How accurate are Win Shares?

[Bill Burgess]

I'd like to have a conversation with respect to the accuracy of the Win Shares system.

[Bill Burgess]

Here is a chart I did, and I listed WS as one column. WS. Notice how it differs from OPS+ and Relative slugging in some cases to the extreme.

Win Shares says that Honus Wagner had the greatest season in history. 59 WS. I beg to differ. Ruth/Bonds registered some seasons over 240 OPS+, while Wagner's 1908 OPS+ was a great 205.

Any thoughts?


According to Relative Slugging Average. Those in red are pre-1920.

----------------Rel.SLG-Rel.OBP--Rel.BA.-OPS+-INK--PCA-----WS--TPR
Ruth, 1920-------2.08----1.47-----1.27---256---16--28.83---51--10.0
Ruth,1921--------2.07----1.43-----1.29---239---16--24.79---53---9.8
Bonds, 2001------2.04 ---1.57-----1.27---262---09--20.82---54--12.2
Ruth, 1923-------1.96----1.55-----1.39---239---16--22.2----55--11.2
Bonds, 2002------1.96----1.77-----1.45---275---09--20.6----49--11.2
Gehrig,1927------1.91----1.34-----1.30---221---07--17.17---44---9.1
Williams, 1941---1.88----1.61-----1.52---235---16--22.1----42---8.8
Ruth, 1924-------1.86----1.43-----1.30---220---16--20.8----45---8.5
Foxx,1932--------1.85----1.35-----1.31---205---14--18.54---40---7.2
Bonds, 2004------1.83----1.76-----1.32---260---09--21.8----53--12.5
Musial,1948------1.83----1.35-----1.44---200---20--20.30---46---7.3
Mantle,1956------1.78----1.36-----1.35---210---18--22.96---49---8.8
Cobb, 1917-------1.78----1.39-----1.54---209---16--19.7----46---8.4
Hornsby,1922-----1.78----1.31-----1.37---207---23--20.37---42---9.2
Yaz,1967---------1.77----1.38-----1.38---195---21--16.81---42---6.5
Wagner,1908------1.70----1.35-----1.43---205---19--18.42---59--10.2
Cobb,1911--------1.73----1.38-----1.53---196---22--20.74---47---6.4

[SABR Matt]

Personally...and I know this may SHOCK you guys...but I think PCA is significantly more accurate. Obviously I don't have any way of proving this beyond a shadow of a doubt aside from my own intuition when looking at my conclusions vs. James' conclusions.

That said, I don't think WS are inaccurate...in general I think they have the right idea. Not sure how to expand on that thought.

[Mariano_Rivera]

Personally...and I know this may SHOCK you guys...but I think PCA is significantly more accurate. Obviously I don't have any way of proving this beyond a shadow of a doubt aside from my own intuition when looking at my conclusions vs. James' conclusions.

That said, I don't think WS are inaccurate...in general I think they have the right idea. Not sure how to expand on that thought.

No really, are you serious?

[Tango Tiger]

Win Shares includes fielding and positioning.

At the very least, compare Wagner's OPS to other SS of his era.

[Bill Burgess]

OK. YOu guys are not picking up on the many discrepancies. So, I will lead the way in showing some.

----------------Rel.SLG-Rel.OBP--Rel.BA.-OPS+-INK--PCA-----WS--TPR
Wagner,1908------1.70----1.35-----1.43---205---19--18.42---59--10.2
Cobb,1911--------1.73----1.38-----1.53---196---22--20.74---47---6.4

Here, Honus is shown to have a 59-47 WS lead. Would seem to be a crushing, over-whelming, level of stats. But wait. When we look at them, where's the crushing difference?

Is somebody telling me that Wagner's being a better fielder worth 12 whole WS points? How is better fielding possible to account for such an avalanche of points?

Can Wagner really save his team that many runs? Wagner's slight .03 point slg. edge is nullified by Ty's slight .05 point BA lead, plus better running.

Cobb stole 83 bases, to set a short-lived record, while Honus stole 53 bases. Honus walked 10 more times than Ty. And what about LQ?

So, where does the 12 points really come from?

[Honus Wagner Rules]

OK. YOu guys are not picking up on the many discrepancies. So, I will lead the way in showing some.

----------------Rel.SLG-Rel.OBP--Rel.BA.-OPS+-INK--PCA-----WS--TPR
Wagner,1908------1.70----1.35-----1.43---205---19--18.42---59--10.2
Cobb,1911--------1.73----1.38-----1.53---196---22--20.74---47---6.4

Here, Honus is shown to have a 59-47 WS lead. Would seem to be a crushing, over-whelming, level of stats. But wait. When we look at them, where's the crushing difference?

Is somebody telling me that Wagner's being a better fielder worth 12 whole WS points? How is better fielding possible to account for such an avalanche of points?

Can Wagner really save his team that many runs? Wagner's slight .03 point slg. edge is nullified by Ty's slight .05 point BA lead, plus better running.

Cobb stole 83 bases, to set a short-lived record, while Honus stole 53 bases. Honus walked 10 more times than Ty. And what about LQ?

So, where does the 12 points really come from?

Bill,

Let's see what the other Bill said about Wagner's 1908 season. This is from the NBJHBA:

Wagner’s 1908 season ranks, by the Win Shares system, as the greatest season of the 20th century; even Babe Ruth never matched it. Why? Well, Wagner hit .354 and drove in 109 runs. This is no big deal; Wagner hit .354 and drove in 109 runs pretty much every year. In baseball history there are lots of guys who hit .350 and drove in 150 runs. What makes Wagner different is (a) defense and (b) a quite exceptional ratio of wins to runs scored.

The National League ERA in 1908 was 2.35-the lowest of the dead ball era, the lowest ERA for a league in the 20th century.

In modern baseball, the league ERAs are just about twice that, about 4.70. So double those numbers: if you had a shortstop, like Wagner, who drove in 218 runs, what would that be worth?

In additon to that, the Pirates were playing in the poorest hitter’s park, which reduced scoring by 16%…In the context, where runs were extremely scarce, Wagner led the National League in hits (201), doubles (by 30%), in triples, in total bases (by 40), in stolen bases (53), in runs created (by 28%), in batting average (by 20 points), in RBIs, in on base percentage (.415), and in slugging percentage (by almost 100 points). He was second in home runs and runs scored. At shortstop, he led the league in putouts, by 40. Even Babe Ruth never had as much impact on the game he was playing as Honus Wagner did in that one season. (page 548-549)

My interpretation of this is that WS places heavy emphasis or league context. Let's compare the 1908 NL vs the 1911 AL:

1908 NL: .239/.299/.306. 3.33 R/G, 2.34 ERA
1911 AL: .273/.338/.358, 4.61 R/G, 3.34 ERA

Obviously, Cobb's league was much more of an offensive league.

[AstrosFan]

Remember, the context for Win Shares is not based on OBP or SLG, or BA, but on runs. The context Wagner was playing in is around the lowest of all-time, that or the 1968 Dodger Stadium context. Cobb, on the other hand, was playing in what was considered a lively era at the time, as a new cork-centered ball was introduced.

The actual wins versus Pythagenpat wins for both teams actually gives the advantage to Tyrus, the Pirates exceed their Pythagenpat by 5.89 wins, the Tigers by 6.86 wins. So Honus isn't getting any advantage there.

So what it basically comes down to is the run context being a lot different, and Wagner being far superior defensively. In spite of Cobb leading the league in everything, his year isn't necessarily better than Wagner's offensively, and he's quite a bit behind defensively. There are some other quirks, I expect, that cause Wagner to rate so much further ahead, but those are the two most crucial reasons.

[brett]

OK. YOu guys are not picking up on the many discrepancies. So, I will lead the way in showing some.

----------------Rel.SLG-Rel.OBP--Rel.BA.-OPS+-INK--PCA-----WS--TPR
Wagner,1908------1.70----1.35-----1.43---205---19--18.42---59--10.2
Cobb,1911--------1.73----1.38-----1.53---196---22--20.74---47---6.4

Here, Honus is shown to have a 59-47 WS lead. Would seem to be a crushing, over-whelming, level of stats. But wait. When we look at them, where's the crushing difference?


Actually, on second glance, 20 defensive runs saved would have been 9-10 win shares in that context.

I am all for rating hitters relative to the hitting of their time, but when hitting changes the value of a defensive season then I am more for looking at the defensive greatness. Wagner wasn't a better fielder because fielding was particularly low that year, but because hitting was low, so the same relative defensive season is worth more in the offensive context.

Another thing. Cobb's league had a higher LQ for sure. If you can't make up as much value with steals and fielding then it is harder for a great all around player to separate from the league. Win shares has no league quality adjustment.

OPS+ scores might not have varied much in NL '08, but total value did because the value of defense and steals varied much more than usual.

Bill, I think that they were basically equal as hitters.

A great defensive SS probably saves 20 runs on an average one. That is about 7-8 win shares. A mid level positional adjustment would be another 5 win shares to the SS so that is the 12. The mid-level positional adjustment would be only about half of the difference between the offensive production of an average SS and an average CF so he was pretty conservative here.

That still assumes that Cobb was an average centerfielder, and that they were equal baserunners in every phase of the game.

Also, Wagner got a little lucky. His good but not great steal rate was actually very productive just because there was so little offense. Based on skills, he wasn't a better relative basestealer than in other years, but he gets more credit because hitters produced less which is one of the cases where I think it makes sense to look at skills versus "value."

It was a year where Wagner happened to put together his best realtive hitting, fielding and baserunning all at once. Cobb had better hitting years where his steals or defense were less valuable.

[538280]

I think WS are generally accurate but I"m not going to form my whole opinion based on them, and I've said that four times in the Morgan/A-Rod thread yet no one seems to be listening I first look at the stats myself (knowing what I know about how teams win games) and determine how good I think a player is.

[Bench 5]

Of the 59 Win Shares that Hans receives credit for under this system, how many are for offensive production and how many are for defense? Same question for Cobb?

I thought I read somewhere that Wagner got a big bump on the defensive end which made this season rate so high per James' system. I think that on the offensive side Wagner's season is considered great under Win Shares - but not in the same category as the best of Ruth and Bonds.

[AstrosFan]

Of the 59 Win Shares that Hans receives credit for under this system, how many are for offensive production and how many are for defense? Same question for Cobb?

I thought I read somewhere that Wagner got a big bump on the defensive end which made this season rate so high per James' system. I think that on the offensive side Wagner's season is considered great under Win Shares - but not in the same category as the best of Ruth and Bonds.

Wagner got 49.21 OWS in 1908. Bonds got 52.22 in 2001, his highest. Ruth's 1923 received 48.48. Wagner also got 9.74 DWS.

[Bill Burgess]

Bill,

Let's see what the other Bill said about Wagner's 1908 season. This is from the NBJHBA:



My interpretation of this is that WS places heavy emphasis or league context. Let's compare the 1908 NL vs the 1911 AL:

1908 NL: .239/.299/.306. 3.33 R/G, 2.34 ERA
1911 AL: .273/.338/.358, 4.61 R/G, 3.34 ERA

Obviously, Cobb's league was much more of an offensive league.

But Adam, I showed RELATIVE STATS, not raw numbers. They are indexed to L. avaerages, and hence they were about on par for hitting, while Ty led in SBs. Since we don't know what their CS ratios were, I assume they were about the same. Cobb was a normal 70% stealer.

So, the 12 WS must all come from fielding.

Yes, Honus was a great defensive SS, but Ty was a very, very good defensive CF. And an even better RFer. Matt Souders gives Ty a PCA gold glove in CF in 1911.

So, if Honus saves his team 20 runs at SS, how many does Ty save his from CF?

[Bill Burgess]

Remember, the context for Win Shares is not based on OBP or SLG, or BA, but on runs. The context Wagner was playing in is around the lowest of all-time, that or the 1968 Dodger Stadium context. Cobb, on the other hand, was playing in what was considered a lively era at the time, as a new cork-centered ball was introduced.

The actual wins versus Pythagenpat wins for both teams actually gives the advantage to Tyrus, the Pirates exceed their Pythagenpat by 5.89 wins, the Tigers by 6.86 wins. So Honus isn't getting any advantage there.

So what it basically comes down to is the run context being a lot different, and Wagner being far superior defensively. In spite of Cobb leading the league in everything, his year isn't necessarily better than Wagner's offensively, and he's quite a bit behind defensively. There are some other quirks, I expect, that cause Wagner to rate so much further ahead, but those are the two most crucial reasons.

But AF, relative stats already take into consideration the league strengths. If anything, Cobb's L. was the stronger.

True, Honus had to overcome his homepark, which Bill James calls the poorest hitting park in the game. Bill says Exhibition Park reduced scoring by 16%, so that might account for some of it. And that vindicates my former argument that Honus was a better player than Mays, and some disputed that. Said it was good for Wagner to hit in a park which reduced scoring 16%! The naiveté!

And I didn't think Bill James used a Positional Adjustment. Am I mistaken on this?

PCA says that Cobb's season was the better of the two. What say ye, Mathew?

[ChrisLDuncan]

Well, if you look at career adjusted EqA Wagner's EqA was .356 for that season, so a below average season for Ruth , and Cobb's EqA in 1911 was .343, so Wagner was also better offensively by that metric too.

But AF, relative stats already take into consideration the league strengths. If anything, Cobb's L. was the stronger.

True, Honus had to overcome his homepark, which Bill James calls the poorest hitting park in the game. Bill says Exhibition Park reduced scoring by 16%, so that might account for some of it. And that vindicates my former argument that Honus was a better player than Mays, and some disputed that. Said it was good for Wagner to hit in a park which reduced scoring 16%! The naiveté!

And I didn't think Bill James used a Positional Adjustment. Am I mistaken on this?

PCA says that Cobb's season was the better of the two. What say ye, Mathew?

I'm not going to comment on LQ because I haven't researched it.

To the Honus/Mays point, that and calling him a freakishly good defensive SS is an understatement...also James articulates some other Mays/Wagner points too.

Bill James uses a positional adjustment in a round about way. His win shares are 52% offensive and 48% defensive and each defender can only earn so large of a chunk of the DWS. For more on his Win Shares check this out

http://en.wikipedia.org/wiki/Win_shares
http://espn.go.com/mlb/columns/neyer_rob/42798.html
http://www.stats.com/store/graphics/wins_intro.pdf
http://www.tangotiger.net/winshares.htm
http://www.tangotiger.net/billjames.htm
http://www.tangotiger.net/winsloss.html
http://www.tangotiger.net/winshares.pdf this one's a lil long


You can also buy his book WinShares but I believe it's out of print and it costs 50 bucks from some guy on Amazon. Overall, I think Win Shares are a place to start on to what a player did, however I don't think it's the end all be all as to who was better.

[SABR Matt]

Actually no...it's 52% DEFENSIVE and 48% OFFENSIVE. It's a kuldge factor to answer his concern that pitchers weren't getting enough wins in his mind.

One of many kludgy adjustments he makes in order to force his system to fit his world view.

In any event...Cobb's 1911 season is viewed as slightly INFERIOR to Wagner's 1908 seaosn by normalized PCA. It's very very close, however.

Cobb scored 5.28 defensive wins (by far his best season defensively...his normal in CF was closer to 2 wins even in his youth) and 17.42 offensive wins in 1911 (these are normalized win counts...normalized such that the mean and standard deviation of win scoring rates are the same for every league).

Wagner scored 3.30 defensive wins in 1908 (a slightly down year for him by PCA...his normal was closer to 4.0-4.5 wins) and 20.55 offensive wins.

Cobb loses by the narrow margin of 22.70 wins to 23.85 wins.

I think it is reasonable to assert that Wagner was a slightly better overall player in 1908 than Cobb was in 1911...I do NOT think it is reasonable to assert as WS do that Wagner is 25% better than Cobb...that strikes me as a bit of an oddball assertion.

[Bill Burgess]

Actually no...it's 52% DEFENSIVE and 48% OFFENSIVE. It's a kuldge factor to answer his concern that pitchers weren't getting enough wins in his mind.

One of many kludgy adjustments he makes in order to force his system to fit his world view.

In any event...Cobb's 1911 season is viewed as slightly INFERIOR to Wagner's 1908 seaosn by normalized PCA. It's very very close, however.

Cobb scored 5.28 defensive wins (by far his best season defensively...his normal in CF was closer to 2 wins even in his youth) and 17.42 offensive wins in 1911 (these are normalized win counts...normalized such that the mean and standard deviation of win scoring rates are the same for every league).

Wagner scored 3.30 defensive wins in 1908 (a slightly down year for him by PCA...his normal was closer to 4.0-4.5 wins) and 20.55 offensive wins.

Cobb loses by the narrow margin of 22.70 wins to 23.85 wins.

I think it is reasonable to assert that Wagner was a slightly better overall player in 1908 than Cobb was in 1911...I do NOT think it is reasonable to assert as WS do that Wagner is 25% better than Cobb...that strikes me as a bit of an oddball assertion.

But Matt? You formerly gave me 20.74 for Cobb for 1911, and 18.42 for Wagner for 1908. Were those older, unrevised PCA values?

If you notice my chart, I give the PCA values you assigned me previously. Are all of them now outdated? I used PCA to sequence my entire 'Some of the Greatest Hitting Seasons List'.

At some point, if the values you gave me no longer apply, I will be needing the new set. Is that alright? And by that, I refer to the Offensive set, and the overall set.

In any case, I was only referencing the Cobb/Wagner seasons to dramatize the differences between PCA/WS. I will show some others too in the chart below.

Can anyone else besides me see the huge differences?

Maybe before I was just using the Offensive components, Matt. Does that conform to your present Cobb, 1911, Wagner, 1908 offensive PCA values?

Bill

[Bill Burgess]

True, Honus had to overcome his homepark, which Bill James calls the poorest hitting park in the game. Bill says Exhibition Park reduced scoring by 16%, so that might account for some of it. And that vindicates my former argument that Honus was a better player than Mays, and some disputed that. Said it was good for Wagner to hit in a park which reduced scoring 16%! The naiveté!

Chris? Does hearing Bill tell us that Wagner homepark, Exhibition Park, 1900-1908, influence you any in realizing that Wagner's hitting stats deserve a bump, due to 16% less scoring?

I hope this assists you in realizing that perhaps you shorted Wagner some in previous debates, in his hitting.

[Bill Burgess]

In any event, I would like to look further into the accuracy of the Win Shares system, as a reference guide to some of our many comparison debates, OK?

I seldom venture here, into the scary realm of stats, but sometimes I must, in order to further my stat education. So, go easy, fellas. These are my baby steps. Just learning to walk here.

Notice in the chart below, I give both the WS and PCA offensive values. And I realize that WS includes defense while the PCA numeral does not. Can anyone else identify some discrepancies? Please don't make me do all the heavy lifting here, guys.


Some of the Greatest Hitting Seasons Ever: Listed according to PCA Offensive Values.
Seasons pre-1920 are high-lighted in red, for convenience.

----------------Rel.SLG-Rel.OBP--Rel.BA.-OPS+-INK--PCA-----WS--TPR
Ruth, 1920-------2.08----1.47-----1.27---256---16--28.83---51--10.0
Ruth,1921--------1.99----1.39-----1.25---239---16--24.79---53---9.8
Mantle,1956------1.75----1.34-----1.35---210---18--22.96---49---8.8
Ruth, 1923-------1.89----1.50-----1.34---239---16--22.2----55--11.2
Williams, 1941---1.79----1.56-----1.46---235---16--22.1----42---8.8
Bonds, 2004------1.83----1.76-----1.32---260---09--21.8----53--12.5
Bonds, 2001------2.04 ---1.57-----1.27---262---09--20.82---54--12.2
Ruth, 1924-------1.80----1.39-----1.27---220---16--20.8----45---8.5
Cobb,1911--------1.73----1.32-----1.53---196---22--20.74---47---6.4
Bonds, 2002------1.96----1.77-----1.45---275---09--20.6----49--11.2
Hornsby,1922-----1.76----1.30-----1.35---207---23--20.37---42---9.2
Musial,1948------1.72----1.28-----1.36---200---20--20.30---46---7.3
Cobb, 1917-------1.72----1.36-----1.50---209---16--19.7----46---8.4 
Lajoie, 1901-----1.53----1.38-----1.73---200---23--18.94---42---7.4
Morgan, 1975-----1.37----1.31-----1.22---169---?----?------44---9.8
Morgan, 1976-----1.33----1.32-----1.20---187---?----?------37---7.5
Foxx,1932--------1.76----1.30-----1.26---205---14--18.54---40---7.2
Wagner,1908------1.70----1.35-----1.43---205---19--18.42---59--10.2
Speaker, 1912----1.44----1.39-----1.54---188---06--18.13---51---7.1
Medwick, 1937----1.37----1.24-----1.42---180---24--17.47---40---5.2
Cash, 1961-------1.41----1.46-----1.37---201---07--17.30---42---8.3
T.Williams,1949--1.71----1.38-----1.30---192---19--17.24---40---6.6
Gehrig,1927------1.88----1.34-----1.30---221---07--17.17---44---9.1
J.Jackson, 1911--1.49----1.38-----1.49---193---00--17.13---39---6.8
Yaz,1967---------1.65----1.30-----1.29---195---21--16.81---42---6.5
F.Robinson, 1966-1.28----1.31-----1.68---199---18--16.48---41---6.8
H.Duffy, 1894----1.42----1.26-----1.58---177---20--16.44---33---4.0
Delahanty, 1899--1.44----1.33-----1.59---189---16--16.36---41---5.4
Babe Herman,1930-1.51----1.26-----1.29---170---00--16.26---32---3.5
Wagner, 1900-----1.36----1.28-----1.56---175---10--16.05---34---4.3
Hack Wilson,1930-1.61----1.26-----1.17---178---13--15.81---35---4.9
Carew, 1977------1.45----1.35-----1.40---178---10--15.74---37---6.3
Vaughan, 1935----1.38----1.48-----1.31---190---09--15.72---39---7.6
Kiner, 1951------1.18----1.31-----1.60---184---12--15.64---35---5.1
Sisler,1920------1.63----1.29-----1.43---181---08--15.40---33---7.6
Al Rosen, 1953---1.28----1.23-----1.60---180---14--15.33---42---6.5
Burkett, 1901----1.42----1.35-----1.49---181---11--15.21---38---5.8
Aaron, 1959------1.36----1.24-----1.59---181---10--15.21---38---7.2
Mize, 1937-------1.33----1.24-----1.55---172---00--14.95---34---3.3
Terry, 1930------1.32----1.25-----1.38---158---07--14.46---32---5.5
Klein,1930-------1.53----1.21-----1.27---159---06--14.17---28---5.6
Sisler, 1922-----1.49----1.34-----1.47---170---13--14.08---29---5.2
Zimmerman, 1912--1.36----1.19-----1.54---169---16--13.16---34---5.0
Ott, 1929--------1.11----1.22-----1.49---165---02--13.15---31---5.8
Simmons,1930-----1.68----1.20-----1.32---176---07--12.94---36---4.5
Averill, 1936----1.30----1.16-----1.48---159---04--12.86---27---4.0
DiMaggio,1937----1.62----1.16-----1.32---168---07--12.78---39---5.9
Clemente, 1967---1.13----1.24-----1.52---170---07--12.59---35---5.6
Brett, 1980------1.44----1.38-----1.66---202---07--12.43---36---7.4
Greenberg, 1937--1.19----1.22-----1.60---172---04--12.38---33---5.5
Snider, 1954-----1.28----1.20-----1.58---170---03--12.33---39---4.2
J.Robinson, 1949-1.30----1.23-----1.35---159---06--12.22---36---6.5
Heilmann, 1921---1.34----1.24-----1.51---167---07--12.13---28---3.4
Mays, 1954-------1.30----1.22-----1.63---175---08--12.11---40---6.8
Rose, 1969-------1.39----1.27-----1.38---158---07--11.94---37---3.7
B.Williams, 1970-1.24----1.10-----1.49---147---07--11.88---29---3.5
O'Doul, 1929-----1.35----1.30-----1.46---159---07--11.83---31---5.1
T.Davis, 1962----1.31----1.13-----1.34---148---11--11.70---36---3.6
Crawford, 1911---1.38----1.23-----1.46---163---00--11.49---32---2.2
Keeler, 1897-----1.42----1.28-----1.49---164---07--11.28---32---4.0
Manush, 1928-----1.34----1.15-----1.44---154---00--11.08---35---2.9
Lindstrom, 1930--1.25----1.07-----1.28---132---00--10.78---32---4.7
Z.Wheat, 1924----1.32----1.25-----1.40---163---00--10.67---35---4.5
Z.Wheat, 1925----1.22----1.13-----1.30---142---00--10.61---27---2.5
P.Waner, 1927----1.27----1.22-----1.33---155---13--10.45---36---3.4
Gehringer, 1936--1.22----1.15-----1.31---142---02--10.37---34---5.9
Reiser, 1941-----1.32----1.18-----1.54---165---13--10.22---34---4.7
Colavito, 1961---1.13----1.17-----1.46---157---01--10.10---33---4.0
J.Rice, 1978-----1.20----1.08-----1.55---158---17--10.08---36---4.0
KiKi Cuyler,1930-1.13----1.15-----1.17---133---03---9.82---29---3.0
Trosky, 1936-----1.14----1.01-----1.46---148---04---9.66---21---1.9
Oliva, 1964------1.30----1.10-----1.45---150---12---9.43---27---3.2
K.Williams, 1922-1.16----1.13-----1.57---164---08---8.95---30---4.3
J.Tobin, 1921----1.20----1.05-----1.19---119---02---8.24---25---0.5

-------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------------------------------------
Same chart, listed according to Relative Slugging Average.

----------------Rel.SLG-Rel.OBP--Rel.BA.-OPS+-INK--PCA-----WS--TPR
Ruth, 1920-------2.08----1.47-----1.27---256---16--28.83---51--10.0
Bonds, 2001------2.04 ---1.57-----1.27---262---09--20.82---54--12.2
Ruth,1921--------1.99----1.39-----1.25---239---16--24.79---53---9.8
Bonds, 2002------1.96----1.77-----1.45---275---09--20.6----49--11.2
Ruth, 1923-------1.89----1.55-----1.39---239---16--22.2----55--11.2
Gehrig,1927------1.88----1.34-----1.30---221---07--17.17---44---9.1
Williams, 1941---1.88----1.61-----1.52---235---16--22.1----42---8.8
Bonds, 2004------1.83----1.76-----1.32---260---09--21.8----53--12.5
Ruth, 1924-------1.80----1.43-----1.30---220---16--20.8----45---8.5
Mantle,1956------1.78----1.36-----1.35---210---18--22.96---49---8.8
Yaz,1967---------1.77----1.38-----1.38---195---21--16.81---42---6.5
Hornsby,1922-----1.76----1.31-----1.37---207---23--20.37---42---9.2
Foxx,1932--------1.75----1.35-----1.31---205---14--18.54---40---7.2
Musial,1948------1.72----1.35-----1.44---200---20--20.30---46---7.3
Cobb, 1917-------1.72----1.39-----1.54---209---16--19.7----46---8.4
T.Williams,1949--1.71----1.38-----1.30---192---19--17.24---40---6.6
Wagner,1908------1.70----1.35-----1.43---205---19--18.42---59--10.2
Simmons,1930-----1.68----1.20-----1.32---176---07--12.94---36---4.5
Cobb,1911--------1.64----1.38-----1.53---196---22--20.74---47---6.4
Sisler,1920------1.63----1.29-----1.43---181---08--15.40---33---7.6
DiMaggio,1937----1.62----1.16-----1.32---168---07--12.78---39---5.9
Hack Wilson,1930-1.61----1.26-----1.17---178---13--15.81---35---4.9
Lajoie, 1901-----1.53----1.38-----1.73---200---23--18.94---42---7.4
Klein,1930-------1.53----1.21-----1.27---159---06--14.17---28---5.6
Babe Herman,1930-1.51----1.26-----1.29---170---00--16.26---32---3.5
Sisler, 1922-----1.49----1.34-----1.47---170---13--14.08---29---5.2
J.Jackson, 1911--1.49----1.38-----1.49---193---00--17.13---39---6.8
Carew, 1977------1.45----1.35-----1.40---178---10--15.74---37---6.3
Speaker, 1912----1.44----1.39-----1.54---188---06--18.13---51---7.1
Brett, 1980------1.44----1.38-----1.66---202---07--12.43---36---7.4
Delahanty, 1899--1.44----1.33-----1.59---189---16--16.36---41---5.4
Burkett, 1901----1.42----1.35-----1.49---181---11--15.21---38---5.8
H.Duffy, 1894----1.42----1.26-----1.58---177---20--16.44---33---4.0
Keeler, 1897-----1.42----1.28-----1.49---164---07--11.28---32---4.0
Cash, 1961-------1.41----1.46-----1.37---201---07--17.30---42---8.3
Rose, 1969-------1.39----1.27-----1.38---158---07--11.94---37---3.7
Crawford, 1911---1.38----1.23-----1.46---163---00--11.49---32---2.2 
Vaughan, 1935----1.38----1.48-----1.31---190---09--15.72---39---7.6
Morgan, 1975-----1.37----1.31-----1.22---169---?----?------44---9.8
Medwick, 1937----1.37----1.24-----1.42---180---24--17.47---40---5.2
Aaron, 1959------1.36----1.24-----1.59---181---10--15.21---38---7.2
Wagner, 1900-----1.36----1.28-----1.56---175---10--16.05---34---4.3
Zimmerman, 1912--1.36----1.19-----1.54---169---16--13.16---34---5.0
O'Doul, 1929-----1.35----1.30-----1.46---159---07--11.83---31---5.1
Heilmann, 1921---1.34----1.24-----1.51---167---07--12.13---28---3.4
Manush, 1928-----1.34----1.15-----1.44---154---00--11.08---35---2.9
Morgan, 1976-----1.33----1.32-----1.20---187---?----?------37---7.5
Mize, 1937-------1.33----1.24-----1.55---172---00--14.95---34---3.3
Terry, 1930------1.32----1.25-----1.38---158---07--14.46---32---5.5
Z.Wheat, 1924----1.32----1.25-----1.40---163---00--10.67---35---4.5
Reiser, 1941-----1.32----1.18-----1.54---165---13--10.22---34---4.7
T.Davis, 1962----1.31----1.13-----1.34---148---11--11.70---36---3.6
Averill, 1936----1.30----1.16-----1.48---159---04--12.86---27---4.0
J.Robinson, 1949-1.30----1.23-----1.35---159---06--12.22---36---6.5
Mays, 1954-------1.30----1.22-----1.63---175---08--12.11---40---6.8
Oliva, 1964------1.30----1.10-----1.45---150---12---9.43---27---3.2
Al Rosen, 1953---1.28----1.23-----1.60---180---14--15.33---42---6.5
F.Robinson, 1966-1.28----1.31-----1.68---199---18--16.48---41---6.8
Snider, 1954-----1.28----1.20-----1.58---170---03--12.33---39---4.2
P.Waner, 1927----1.27----1.22-----1.33---155---13--10.45---36---3.4
Lindstrom, 1930--1.25----1.07-----1.28---132---00--10.78---32---4.7
B.Williams, 1970-1.24----1.10-----1.49---147---07--11.88---29---3.5
Z.Wheat, 1925----1.22----1.13-----1.30---142---00--10.61---27---2.5
Gehringer, 1936--1.22----1.15-----1.31---142---02--10.37---34---5.9
J.Tobin, 1921----1.20----1.05-----1.19---119---02---8.24---25---0.5
J.Rice, 1978-----1.20----1.08-----1.55---158---17--10.08---36---4.0
Greenberg, 1937--1.19----1.22-----1.60---172---04--12.38---33---5.5
Kiner, 1951------1.18----1.31-----1.60---184---12--15.64---35---5.1
K.Williams, 1922-1.16----1.13-----1.57---164---08---8.95---30---4.3
Trosky, 1936-----1.14----1.01-----1.46---148---04---9.66---21---1.9
Clemente, 1967---1.13----1.24-----1.52---170---07--12.59---35---5.6
Colavito, 1961---1.13----1.17-----1.46---157---01--10.10---33---4.0
KiKi Cuyler,1930-1.13----1.15-----1.17---133---03---9.82---29---3.0
Ott, 1929--------1.11----1.22-----1.49---165---02--13.15---31---5.8

The home/away splits were located at Total Baseball, Vol. 1, ed. by John 
Thorn & Pete Palmer with David Reuther, 1989, pp. 2200-2213.

[Bill Burgess]

Since you guys are a little bashful, I will continue on. Here is another comparison.

----------------Rel.SLG-Rel.OBP--Rel.BA.-OPS+-INK--PCA-----WS--TPR
Morgan, 1976-----1.52----1.34-----1.21---187---?----?------37---7.5
Wagner,1908------1.76----1.38-----1.43---205---19--18.42---59--10.2

Here, Honus crushes one of Joe's peak seasons by an even more devastating margin, 22 whole WS points.

Since some members said previously that Joe Morgan's peak was so good, how can Honus get 22 whole points on him? In this case, they can't all be DWS, can they?

It appears that Honus does crush Joe in relative slugging/BA. Any thoughts?

To cut to the chase, does everyone agree that Wagner's 1908 season was the very best season that any baseball player ever had, including some of Ruth's/Williams' monsters?

And if not, how does that square with WS being as accurate as advertised?

[Bill Burgess]

Since Bill uses Wagner's 1908 season as his reference season, here are some other seasons for comparison. Does Wagner's 1908 still look like the topper?

----------------Rel.SLG-Rel.OBP--Rel.BA.-OPS+-INK--PCA-----WS--TPR
Ruth, 1920-------2.08----1.47-----1.27---256---16--28.83---51--10.0
Ruth,1921--------2.07----1.43-----1.29---239---16--24.79---53---9.8
Ruth, 1923-------1.96----1.55-----1.39---239---16--22.2----55--11.2
Williams, 1941---1.88----1.61-----1.52---235---16--22.1----42---8.8
Bonds, 2004------1.83----1.76-----1.32---260---09--21.8----53--12.5
Bonds, 2001------2.04 ---1.57-----1.27---262---09--20.82---54--12.2
Ruth, 1924-------1.86----1.43-----1.30---220---16--20.8----45---8.5
Bonds, 2002------1.96----1.77-----1.45---275---09--20.6----49--11.2
Wagner,1908------1.70----1.35-----1.43---205---19--18.42---59--10.2

[baseballPAP]

Some time ago I set out to work on a system that would factor in known numbers and come up with a number that represents the hits saved by a team in any given season. I took into account groundball/flyball rates, BABIP, and range factor, as applied to the league averages. My analysis is taking a lot longer than expected (mostly because I get bored easy, and I have a lot of these little pet projects, in various stages of completion), but I have finished 1901 through 1911, just enough to bear a little weight on this discussion.

The first is easy...Wagner is a SS, and in 1908 he played 151 games there. The Pirates played 155, with little used utility man Charlie Starr playing 5 games (4 starts at least...one game both players appeared in). Pittsburgh SS's, by my overly complicated system, saved about 94 hits. Starr was considerably below average in all facets as a SS that year, so Wagner would rate at about +95. Among all teams, this rates 15th (of 192)...the other 14, in no particular order:
1901 Brooklyn (Bill Dahlen)
1902 Braves (Herman Long)
1905 and 1907 Highlanders(Yankees) (Kid Elberfeld)
1905 Senators (Joe Cassidy....only lasted 2 years due to his anemic offense)
1906 and 1907 Phillies (Mickey Doolan, 06 is the best so far, 114 hits saved)
1908 and 1910 Senators (George McBride)
1910 Brooklyn (Tony Smith, only full season hit .181)
1913 ChiSox (Buck Weaver)

If you're counting, you'll notice only 11 seasons, when I promised 14. Thats because Wagner in 1909, 1910 and 1913 also makes the top 15. In other words, Wagner had 4 of the best 15 defensive seasons by a SS from 1901-1913. I would say that qualifies him as an A+ defender by my system, although it isn't as yet designed to rate individual players, just teams.

Cobb is MUCH harder to nail down, as I haven't yet found a good way to lock down the expected balls hit to LF, CF, and RF. I can say that Cobb's 1911 OF with Ty in CF, Wahoo in RF and Davy Jones and Delos Drake splitting time in RF was in the bottom third of all OFs to date. The league RF for OFs combined on a team level was 6.03, but Detroit had a low strikeout/flyball staff that year, and the Tigers should have managed 6.54. The came in at 6.24, losing about one third of a ball per game as compared to the expected average. The difficult part is in locating where the fault lies between the OFers. Crawford was no gazelle, but was only a few years removed from playing center. The two headed LFer, Drake and Jones both seem to be at least adequate,maybe even better than that. Cobb is most of the rest, and at least as he compares to the league average RF, had one of his best seasons in center. That is as far as I can go with this system though....at least until I figure out an accurate way to project how many balls each OFer should be getting to. I'd give Cobb an A for this season, but no more than a lowish B, maybe even B- career in CF.

I really like this system, and I think it has some application for rating players. At least before more modern defensive measures that utilize PBP data can be used anyway.... now if only I had someone to enter in the other 100 seasons or so of data It does seem to back up a lot of general consensus opinions from the era...Nap Lajoie really stands out, as does Hal Chase and Eddie Collins. When teams employed several players, most often the team comes in with a poor score. Most players with a long career are fairly consistent with their impact....they don't have one awesome season mixed in with a bunch of poor ones.

[brett]

Wagner got 49.21 OWS in 1908. Bonds got 52.22 in 2001, his highest. Ruth's 1923 received 48.48. Wagner also got 9.74 DWS.

Are offensive WS adjusted to position, or does that go into the defensive bank. It seems that they are putting the positional value into the offensive end.

[brett]

Actually no...it's 52% DEFENSIVE and 48% OFFENSIVE. It's a kuldge factor to answer his concern that pitchers weren't getting enough wins in his mind.

I thought it was offensive players don't get credit when runs are produced by errors.

Also, an average team will have 5.5% more defensive innings than offensive innings in a win, 2.75% more per game on average.

[538280]

Of the 59 Win Shares that Hans receives credit for under this system, how many are for offensive production and how many are for defense? Same question for Cobb?

I thought I read somewhere that Wagner got a big bump on the defensive end which made this season rate so high per James' system. I think that on the offensive side Wagner's season is considered great under Win Shares - but not in the same category as the best of Ruth and Bonds.

Wagner's 1908 was rated by WS through 2001, when the WS book was published, as the 2nd best offensive season of all time behind only Bonds' 2001. It was ranked ahead of any offensive season that Babe Ruth ever had and ahead of any other season any player had except Bonds' 2001. I don't agree with that at all. I think that is definitely a result of a flaw in WS in that it is based on actual team wins and that sometimes the team in question's runs to wins ratio is very low, and this causes the player to look a lot better than his individual stats would suggest. Basically what WS is assuming is that since the team got more wins out of less runs than other teams they are better than their statistics. In some cases this may be true, but IMO in most cases it is more random and it causes funny results sometimes. I think that is the case with Wagner's 1908 and why they see it as SO great offensively, and I don't agree with it at all. Defenisvely they show it being a great fielding season with 9.74 DWS, but not unbelieveably great, according to DWS it was Wagner's 3rd best defensive season, and WS doesn't even show him deserving of the SS GG in 1908 (Joe Tinker beat him out). Anyway, I still think Wagner's 1908 is legitimately one of the best seasons ever overall, because of Wagner's great SS defense, not the best though and certainly not the 2nd best offensive season of all time. I think clearly offensively WS badly overrates Wagner due to the reasons I said above.