Accuracy of ERA+ [Archive] - NYYFans.com Forum

pacewon

03-31-04, 11:34 PM

I'm admittingly far from an expert in ERA+, so I say:

Bob Feller's career ERA+ is 122 and he's a Hall of Famer.

Juan Marichal's career ERA+ is 122 and he's a Hall of Famer. Ditto for Eddie Plank.

Don Drysdale's career ERA+ is 121 and he's a Hall of Famer. So is Clark Griffith's and Joe McGinnity's, both Hall of Famers.

Tom Glavine's career ERA+ is 121 and he's a future Hall of Famer.

Kevin Appier's career ERA+ is 121, and in 5 years, nobody is gonna remember him. What gives? Is it just a simple exception to the rule?

markp

04-01-04, 05:07 AM

there are pitchers from the 20's and 30's this applies to as well. The HOF put a lot more hitters from that era (Lloyd Waner) and pitchers from the oughts (Jack Chesbro), teens, and sixties than they should have. It's the deception of the gaudy, out-of-context numbers.

pacewon

04-01-04, 12:15 PM

Originally posted by markp
there are pitchers from the 20's and 30's this applies to as well. The HOF put a lot more hitters from that era (Lloyd Waner) and pitchers from the oughts (Jack Chesbro), teens, and sixties than they should have. It's the deception of the gaudy, out-of-context numbers.

I was hoping you would respond mark. Thanks

AngelAstro

04-05-04, 02:02 PM

I have a hard time grasping ERA+, especially when it comes to comparing a reliever to a starter. What are good ERA+'s for a reliever and a starter?

Thanks.

markp

04-05-04, 09:29 PM

but that's for starters. If you look over the last 60 years, there are a wole lot more 1.89 ERAs in the bullpen than in the rotation. I'm not sure where the RP break even point is, but it's a lot higher than 100.

I think OPS+ and ERA+ are a lot better than Bill James forumlas in Abstract II, and are at ealst as good as such things as EqA, RC, and for the pitchers WHIP and DIPS.

Monkeyman

04-06-04, 09:28 AM

Originally posted by markp
but that's for starters. If you look over the last 60 years, there are a wole lot more 1.89 ERAs in the bullpen than in the rotation. I'm not sure where the RP break even point is, but it's a lot higher than 100.

I think OPS+ and ERA+ are a lot better than Bill James forumlas in Abstract II, and are at ealst as good as such things as EqA, RC, and for the pitchers WHIP and DIPS.

I would think a major problem with park adjusted stats is they assume the walls are rounded, and an even distance from the plate at all points. A dead pull right handed hitter in yankee stadium isn't going to have a good OPS+ because it isn't assumed to be that bad of a hitters park, since lefties hit so well there. Shouldn't they at least figure out a way to seperate lefty and righty hitters?

AngelAstro

04-06-04, 09:33 AM

Originally posted by Monkeyman

I would think a major problem with park adjusted stats is they assume the walls are rounded, and an even distance from the plate at all points. A dead pull right handed hitter in yankee stadium isn't going to have a good OPS+ because it isn't assumed to be that bad of a hitters park, since lefties hit so well there. Shouldn't they at least figure out a way to seperate lefty and righty hitters?

Anyone can correct me if I am wrong, but I believe park-adjusted stats measure how well a player would hit or pitch in a neutral ballpark, so that players could be compared to one another easily. Of course right handed hitters will hit worse at Yankee stadium. Park adjusted stats would probably give the dead pull right handed hitter a higher OPS, more in line with his abilities than his normal Yankee stadium dampened stats.

AngelAstro

04-06-04, 09:52 AM

Originally posted by markp
but that's for starters. If you look over the last 60 years, there are a wole lot more 1.89 ERAs in the bullpen than in the rotation. I'm not sure where the RP break even point is, but it's a lot higher than 100.

I think OPS+ and ERA+ are a lot better than Bill James forumlas in Abstract II, and are at ealst as good as such things as EqA, RC, and for the pitchers WHIP and DIPS.

For offense I really like GPA.

http://www.aarongleeman.com/2003_11_23_baseballblog_archive.html#106991283868290519

Also, click on the Introducing GPA link in the above article.

Monkeyman

04-06-04, 10:41 AM

Originally posted by AngelAstro

Anyone can correct me if I am wrong, but I believe park-adjusted stats measure how well a player would hit or pitch in a neutral ballpark, so that players could be compared to one another easily. Of course right handed hitters will hit worse at Yankee stadium. Park adjusted stats would probably give the dead pull right handed hitter a higher OPS, more in line with his abilities than his normal Yankee stadium dampened stats.

But isn't their definition of a neutral park just a park where an average # of runs are scored?

AngelAstro

04-06-04, 11:13 AM

Originally posted by Monkeyman

But isn't their definition of a neutral park just a park where an average # of runs are scored?

Here's a link that describes (in excruciating detail) the what and why of Park Adjustment.

http://highboskage.com/STATINFO.HTM

Yes, the definition of a neutral park is a park where an average number of runs are scored. But the purpose is to gain a feel of what any individual player or team would produce given a neutral playing field.

WiffleWOOD

04-06-04, 11:35 AM

here's a little more info on park adjustment:

http://www.baseballprospectus.com/cards/glossary.shtml#eqa

"An adjustment made to account for the fact that some parks are easier to hit in than average, giving an advantage (in raw statistical terms) to hitters who play for that team. Park factors are always made relative to a league average of 1.00. The park adjustments in the BP are made only on the park factor for runs, averaged over five years."

Monkeyman

04-06-04, 11:41 AM

Originally posted by AngelAstro

Here's a link that describes (in excruciating detail) the what and why of Park Adjustment.

http://highboskage.com/STATINFO.HTM

Yes, the definition of a neutral park is a park where an average number of runs are scored. But the purpose is to gain a feel of what any individual player or team would produce given a neutral playing field.

But it still doesn't take actual park shape into account, essentially averaging out the dimensions. A park factor can't be one number, it would have to account for many different types of hitters. For instance, a dead pull hitter like Giambi, a right-center guy like Mike Stanley, or Raul Mondesi who hit everything to left field. Maybe if we broke down every ball that came off of a player's bat, and where it would land with the average park dimensions, taking into account if it was a line drive, high fly or whatever, then we could do things like OPS+ and ERA+ correctly. But that still wouldn't take into account guys who hit a ton of doubles because of hard turf. Its really still just fudging it.

AngelAstro

04-06-04, 02:17 PM

Originally posted by Monkeyman

But it still doesn't take actual park shape into account, essentially averaging out the dimensions. A park factor can't be one number, it would have to account for many different types of hitters. For instance, a dead pull hitter like Giambi, a right-center guy like Mike Stanley, or Raul Mondesi who hit everything to left field. Maybe if we broke down every ball that came off of a player's bat, and where it would land with the average park dimensions, taking into account if it was a line drive, high fly or whatever, then we could do things like OPS+ and ERA+ correctly. But that still wouldn't take into account guys who hit a ton of doubles because of hard turf. Its really still just fudging it.

I'm not sure why you believe that it would have to account for many types of hitters. If you averaging all of these numbers together, you are getting a very good idea about what is happening in individual ballparks. I guess the numbers won't be perfect, but it will give you a much better idea about the skill of someone like Alfonso Soriano who was probably hampered by death valley.

Monkeyman

04-06-04, 02:34 PM

Originally posted by AngelAstro

I'm not sure why you believe that it would have to account for many types of hitters. If you averaging all of these numbers together, you are getting a very good idea about what is happening in individual ballparks. I guess the numbers won't be perfect, but it will give you a much better idea about the skill of someone like Alfonso Soriano who was probably hampered by death valley.

I was under the impression that the purpose of park adjusted stats was to figure out how good a guy would be in the "average" stadium. The only way to really do that would be to have him play against an average team with an equal number of games in each stadium.

A stadium isn't "average" just because an "average" number of runs are scored there. Bear with me here. If a stadium gives an advantage to a certain type of hitter because of some quirk in its construction, be it the Green Monster, or short right field porches in Yankee Stadium or wherever the hell the Giants play now, its going to give an advantage to the guy who can best take advantage of THAT particular thing. Lets say you had two Barry Bonds on the Giants, one right-handed, and one left-handed. We'll call the lefty Barry Bonds, and the righty... oh... Jon Dowd. They are identical in their abilities, however Bonds would have better #s than Dowd, because the stadium was, essentially, built with Bonds in mind, giving him a nice short porch in right to shoot for. Does that make Bonds the better hitter? According to OPS+, it does. But, with my subjective analysis (gasp!) I can tell thats not really the case.

AngelAstro

04-06-04, 02:41 PM

Originally posted by Monkeyman

I was under the impression that the purpose of park adjusted stats was to figure out how good a guy would be in the "average" stadium. The only way to really do that would be to have him play against an average team with an equal number of games in each stadium.

A stadium isn't "average" just because an "average" number of runs are scored there. Bear with me here. If a stadium gives an advantage to a certain type of hitter because of some quirk in its construction, be it the Green Monster, or short right field porches in Yankee Stadium or wherever the hell the Giants play now, its going to give an advantage to the guy who can best take advantage of THAT particular thing. Lets say you had two Barry Bonds on the Giants, one right-handed, and one left-handed. We'll call the lefty Barry Bonds, and the righty... oh... Jon Dowd. They are identical in their abilities, however Bonds would have better #s than Dowd, because the stadium was, essentially, built with Bonds in mind, giving him a nice short porch in right to shoot for. Does that make Bonds the better hitter? According to OPS+, it does. But, with my subjective analysis (gasp!) I can tell thats not really the case.

I think that your premise is wrong. I'm almost 100% positive that Bonds and Dowd would have the same OPS+ even if Bonds had better numbers before adjustment.

Monkeyman

04-06-04, 02:45 PM

Originally posted by AngelAstro

I think that your premise is wrong. I'm almost 100% positive that Bonds and Dowd would have the same OPS+ even if Bonds had better numbers before adjustment.

But how is that possible, if it only takes the # of runs scored there into account, and not whom they are scored by?

AngelAstro

04-06-04, 02:55 PM

Originally posted by Monkeyman

But how is that possible, if it only takes the # of runs scored there into account, and not whom they are scored by?

I'll be honest, I'm not exactly sure how the math works. Here is what I believe is a good example of why it is good to look at OPS+. I am going to compare Soriano to Preston Wilson. The first numbers are the AVG/OBP/SLG from last year.

Soriano - .290/.338/.525

Wilson - .282/.343/.537

So both are batters that don't walk a whole lot, but hit for power.

Looking at the stadiums park factor from last year, in which >100 is favorable to batters:

Yankee Stadium - 96

Coors Field - 112

So Yankee Stadium played as a pitchers park last year, while Coors is obviously favorable to hitters.

The OPS+ of these two were:

Soriano - 128

Wilson - 115

So, we see that even though these two hitters peripheral numbers were very similar, their OPS+ are not as close because Soriano played half his games in a park which would reduce his numbers while Wilson played half his games in a park which pad his numbers.

Monkeyman

04-06-04, 03:35 PM

Originally posted by AngelAstro

I'll be honest, I'm not exactly sure how the math works. Here is what I believe is a good example of why it is good to look at OPS+. I am going to compare Soriano to Preston Wilson. The first numbers are the AVG/OBP/SLG from last year.

Soriano - .290/.338/.525

Wilson - .282/.343/.537

So both are batters that don't walk a whole lot, but hit for power.

Looking at the stadiums park factor from last year, in which >100 is favorable to batters:

Yankee Stadium - 96

Coors Field - 112

So Yankee Stadium played as a pitchers park last year, while Coors is obviously favorable to hitters.

The OPS+ of these two were:

Soriano - 128

Wilson - 115

So, we see that even though these two hitters peripheral numbers were very similar, their OPS+ are not as close because Soriano played half his games in a park which would reduce his numbers while Wilson played half his games in a park which pad his numbers. I wasn't doubting that OPS+ is a good stat, and that certainly demonstrates it. However, does it reflect really how much Soriano's #'s were affected by Yankee Stadium? Or does it just reflect that, overall, its a pitchers park?

AngelAstro

04-06-04, 03:59 PM

markp

04-07-04, 09:31 AM

Originally posted by Monkeyman
I wasn't doubting that OPS+ is a good stat, and that certainly demonstrates it. However, does it reflect really how much Soriano's #'s were affected by Yankee Stadium? Or does it just reflect that, overall, its a pitchers park?

OPS+ is park adjusted. It does reflect how much his numbers were affected by Yankee Stadium.

One of the great fallacies that exists in baseball is that Yankee Stadium is still death to RH hitters. It's not. When they remodeled it, they made it close to even as far as RH vs. LH. But since the media is simply unable to grasp complicated things like moving the fences much closer in left-center while moving them out in RF and right center would change who it favors, they're simply unaware of it.

Soriano went from what is essentially a fair park to a very good hitter's park (though it's certainly not the best hitter's park in baseball as Kaat described it last night.)