I agree. ERA+ is just a flawed metric. The scale isn't what it appears to be. I wish people wouldn't use it.

Seven of the top ten in ERA+ began their careers between 1986 and 1991. With Finley, Kevin Brown, and Mike Mussina among the names, it seems ERA+ isn’t the best metric to use for such comparisons.

I'm inclined to agree. ERA+ seems to be decentralizing (more variance around the mean), for a couple of reasons:

1. Starting pitchers are being pulled earlier in games, which tends to help their ERAs;
2. Runs are being given up in low-leverage situations by marginal pitchers, which tends to raise the overall league ERA and makes the best pitchers look better compared to the league even though those runs have very little to do with the actual outcome of games.

-- MWE

One of the comments on that site mentioned one of my pet annoyances, namely that Lou Whitaker got knocked off the ballot by the 5% rule, while Alan Trammell gets steady support. I voted for Trammell as one of my four HoF choices, but I would've liked to be able to vote for Whitaker, too.

He’s eighth in ERA+ among pitchers eligible for the HOF with 3,000 or more innings

Aside from inherent problems with ERA+, Finley's also 185th in career ERA+. Those two qualifiers – eligibles with 3,000+ IP – are more magnifiers than qualifiers.

The ERA+ discussion is interesting, and I am probably guilty of using it too much so I particularly appreciate the persuasive remarks above.

For this piece though I think Rich is just pointing out how little rhyme or reason there seems to be when it comes to the non-inner circle HOF'ers. Some interchangeable players in terms of quality run the gamut from inducted, to gaining enough support to hang around the ballot, to being one and done.

Studes: what's your objection to ERA+?

What I don't like is that it doesn't provide a consistent baseline, since the denominator is the pitcher's own ERA. If a pitcher is 20% better than average, his ERA+ is 125. If he's 25% above avg, ERA+ is 133. And if 33% better, ERA+ soars to 150. And it makes bad pitchers look better than they are (25% below average yields an 80 ERA+). But this could easily be fixed, by defining ERA+ = 100 + 100*(LgERA-PitERA)/LgERA.

Seven of the top ten in ERA+ began their careers between 1986 and 1991. With Finley, Kevin Brown, and Mike Mussina among the names, it seems ERA+ isn’t the best metric to use for such comparisons

Umm, since this list is limited to non-HOFers, wouldn't you expect it to be filled with recent pitchers? More generally, I don't agree that ERA+ helps contemporary starters -- just the reverse, in fact. The move to having more relievers pitch in short relief has increased the "reliever advantage" over starters. That means about 1/3 of all IP are thrown by pitchers who are basically hitting from the ladies' tees, making league ERA an unfair benchmark for starting pitchers. These days, a 100 ERA+ from a starter is pretty darn good. In contrast, 40 or 50 years ago relievers threw fewer IP and had less of an advantage, so league ERA was a more appropriate reference point for evaluating starters.

Seven of the top ten in ERA+ began their careers between 1986 and 1991. With Finley, Kevin Brown, and Mike Mussina among the names, it seems ERA+ isn’t the best metric to use for such comparisons.

As GuyM pointed out, of course this is true before these players have a chance to be elected.

http://www.baseball-reference.com/pi/shareit/yyxE

I'll assume Clemens, Maddux and the Unit are in. Of Schilling, Brown, Smoltz and Mussina, at least one will make it. Glavine at 12 is in. So those five are gone.

Of the remaining list, 3 of the top 10 are the most recent era, and 5 of the top 20 (includes Orel, but not Reuschel).

This isn't a horrible distribution.

Yet another link to an article about the HOF. Will someone please explain why it really matters who does or doesn't get into the HOF. I'm a lifelong baseball fan. If a game's on TV, I'll watch it, even if it's not the Braves, which are my team, but I have never been able to undersand why anyone cares who's in the HOF, yet people argue over it like a bunch of Catholic cardinals arguing over who should and shouldn't be deemed saints.

Will someone please explain why it really matters who does or doesn't get into the HOF.

It doesn't, really, but what else is there to talk about on Dec. 31?

It doesn't, really, but what else is there to talk about on Dec. 31?

My wife would argue that how to make a good appletini is a valid discussion topic...

It doesn't, really, but what else is there to talk about on Dec. 31?

I'm focusing on last minute additions to my 2008 Death Pool list.

I am going to start popping into threads I don't care about and assert that I do not care.

Good stuff.

Personally, I'm way more interested in who gets into the Hall than who gets sainted. (I was going to use "gets beatified" but I'm not sure that beatification actually makes someone a saint...)

(I was going to use "gets beatified" but I'm not sure that beatification actually makes someone a saint...)

Beatification is a lesser honor than sainthood.

Is Jim hust referring to non-HOFers up top?

It doesn't, really, but what else is there to talk about on Dec. 31?

Darin Erstad, So Taguchi, Jim Leyritz, Miguel Olivo and Oscar Villarreal show up in Rotoworld's current "Top Ten Player Searches." It's hard to get deader than that.

More generally, I don't agree that ERA+ helps contemporary starters -- just the reverse, in fact.

Then why are the ERA+ charts dominated by active pitchers?

The list of single-season ERA+s over 200 since the live-ball:

Lefty Grove, 1931
Billy Pierce, 1955
Bob Gibson, 1968
Ron Guidry, 1978
Doc Gooden, 1985

Roughly one per generation. After 1990:

Clemens, 1990
Maddux, 1994
Maddux, 1995
Brown, 1996
Clemens, 1997
Pedro, 1997
Pedro, 1999
Pedro, 2000
Pedro, 2002
Pedro, 2003
Clemens, 2005

I don't think that an incredible constellation of talent (which this list obviously represents) is enough to explain the disparity away. Prior to 1990, appearing on this list meant that you had one the absolute best seasons in the history of the game, and it was never done by Koufax or Seaver or many other greats.

Explanations? Expansion? Only needing to go 6-7 innings? Different run scoring environment?

Or, to make it even more simple, consider some of the names with a career ERA+ of 127:

Kevin Brown
Curt Schilling
Bob Gibson
Tom Seaver
John Smoltz

All incredible pitchers. But the two names from a previous era were arguably the best pitchers of their generations. Smoltz, Schilling and Brown are all pitchers whose HOF candidacy will be debated because they were always a step below the league's very very best. If ERA+ is correct, and they were all of a similar quality, then it was just a quirk of history that superhumans like Maddux, Clemens and Pedro existed in the 90s and prevented Brown/Schilling/Smoltz from being considered the very very best pitchers of their generation. Or, ERA+ is skewed, and Gibson/Seaver ought to be closer to Maddux/Clemens.

Wouldn't it be easier for a good pitcher to put up a good ERA+ in a high-scoring environment than in a low-scoring environment? It seems to me that there'd be a lot more skew around a league ERA of 4.5 than a league ERA of 3.5.

Then why are the ERA+ charts dominated by active pitchers?

That's a fair point. It probably is somewhat easier to allow 2.50 R/G in a 5.00 league than to allow 1.80 R/G in a 3.60 league. In the latter scenario, a great pitcher will not infrequently pitch a game that is "better than a shutout" -- allowing an OPS against better than you need to (typically) hold a team to zero runs. But since you can't post negative runs allowed, the full impact of that performance isn't captured in ERA, creating a kind of "floor" for ERA even in low-run eras.

Still, that's mostly a concern for single-season records, which are more extreme. I don't think it's a lot harder to have a career ERA+ of 125 in a 4.50 era (3.60) than in a 4.00 era (3.20). Also, the extreme run environment of 1994-2000 has abated. So while a slight adjustment is probably appropriate when comparing pitchers from the 1994-2000 period to those in the 1960s, that doesn't invalidate the metric as a pretty useful tool in general.

I would like to echo the sentiment of number 5 and plead guilty to over using ERA+ in numerous discussions. What would be a better park/league adjusted rate stat though?

Wouldn't it be easier for a good pitcher to put up a good ERA+ in a high-scoring environment than in a low-scoring environment? It seems to me that there'd be a lot more skew around a league ERA of 4.5 than a league ERA of 3.5

One problem with this idea is that a lot of the top ERA+ seasons are from the deadball era, while very few are from the high-scoring years of the 1920s and 30s.

Darin Erstad, So Taguchi, Jim Leyritz, Miguel Olivo and Oscar Villarreal show up in Rotoworld's current "Top Ten Player Searches." It's hard to get deader than that.

I'm guessing unintentional and yet...

Ease of racking up a fine ERA+; it is NOT high or low-scoring environ.

It IS that with a five-man rotation and 7 relievers per team, you have a lot of poor pitchers getting innings. It's as if the majors expanded from 1965 to today not from 20 full teams to 30, but from 20 teams to 38 pitching staffs and 26 hitting rosters.

from the above list:
Kevin Brown
Curt Schilling
Bob Gibson
Tom Seaver

** Modern--
Brown led his league once in ERA+, with 7 top 10 finishes overall.
Schilling - zero leagues led, 10 top 10s.
** OTOH, from the previous time period--
Seaver finished first 3 times, with 10 top 10s.
Gibson finished first twice, with 10 top 10s.

One problem with this idea is that a lot of the top ERA+ seasons are from the deadball era, while very few are from the high-scoring years of the 1920s and 30s.

That's because the range of pitcher ability was much greater in the deadball era. Over time, as competition intensifies and the quality of play improves, the gap between the best and worst players narrows. So the best ERA+ should decline over time, IF you could hold the run environment constant. BTW, that's what makes the performance of Pedro, Maddux, RJ etc. so impressive, even if the high run-scoring environment gives them a small ERA+ boost: to tower over other players today is MUCH harder than 50 or 75 years ago.

Or, to make it even more simple, consider some of the names with a career ERA+ of 127:
Kevin Brown
Curt Schilling
Bob Gibson
Tom Seaver
John Smoltz

I don't find this list that troubling. W%: Kevin Brown .594, Curt Schilling .597, Bob Gibson .591, Tom Seaver .603, Smoltz (.588). Smoltz gains a bit of an edge from his relief years, otherwise he'd be lower (AND, he's a great pitcher). Clearly, Seaver's durability/longevity puts him in a higher class, but if Schilling had as many IP at the same ERA+ wouldn't he in fact be an inner-circle guy? Gibby I think probably is disadvantaged a bit in ERA+ by pitching so exclusively in low-run seasons, but again, I don't think it "invalidates" the stat -- we just have to be smart about how we use and interpret it.

Wouldn't it be easier for a good pitcher to put up a good ERA+ in a high-scoring environment than in a low-scoring environment? It seems to me that there'd be a lot more skew around a league ERA of 4.5 than a league ERA of 3.5.

For quite a while, I've thought that this was the case, JRE. But I don't have the math skills to test it.

but again, I don't think it "invalidates" the stat -- we just have to be smart about how we use and interpret it.

Well, define "invalidate." If we reject the face value conclusion that Gibson and Schilling were of the same relative per-inning quality, it means the stat is partially not to be trusted. Comparing across eras and contexts is the entire reason for the stat to exist. That doesn't it make it forever useless, but it needs some adjustment that I'm not able to casually define or eyeball.

re: Seaver and Schilling, you are correct, give Schilling another 1,500 innings of the same quality and he's a no-brainer.

GuyM is precisely right. ERA+ remains a very useful tool in evaluating pitchers across eras. The environment (including pitcher usage) does however need to be taken into account in interpreting its meaning, in the same way that we interpret 4000 career IP differently for a starter whose career began in 1875 than one whose career began in 1975.

My theory is that you end up with a law of diminishing returns; as you attempt to drive your ERA lower and lower, it becomes progressively harder to do so. ERA+ assumes that the function of pitcher quality is linear, when it is pretty clear that it is not. IOW I doubt Greg Maddux would have allowed significantly fewer than 347 hits in 412 innings ('94-'95) if he pitched in the 60's.

So why did we not see any guys, other than Grove, with an ERA+ > 200 in the 20's and 30's? Because strikeouts were much lower and (absent a nuclear sinker and movement and location from Hell like Maddux had) you need tons of strikeouts to do that-and few outside of Grove and Dazzy Vance struck out many guys in that era. Fast forward to 90's/00's and the very best starters all (except for G.M.) had insane K rates in those seasons.

Of Schilling, Brown, Smoltz and Mussina, at least one will make it. Glavine at 12 is in.

I'd say that Brown's appearance in the Mitchell Report pretty much torches what would have been a borderline case anyway. I'd be very surprised if Schilling and Smoltz didn't make it, between Schill's postseason heroics in both Phoenix and Boston and high visibility in general, and Smoltz's impressive dual act and the fact that he played on one of the dynastic teams of the 1990s. I sometimes wonder if Mussina's rep as a clean, smart kind of player might help him in the context of the steroid era. I'm not sure it should, and allegations may come to light that he was using as well, but if they don't, I imagine a lot of sportswriters will look at an admittedly great career and essentially give him a bonus for being a well-shorn Stanford grad -- sort of the opposite of the effect that will no doubt torpedo Brown's candidacy.

My theory is that you end up with a law of diminishing returns; as you attempt to drive your ERA lower and lower, it becomes progressively harder to do so.

This is exactly how I would express it.

Chuck Finley also got really screwed by his defenses. Despite being a fly ball pitcher (47% GB), Finley suffered defenses that allowed a BABIP of .300 behind him.

Despite being a fly ball pitcher (47% GB), Finley suffered defenses that allowed a BABIP of .300 behind him.

47% GB is NOT a flyball pitcher. That's about league average.

-- MWE

Are you sure, Mike? Any times I look at GB/FB splits for multiple players, there's more GB than FB. From memory (back when espn.com used to have easy-to-find GB:FB splits for teams) there are 6 groundballs to every flyball.

Of Schilling, Brown, Smoltz and Mussina, at least one will make it. Glavine at 12 is in.

Schilling & Smoltz are in. Mussina will have to wait for the VC. Brown better not plan any celebrations.

As long as we're kvetching about ERA+ - let's not forget UER disparity. Check out UER% for Kevin Brown & Curt Schilling to get an inkling of what I'm talking about.

I think a lot of problems would be solved simply if the standard deviation were taken into account rather than just the mean ie something like (pitcher ERA - mean ERA)/(stdev ERA). This would be an rough way to normalize for the different run environments and wouldn't be that hard to do (you'd have to fiddle a bit with minimum innings pitched).

You could get a lot more complicated than this of course but the point of using ERA or ERA+ is not that it is the best measure of performance but that it's a simple intuitive stat and you don't want to lose that...

If we reject the face value conclusion that Gibson and Schilling were of the same relative per-inning quality

Why would we do that? Now, I think many of Gibson's best seasons are clearly more valuable than Schilling's best because Gibson was pitching 300+ IP. But why is it obvious that Gibby was better per inning? (And this is someone asking who grew up with a Gibson poster over his bed.)

BTW, Tango says that it's no easier to post a great ERA+ in high-RS environments. Not sure he's right, but he usually is.....

Are you sure, Mike? Any times I look at GB/FB splits for multiple players, there's more GB than FB. From memory (back when espn.com used to have easy-to-find GB:FB splits for teams) there are 6 groundballs to every flyball.

It's usually in the vicinity of 45 GB, 20 LD, 35 FB, give or take....

The main problem with ERA+ for HoF discussions, IMHO, is that it's a rate stat. Now limiting it to guys with at least 3000 IP helps, but the main difference between Seaver and Schilling/Smoltz is still 1500 IP. That's 5-7 seasons depending on whether you want to use a 200 IP or 300 IP baseline.

Gibson's a fairer comparison to those two (3800 IP) but, much as I loved him as a kid, Gibson's one of the more over-rated pitchers. He's kind of the Jim Rice of pitchers. He had some great seasons, he had a lot of good seasons, but he's remembered as far more dominant than he was. And he ended up with a relatively short career compared to many of his contemporaries (and it didn't really start until his age 25 season). His black ink was "just" 20 -- Seaver had 57, Fergie Jenkins had 36 and an equal amount of gray ink, Palmer 41 black and an equal amount of gray ink, even Blyleven (in a much longer career) nearly matches Gibson in black and beats him in gray. With the current comparisons, Smoltz has 34 black and only 11 fewer gray; Schilling has 42 black and an equal amount of gray; Brown has an equal amount of black and far less gray. Gibson's career comps are nice but not inner-circle at all; even his age-based comps are not impressive at all (due to his late start) until the end of his career.

Like Rice, Gibson was dominant -- for a while. Like Rice, he was intimidating. He's deservedly an HOFer (while I don't consider Rice to be one). But he does not belong in a comparison to Seaver in terms of either peak or career and I see no reason to consider him a significantly better pitcher than Smoltz or Schilling (at least in career value).

My wife would argue that how to make a good appletini is a valid discussion topic...

Of course she would. Girly drinks are always a subject for debate. Now a good stiff gin and tonic... not so much.

I do (see him much superior to S&S;above). He could hit, which helped his record and his team beyond his ERA+. Altho his spot on the all-century team surely could have gone to someone else, like Seaver or Grover Alexander or Maddux.

But why is it obvious that Gibby was better per inning?

It's quite possible that I'm overrating Gibson. I'll give you the entirety of my thought process going into that statement: Gibson was one of the top 3 pitchers of his generation (Koufax, Seaver, Marichal?) and considered a no-doubt HOFer, whereas Schilling was one of the top 10 (Maddux, Clemens, Pedro, Unit, Glavine, battling with Brown and Mussina and Smoltz) and considered a borderline HOFer.

Ease of racking up a fine ERA+; it is NOT high or low-scoring environ.

It IS that with a five-man rotation and 7 relievers per team, you have a lot of poor pitchers getting innings. It's as if the majors expanded from 1965 to today not from 20 full teams to 30, but from 20 teams to 38 pitching staffs and 26 hitting rosters.

from the above list:
Kevin Brown
Curt Schilling
Bob Gibson
Tom Seaver

** Modern--
Brown led his league once in ERA+, with 7 top 10 finishes overall.
Schilling - zero leagues led, 10 top 10s.
** OTOH, from the previous time period--
Seaver finished first 3 times, with 10 top 10s.
Gibson finished first twice, with 10 top 10s.

For a relatively short thread, this has had a lot of good posts, and the above contribution by TomH is but one of them.

Seems to me as if we've had this discussion many times before, and the consensus has been that the question of rating pitchers over generations usually comes down to differentiating "better" from "more valuable," and then realizing why the two terms aren't always interchangeable.

Since I always prefer using specific examples, I'd just say that in terms of peak value (in order not to get sidetracked by Walt's point about Gibson), the best pitchers of today may have been "better" pitchers than their predecessors from the 60's and 70's, in the sense that they had superior single season ERA+ numbers and were more "dominant" relative to their environments. But in terms of "value" over the course of a year that "superiority" gets negated to a great extent by the lack of complete games, i.e. by the fact that their teams needed to hire a collection of relief pitchers to complement them. And then there's the difference between the 4-man and the 5-man rotation.

The "career value" question introduces a whole different factor into it, such as Walt raises about Gibson and as everyone acknowledges about Koufax, and I don't want to get into that. But in terms of best single seasons, I have no problem in saying that while at their peaks the Big Six of today (Pedro, Johnson, Clemens, Maddux, and to a slightly lesser extent Schilling and Brown) may have been somewhat "better" than the Big Four of the 60's (Koufax, Gibson, Seaver, and Marichal) relative to their eras, the average peak season of the latter group had a bit more "value" to their teams.

And while of course there's no way of proving it, I also feel confident in my belief that if a time machine could have taken all ten of these pitchers and dropped them at their peaks into any era, they would have been able to adjust to the contingencies of that era and dominate its hitters just as easily as they did in their actual careers. Trying to split hairs between a Koufax and a Randy Johnson for anything but career value is like trying to argue the difference between Bach and Beethoven, or Hemingway and Bellow. It's fun, but at bottom you can't really come to any rational conclusion, and you'd have to be insane to bet against either (or any) of them in any given game.

But in terms of "value" over the course of a year that "superiority" gets negated to a great extent by the lack of complete games, i.e. by the fact that their teams needed to hire a collection of relief pitchers to complement them.

1) WTF are you doing up at 6:25 on New Year's?

2) As to the issue you raise, the value question depends on the value of an additional roster slot. Let's say you replaced, e.g., Pedro, with someone capable of throwing 300 innings with multiple complete games. What would that get you? Well, we know that if you replaced the whole pitching staff with guys expected to throw complete games when possible, it would get you two-three extra roster slots that you wouldn't need to fill with pitchers. (We went from 9 man staffs to 12 man staffs.) OTOH, if you replace only one modern starter with a Gibson, you wouldn't even save that; you would just have a more rested bullpen.