are shortened, careers are shortened, batting performance is impaired. Baserunning usually disappears. Except for pitchers, everyone else plays the field more or less relaxed comfort. Catchers play in continual discomfort, punctuated by acute pain, punctuated by occasional excruciating pain. Neither WAR nor any other stat compensates for these extra obstacles, but fans can, do, and should.

(Edit: One consequence of this is that most players would probably not be catchers if they had a choice. I don't know how many good hitters chose to play other positions, but we all know of players who would have never made it at any other position.)

In his second historical abstract, Bill James ranks the top 100 players. The highest ranked catcher is Josh Gibson, number 9. The highest ranked MLB catcher is Berra at number 41, then come Bench and Campy close behind, Cochrane and 1999 Piazza in the seventies, and Fisk in the nineties, and that's it. My point is that by far the highest ranked catcher is the one ranked by reputation, not by numbers.

His approach begins with team wins, so it works in reverse from WAR. James was uncomfortable with the low ranking that Bench got and he pointed out it was still better than what other statisticians could come up with.

But think about it. Suppose Berra or Gabby Hartnett, or Fisk, or even Cochrane and Dickey had put up the same numbers as third basemen. How many would be top tier HOFers? Suppose Hodges had put up the same numbers as a catcher. Would there even be an argument over his qualifications? No, because we quite properly grant an extrastatistical hardship bonus to catchers.

"But they're CATCHERS! It's not just that it's a debilitating position; it's a premier defensive one. Look at Jim Hegan and Bob Boone. They spent long careers helping their teams win pennants and world series, and neither could hit his hat size."

That's as may be. No one doubts they were excellent catchers, and I think no one really knows how much good that did their teams. Maybe Ubi could weigh in with the WOWY stat to see how their teams fared with them and without them. Otherwise, it's notoriously difficult to capture catching excellence statistically. A catcjer makes over a hundred plays a game that are pretty much routine. Passed balls, dropped popups, errant throws, kicked bunts just aren't frequent enough to make a big difference. Containing the running game is not a vital issue nowadays. But what about Ernie Lombardy? Well, yes, what about him? Here was a hall of fame catcher with the foot speed of my living room sofa.

I'm not saying there's no difference between a good and a bad fielding catcher. I am saying that a bad fielding catcher can hold down a job. And the difference is notoriously hard to discern statistically, and consequently, hard for WAR or any other uberstat to account for. I think Willshad's solution is a sound one--give the catcher a hardship bonus that you think is appropriate.

[willshad]

Also Posey played a fair amount of first base this season as well (29 starts). From what I understand this affects Posey's defensive WAR negatively.

Would it really make such a big difference? We are talking about the best hitter in the league, who played the vast majority of his games as catcher, and was a positive in the field. This is a borderline historical season. A 7.2 WAR score just doesn't seem to give such a season the credit it deserves. It is a fairly common score. Even if he was a first baseman for the whole season, a 7.2 score seems kind of low.

He was basically Babe Ruth for the entire second half, and was good enough in the first half to make the All Star team...you're telling me that Mike Trout deserves 2.5 more WAR than him?

[willshad]

If offensive WAR is determined by how the player fared hitting-wise against others at his position, then does that mean that base running value is also determined compared only to others at the same position? That would make sense to me.

[dominik]

If offensive WAR is determined by how the player fared hitting-wise against others at his position, then does that mean that base running value is also determined compared only to others at the same position? That would make sense to me.

I don't think WAR does a positional adjustment for baserunning.

[JDanger]

If offensive WAR is determined by how the player fared hitting-wise against others at his position, then does that mean that base running value is also determined compared only to others at the same position? That would make sense to me.

Each of the offensive components of WAR (batting, baserunning,) is compared to the average player. After those values are summed up, then the positional adjustment is made.