OPS+ and Yankee Stadium

[leecemark]

--Most of us would agree that one of the biggest flaws of OPS+ is that it does not adjust for handedness (or type of hitter for that matter). Parks do not affect all hitters equally, but OPS+ adjusts them all evenly. Perhaps the greatest single example of this in our discussions is how OPS+ underrates Joe DiMaggio due to the dimensions of Yankee Stadium.
--I agree that OPS+ does sell DiMaggio short, but the flip side of that is that if OPS+ under rates RH power hitters it MUST overrate LH power hitters. So if you are going to say DiMaggio would be 10% better in a balanced park then he leaps up to the 170 area in OPS+, which combined with outstanding defense in CF gives him a reasonable arguement for top 10 status.
--If you subtract that same 10% from LH power hitters though then Ruth drops to less than 190 and, while still amazing, his claim to best hitter/player ever is much more questionable. Gehrig drops down close to 160 and his spot as best 1B ever becomes much more questionable. Berra drop below Gary Carter and clearly out of best catcher ever discussions. Is 10% too stong? Do Yankee fans even want this door opened?

[Mariano_Rivera]

--Most of us would agree that one of the biggest flaws of OPS+ is that it does not adjust for handedness (or type of hitter for that matter). Parks do not affect all hitters equally, but OPS+ adjusts them all evenly. Perhaps the greatest single example of this in our discussions is how OPS+ underrates Joe DiMaggio due to the dimensions of Yankee Stadium.
--I agree that OPS+ does sell DiMaggio short, but the flip side of that is that if OPS+ under rates RH power hitters it MUST overrate LH power hitters. So if you are going to say DiMaggio would be 10% better in a balanced park then he leaps up to the 170 area in OPS+, which combined with outstanding defense in CF gives him a reasonable arguement for top 10 status.
--If you subtract that same 10% from LH power hitters though then Ruth drops to less than 190 and, while still amazing, his claim to best hitter/player ever is much more questionable. Gehrig drops down close to 160 and his spot as best 1B ever becomes much more questionable. Berra drop below Gary Carter and clearly out of best catcher ever discussions. Is 10% too stong? Do Yankee fans even want this door opened?
Interesting thought

[Myankee4life]

During Ruth/Gehrig time YS was only really short straight down the RF line.

CF- 487
deepest RC- 429
left side of bullpen gate in short right center- 350
right side of bullpen gate- 344
RF- 295

So unless they were strictly pull-hitters I really dont see how playing in YS deserves a 10% demotion. Ruth and Gehrig were neither pull-hitters.

LF-301
left side of bullpen gate in short left center- 402
right side of bullpen gate- 415
deepest LC- 457
left side of CF screen- 466
CF-461

Dimaggio was not strictly a pull-hitter. Similar to Ruth/Gehrig exept the dimensions are even worse. He definately deserves a 10% increase in OPS+. The more I look at these dimensions, the more I wonder how was he able to hit 30 HR's a year.

[DoubleX]

--Most of us would agree that one of the biggest flaws of OPS+ is that it does not adjust for handedness (or type of hitter for that matter). Parks do not affect all hitters equally, but OPS+ adjusts them all evenly. Perhaps the greatest single example of this in our discussions is how OPS+ underrates Joe DiMaggio due to the dimensions of Yankee Stadium.
--I agree that OPS+ does sell DiMaggio short, but the flip side of that is that if OPS+ under rates RH power hitters it MUST overrate LH power hitters. So if you are going to say DiMaggio would be 10% better in a balanced park then he leaps up to the 170 area in OPS+, which combined with outstanding defense in CF gives him a reasonable arguement for top 10 status.
--If you subtract that same 10% from LH power hitters though then Ruth drops to less than 190 and, while still amazing, his claim to best hitter/player ever is much more questionable. Gehrig drops down close to 160 and his spot as best 1B ever becomes much more questionable. Berra drop below Gary Carter and clearly out of best catcher ever discussions. Is 10% too stong? Do Yankee fans even want this door opened?

As a Yankee fan, no I don't want this door opened, but I do think you're on to something. :p

I agree that OPS+ would be enhanced if it took into account handedness, of course that raises other issues, first and foremost that not all hitters are strictly pull hitters - Hitters, especially good ones, will often adjust their approach to take advantage of a parks features. Wade Boggs is a good example of this - he was always great in his homepark, and Fenway, Yankee Stadium, and Tropicana Field aren't exactly streamlined.

Also, in this particular case, I don't know if 10% across the board is fair. Old Yankee Stadium was probably harsher on righthanded hitters than it was friendly for lefthanded hitters. I'd probably say 10% boost for righthanders and perhaps a 6 or 7% drop for lefthandeders.

[Ubiquitous]

There is so many ways we could pick apart OPS+ that I think one should just realize that it has its flaws and move on from there. It is just a quick handy tool that gives you a good general picture of a player. It isn't a precision tool.

[leecemark]

As a Yankee fan, no I don't want this door opened, but I do think you're on to something. :p

I agree that OPS+ would be enhanced if it took into account handedness, of course that raises other issues, first and foremost that not all hitters are strictly pull hitters - Hitters, especially good ones, will often adjust their approach to take advantage of a parks features. Wade Boggs is a good example of this - he was always great in his homepark, and Fenway, Yankee Stadium, and Tropicana Field aren't exactly streamlined.

Also, in this particular case, I don't know if 10% across the board is fair. Old Yankee Stadium was probably harsher on righthanded hitters than it was friendly for lefthanded hitters. I'd probably say 10% boost for righthanders and perhaps a 6 or 7% drop for lefthandeders.

--I don't think Yankee Stadium really helped LH hitters. It just didn't disadvantage them to the large extent assumed by OPS+. Yankee Stadium has always played as a pitchers park and sometimes as an extreme pitchers park. That gives hitters a large upward adjustment in OPS+, which should actually be huge for RH hitters (especially RH power hitters), but smaller or none for LH hitters (especially LH power hitters).

[stevebogus]

Park effects are not a constant. While Yankee Stadium always had the same basic shape the actual field dimensions changed as it was remodeled. And I'm not just talking about the recent work. Aside from that the dimensions of all the other parks changed over time. The AL ballparks of 1927 were not the same parks as 1961. And park factors are affected by the talent of the players. Some players are better suited to take advantage of what a park allows while others are not. Since the park factors will shift with the talent, what should the standard be?

For Gehrig and Ruth Yankee Stadium played as an essentially neutral park. It helped Bill Dickey hit a bunch of HRs later in his career when he became more of a pull hitter. It hurt DiMaggio and Elston Howard and Bill Skowron. Pepitone hit better there while Maris didn't.

When we start making adjustments how do we account for all that? Do we downgrade Ruth and Gehrig because some other lefties were better able to take advantage of Yankee Stadium's dimensions?

[Tango Tiger]

There is so many ways we could pick apart OPS+ that I think one should just realize that it has its flaws and move on from there. It is just a quick handy tool that gives you a good general picture of a player. It isn't a precision tool.

Agreed. After all the biggest "flaw" in OBP is that it weights the walk and HR equally. The biggest "flaw" in SB% is that it excludes HR. The biggest "flaw" in pitcher strikeouts is it doesn't tell you about how gutsy he's pitching.

Does what I'm saying sound ridiculous? Every metric is defined a certain way, and it should be evaluated as that. OPS+ comes out and says it will use a generic park factor, and that's that. It's not a flaw. It's a design decision. Its use is limited to the information it contains.

[leecemark]

--Agreed and really the best way to view it is that OPS+ is flawed in many ways and is merely a thumbnail sketch that may greatly over and under value certain players. Unfortunately, many posters of limited imagination will simply say X has an OPS+ 10 points higher than Y and was therefore a better (or more valuable) hitter. The margin more error is much higher than 10 points even leaving aside issues of longevity and league quality. Some types of players actually gain big advantages (or disadvantages) in generically nuetral parks or may not be hurt by a particular pitchers park or helped by some hitters parks.

[The Toy Cannon]

10% seems a bit ridiculous, to me. That's essentially crediting DiMaggio with 40 extra bases a season, when he played only half his games in Yankee Stadium, and it assumes(as already mentioned) that he was a pure pull hitter.

[SABR Matt]

Agreed with Toy Cannon. The YS edge for lefties might have been 3-5% tops IMHO.

[Sliding Billy]

--Most of us would agree that one of the biggest flaws of OPS+ is that it does not adjust for handedness (or type of hitter for that matter).
If OPS+ adjusted for handedness, it would become even less valuable as a stat for estimating offensive value. The fact that Yankee Stadium hurt Joe's performance reduced his value to his team. Sad, maybe, but true. If OPS+ didn't reflect that, it would be worse at doing its job, not better.

What OPS+ is sort of good for is relativizing a player's stats to the value of runs scored in the parks he played in. Joe gets a deserved bump for playing in a pitchers' park.

[leecemark]

--OPS+ no doubt is used in many ways. The way it primarily used here though is to compare players from different eras (in fact I started this in history, not stats - don't know who moved it or why). For that we want the park affects and offensive environment as nuetral as possible.

[DoubleX]

I tend to look at OPS+ as a good leaping off point. It's not the be all and end all, but it gives a good quick picture. From there other information will supplement my analysis. For example, in the case of DiMaggio, I'm aware that Yankee Stadium likely hurt his production, perhaps significantly, and caused his already great numbers to belie just how extraordinary he was. So that is something I include in my analysis. Sure, it's subjective and can't really be quantified, but when you get down to it, subjectivity is often what's making the difference. Plus, I'm only going to allow subjectivity to go so far - I'm not going to speculate that DiMaggio would have been statistically closer to Ted Wlliams if not for Yankee Stadium; to me, it's just an additional element to be factored in. I hope this all makes senses.

[mtortolero]

Agreed. After all the biggest "flaw" in OBP is that it weights the walk and HR equally. The biggest "flaw" in SB% is that it excludes HR. The biggest "flaw" in pitcher strikeouts is it doesn't tell you about how gutsy he's pitching.

Does what I'm saying sound ridiculous? Every metric is defined a certain way, and it should be evaluated as that. OPS+ comes out and says it will use a generic park factor, and that's that. It's not a flaw. It's a design decision. Its use is limited to the information it contains.

What look as a flaw is not OPS+ as a theorical way to compare one player with the league, as Lee Sinnnins RCCA or RSSA do. The flaw is the way as park factors to adjust OPS and leverage with the league are calculated in without take care as the park helps or limits the hitter skills for a RH or LH batter.
Memorial Coliseum had a park factor of 109/108 in 1961 but was a great park for RHB and a hard place to hit by LHB powerhitters. How manny Hrs. stole the park to Duke Snider in the four years he played in Memorial Co?

[Sliding Billy]

What look as a flaw is not OPS+ as a theorical way to compare one player with the league, as Lee Sinnnins RCCA or RSSA do. The flaw is the way as park factors to adjust OPS and leverage with the league are calculated in without take care as the park helps or limits the hitter skills for a RH or LH batter.
Memorial Coliseum had a park factor of 109/108 in 1961 but was a great park for RHB and a hard place to hit by LHB powerhitters. How manny Hrs. stole the park to Duke Snider in the four years he played in Memorial Co?
But that's because the point of the park adjustments, however well or poorly they are done, is not to estimate how well a player would do in some imaginary neutral-for-everybody park. It's to relativize the player's actual offensive productivity to the actual frequency of runs scored in that park at the time he played there. And that's all they're good for.

If you give Snider a boost in Memorial or a ding in Ebbets Field, you lose that. OPS+ isn't designed to smooth out the bumps of fate, just to give an eyeball estimate of the value of what the player accomplished in context of the run-scoring environment. It's utterly useless for comparing players on the basis of what they might have done if they all faced the same degree of difficulty in batting.

What's wrong with just saying, "Dimaggio would have had more home runs if he'd played in a neutral park"? Or "In considering Snider's Brooklyn home run totals, you have to keep in mind that he played in a bandbox facing teams with a 'No lefties against Brooklyn' rule"?

There's never going to be a stat that takes into account every player's advantages or disadvantages relative to every other player. And even if there were, it wouldn't yield anything of actual interest, since those advantages and disadvantages are intrinsic to a player's value, intrinsic to the game itself.

[SHOELESSJOE3]

OPS+ aside, just on the subject of home runs, during the time Williams and Dimaggio played it was more difficult for RH hitters to hit home runs.

Notice I use the words HR hitters. This is just so this does not turn into a Ted or Joe debate, thats not my intention.

For what ever reason, I don't know why parks were built in this fashion, on average there was a considerable difference in distances down the lines from the 1930s to the 1950s. Left field lines more distant than right field lines. Power alleys were about the same in RCF and LCF,not much of a factor. I took in the years from the late 1930s to about 1950. These figures are from seven AL parks, Yankee Stadium excluded.

From some quick math what I came up with.
Average line distance AL parks, seven parks.

351 feet left field.
320 feet right field.
Thats a significant difference. Anyone out there done any math to calculate this. I did it the old fashioned way, add them up and divide by the number 7.

I did one more where I excluded one extreme short distancein RF ( League Park) just to see how much closer the average line distance might be.

In this one I excluded League Park in Cleveland, 290 feet right field line.
This includes only 6 parks, Cleveland and Yankee Stadium excluded.

349 feet left field
323 feet right field.

Now the gap might narrow more if I were to exclude another extreme, Griffith with 405 to left but I don't want to cut too much from numbers. Besides there is a significant number of games played in both these parks by visiting teams under the old schedule. If I recall it would be 22 games for visitors 11 at each park.

No matter how you cut it, some advantage in home run hitting for left handed batters in that time period.

[Ubiquitous]

For what ever reason, I don't know why parks were built in this fashion, on average there was a considerable difference in distances down the lines from the 1930s to the 1950s.




Size constraint probably but that is a guess, but I do see when builders finally got the room to build big stadiums they generally built them to be symmetrical.

Looking at it I see that Tiger Stadium and Fenway had shorter distances in left field, Shibe Park, Comiskey, and Cleveland symmetrical, Griffith, Yankee, and Sportsman Park with the shorter right field fences. So it isn't a blatant favoritism to shorter right field fences in the AL. 3 parks with shorter right fields but two of them were caused by lot sizes. 2 stadiums with shorter left field fences and one at the very least because of lot size. While we have 3 baseball palaces that are neutral.

[Ubiquitous]

Alright the only one I can't show is Sportsman park since it isn't up yet, and neither is league park. This helps visualize a little better what we are talking about. Okay looking at it I see that Tiger Stadium didn't have a shorter left field. I must have looked at the wrong set of numbers for them. It is a confusing mess on ballparks site.

To me it looks like possibly size constraints forced some parks to favor one side over the other and I think it makes sense to give the favor to the lefties. If one wants to keep the offense/defense balanced then you would give the bonus to lefties who do not have as many batters in the league as righties do.

[SHOELESSJOE3]

Size constraint probably but that is a guess, but I do see when builders finally got the room to build big stadiums they generally built them to be symmetrical.

Looking at it I see that Tiger Stadium and Fenway had shorter distances in left field, Shibe Park, Comiskey, and Cleveland symmetrical, Griffith, Yankee, and Sportsman Park with the shorter right field fences. So it isn't a blatant favoritism to shorter right field fences in the AL. 3 parks with shorter right fields but two of them were caused by lot sizes. 2 stadiums with shorter left field fences and one at the very least because of lot size. While we have 3 baseball palaces that are neutral.


The figures I gave, that I dealth with were for the period the late 1930s to about 1950 since the subject was Joe Dimaggio, his OPS+ and the possible effect his home park may have had on him. I just took a quick look at the other AL parks in that time period. I see you displayed some diagrams of some parks, going to take a look.

Nice diagrams, they look like they might be from Clems Baseball, a favorite of mine.

[SHOELESSJOE3]

Sportsman park [/B]since it isn't up yet, and neither is league park. This helps visualize a little better what we are talking about. Okay looking at it I see that Tiger Stadium didn't have a shorter left field. I must have looked at the wrong set of numbers for them. It is a confusing mess on ballparks site.

To me it looks like possibly size constraints forced some parks to favor one side over the other and I think it makes sense to give the favor to the lefties. If one wants to keep the offense/defense balanced then you would give the bonus to lefties who do not have as many batters in the league as righties do.


What I have on Sportsmans.
LF------351 in 1942
LCF----379
RF-----309 in 1939
RCF----354 in 1942


League Park
LF----374 in 1938 and 375 in 1942 No idea where they came up with a foot.
LCF---415 in 1942
RF----290 in 1921

Don't have the info here but I believe the height difference was significant.
A wall in left field 5 or 6 feet high and a concrete wall in right field at about 20 feet topped by a fence of 20 or 25 feet, a total of 40 or 45 feet. Certainly stopped some line drive homers in that short 290 feet in right field.


I see you have a diagram of Cleveland Stadium dated 1933 but I believe the Indians played there only one season in the 1930s, plus some holiday and Sunday games. The Indians used League Park then Cleveland Stadium in 1947. I have no idea why Cleveland Stadium was built in the 1930s, for the Indians or for other events or maybe both.

[SHOELESSJOE3]

I have no idea why Cleveland Stadium was built in the 1930s, for the Indians or for other events or maybe both.

What I did find. It was built in the hope that it might be the site for the 1932 Olympic games. Have to check that one. It seems like you would want it in place earlier than 1931 if your shooting to host the 1932 Olympics.

[BoSox Rule]

I just use EqA, my favorite stat for hitting.

[Joltin' Joe]

I just use EqA, my favorite stat for hitting.

BSR, EqA doesn't adjust for hitter handedness or type does it?

[Chisox]

--Gehrig drops down close to 160 and his spot as best 1B ever becomes much more questionable. Berra drop below Gary Carter and clearly out of best catcher ever discussions. Is 10% too stong? Do Yankee fans even want this door opened?
I'm not sure about Berra (I'd have to look that up), but Gehrig actually did worse at Yankee Stadium than on the road, although I have no idea where the stats are to support that. Berra's for '57 on can be found at http://www.retrosheet.org/boxesetc/Jberry1010.htm.

[Ubiquitous]

Gehrigs can be found in the first edition of Total baseball.

Gehrig was a .329/.436/.620 hitter at home, and a .351/.458/.644 hitter on the road.

Gehrig had 5 seasons out of 14 seasons in which he had better rate stats at home then on the road.

[SHOELESSJOE3]

I'm not sure about Berra (I'd have to look that up), but Gehrig actually did worse at Yankee Stadium than on the road, although I have no idea where the stats are to support that. Berra's for '57 on can be found at http://www.retrosheet.org/boxesetc/Jberry1010.htm.
Just a sample of Gehrig, home/away.
----------AB-----2B----3B---HR-----Ba.-----OBP----Slg-----OPS
Home----3861---206---83---251----.329----.436----.620----1.056
Away----4140---329---79---242----.351----.458----.644----1.102

Did do a bit better on the road in percentage based numbers. Big difference in doubles hit home/away. He did have almost 300 more at bats away but it did not seem to factor in that much in some other cumulative numbers, triples and home runs, not as much as doubles hit.