I've maintained for a while the Sheffield was HOF-quality. More evidence. I wonder how that's going to go over.

I'm amazed by how low Will Clark's number is. Thought he would come off better on such a list.

although it is interesting to note how his overall score stacks up against other career-over-peak guys like Whitaker or Schilling,

I assume this is a typo, as just about all Schilling has going for him is peak; 3-4 CYA caliber seasons, a few post season heroics, and little else.

you mentioned that the NWAA required to be among the league leaders has remained essentially stable since 1970...but has the peak values attained by those leaders increased? something has to account for the considerably higher scores of the modern players. a few guesses:

--longer careers (due to modern medicine, training methods, nutrition and contracts)
--performance enhancing drugs (I have a feeling this actually plays minimal role, given some of the names on the list)
--expansion
--smaller parks

it's also worth noting that your metric *seems* to undervalue relief pitching to some degree (which, admittedly, isn't a bad thing). Note that Smoltz and Glavine share similar career numbers per season as starters (Smoltz having better K/IP, WHIP), yet Glavine is 6+ pts higher.

finally, look at that Pujols guy. Already past 30. What a machine (did anyone catch the interview with him on FSN?).

finally, look at that Pujols guy. Already past 30.

LOL, when I read this I thought you were taking a shot at his reported age.

although it is interesting to note how his overall score stacks up against other career-over-peak guys like Whitaker or Schilling,

I assume this is a typo, as just about all Schilling has going for him is peak; 3-4 CYA caliber seasons, a few post season heroics, and little else.

Schilling's career, in my opinion, is what's impressive. Almost 3000 IP at 128 ERA+. Never won a CYA, but 3 2nds and a 4th (voters don't remember "CYA caliber seasons", they remember the award winners). Never won an ERA title, but 9 seasons in the top 10. If he stays healthy, he'll pass 3000 strikeouts this season, but only lead the league twice.

Certainly sounds like a "career-over-peak" guy to me.

Schilling is a sure thing for the Hall because of the bloody sock.

I actually like this approach as a compromise between peak and career stats, it seems to come out roughly right. I'd love to see it applied backwards in history, to examine guys like Koufax who are at the heart of the peak-versus-career debate.

Interesting how closely bunched A-Rod, Griffey, RJ and Edgar are.

I like this piece and think it is an interesting effort toward balance. But I've always thought the "peak vs. career" debate was more about "greatest ever" titles than HOF worthiness. That is, a guy might get in the Hall with either a fabulous peak (Koufax) or a long, productive career (Sutton), but only be considered for "greatest ever" if he has both (Clemens). At least that is how I view it. I think both Koufax and Sutton are worthy of the Hall, but wouldn't put either in the discussion of greatest pitcher ever (I know some will with Koufax).

Well put, bunyon.

(I know some will with Koufax).

They'd be wrong. There is some argument to be made for Koufax (and Pedro for that matter) as the *best* pitcher ever. There is no argument for either right now for the title of *greatest*. Best is more what you could have done had you been able to do more, greatest is what you exactly did.

As a Hall of Merit voter and an avowed 'peak' voter I find this interesting. I just wonder what these might look like with defense involved as there are some players who would greatly benefite (Ozzie) or be lowered somewhat (Pedro Guerrero is a good case here) if defense were included. Maybe pitcher hitting shoudl also be included, not as big of a deal in today's game as pitcher's pitch fewer innings and fewer innings per game and thus get pinch hit for more often, but there are guys like Wes Ferrell, whose hitting makes him HOF worthy (and a HOMer by the way) instead of another peak centered contender.

In my HOM a player's value by something somehwat higher than average AND average, in order to seperate those players who play at an MVP level (say, a Charlie Keller type, maybe Jason Giambi today?) with those that play at an all-star level (maybe Bernie Williams?) for longer. Could be something to tinker with. Nice article though.

That should read HOM system, the HOM isn't mine.

finally, look at that Pujols guy. Already past 30.

LOL, when I read this I thought you were taking a shot at his reported age.

well read. I intended both meanings, for as much as I enjoy watching him play, I have serious doubts that he's only 26. Then again, I didn't think that was Rafael Furcal speeding past me straddling the lane line a few years back, so I could be wrong.

This is a very good article. A good reminder of just how good Edgar was.

So, Curt doesn't have any great peak years and his career comes up short. Without the bloody sock, he's in no conversation about the HOF.

nice to know you're in all their heads

Ooh, sounds a challenge. I bet any baseball fan could name 60-90% of the CYAs, but close to 0% of the second-place men. I take it you're significantly different? Or perhaps you think the HOF voters are?

(voters don't remember "CYA caliber seasons", they remember the award winners).

nice to know you're in all their heads

Actually, Toilet, it is Bill James in their heads, not me.

James' Hall of Fame Standards (which, according to bbref, "attempts to assess how likely (not how deserving) an active player is to make the Hall of Fame" (emphasis mine). The Standards awards points for winning MVP, CYA, and ROY, but only for winning.

That's HOF Monitor, not Standards.

Thanks for the comments. To address a few points:

I assume this is a typo, as just about all Schilling has going for him is peak; 3-4 CYA caliber seasons, a few post season heroics, and little else.

Schilling has only ranked among the top ten in WAA once in his career--a ninth-place finish in 2004, with 3.61 WAA. That's the lowest peak of any player in the sample with such a high career score. That's why I considered him a "career" guy.

it's also worth noting that your metric *seems* to undervalue relief pitching to some degree (which, admittedly, isn't a bad thing). Note that Smoltz and Glavine share similar career numbers per season as starters (Smoltz having better K/IP, WHIP), yet Glavine is 6+ pts higher.

I don't really know if that's because of undervaluing relief pitching. First of all, the metrics I'm using don't consider things like K/IP or WHIP, but it's true that Smoltz does have better career rate stats in terms of ERA+ as well. However, Glavine's career rates are dragged down by his first four seasons, in which he ranged from below-average to god-awful. Smoltz, outside of his first half-season, has been above average every year of his career. That's exactly the thing this metric is designed to adjust for. Glavine also has a higher peak, by the numbers I'm using--he's had six seasons over 3 WAA (versus three for Smoltz) and three over 4 WAA (Smoltz only has one).

Maybe pitcher hitting shoudl also be included, not as big of a deal in today's game as pitcher's pitch fewer innings and fewer innings per game and thus get pinch hit for more often, but there are guys like Wes Ferrell, whose hitting makes him HOF worthy (and a HOMer by the way) instead of another peak centered contender.

I did include pitcher hitting (except for the early years of Blyleven and John's career), although I don't think it's going to have a real big impact on a pitcher's career stats.

Also interesting to see how far Bonds and also McGwire rate ahead of Sosa, who as a fellow HR record chaser may have seemed to be even with them in terms of name recognition. Possibly he will be seen from a distance the way I might look at Duke Snider-a great player, but whom only Dodger fans of the Brooklyn era could have said was as good as Mays or Mantle.

By the way, look at Barry Larkin's score! And that doesn't include defensive value, which would put him in that "4 M's" range of HOF worthiness, no?

Nice idea and article. My one suggestion: I wouldn't call it NET Wins Above Average; it's more like GROSS Wins Above Average.

You are not subtracting the negative seasons, you're looking only at the positive ones -- and therefore you're looking at a gross figure. If you did subtract the negatives, you'd then have a net figure.

The discussion has some interesting points, about "best" player vs "greatest" player vs "Hall of Fame" player and the differing criteria that we might use. If indeed it is correct to say that one can make the Hall of Fame via either a high peak career (Koufax) or a long career (Sutton), I wonder how NWAA (better to call it GWAA I think) compares to those criteria; from the lists it looks like it might lean a little bit towards being more similar to a long career criterion than a high peak one.

Schilling has only ranked among the top ten in WAA once in his career--a ninth-place finish in 2004, with 3.61 WAA. That's the lowest peak of any player in the sample with such a high career score. That's why I considered him a "career" guy.

Just to clarify--2004 isn't Schilling's best year (2001 is), but it's the only year he ranked in the league's top ten. The point about him having a very low peak for such a high career score stands, though.

Nice idea and article. My one suggestion: I wouldn't call it NET Wins Above Average; it's more like GROSS Wins Above Average.

Heh. You're probably right, now that I think about it--finance is not my strong point. And "GWAA" has the advantage of being easier to pronounce as one syllable.

I think the stat is fun, but I wouldn't use it to try to determine how good someone really is. The stat has a number of ranking that I think are just plain wrong. For example, it has Mike Schmidt as marginally better than Edgar Martinez, and worse than Frank Thomas, when he was substantially better than both of them.

There's something about the list that overrates players that played the last ten years, while underrating the stars from the 80's.

For example, it has Mike Schmidt as marginally better than Edgar Martinez, and worse than Frank Thomas, when he was substantially better than both of them.

With regard to those specific examples--the problem, I think, can be chalked up to the lack of defensive adjustment. The stat treats Schmidt as if he were an average defensive third baseman, when he was considered to be one of the best defenders ever at his position. (Also, Schmidt's peak is rather low for such a great player--it doesn't help that what would have been his best season was interrupted by a strike.) As for DHs, RCAP compares them to the average hitter at their position, but doesn't include any further adjustment for their lack of defensive value. I didn't get into that because it's hard to find any agreement on how such an adjustment should be carried out, but most people would lop some value off the numbers for Thomas, Martinez, and Molitor.

"GWAA" has the advantage of being easier to pronounce as one syllable.

If you did it over replacment, GWAR is an excellent name.

I bet any baseball fan could name 60-90% of the CYAs, but close to 0% of the second-place men. I take it you're significantly different? Or perhaps you think the HOF voters are?

I bet a significant number of HOF voters actually look up, like in a book or online, a player's stats and award finishes.

Some people were wondering how Koufax would score on this system--I took a look at his numbers, and Koufax comes out with only 29.06 GWAA (if that's what we're calling it now). I didn't include his offense, and he was a poor hitter, so that would probably bring his numbers down a little further.

Maybe someday I'll try this looking at pitchers before 1972--but it would be a lot of work to collect all the data.

What would a Gross Wins Above Median list look like? Substantially different?

What would a Gross Wins Above Median list look like? Substantially different?

I wouldn't expect so--but I might not be the best person to ask that question. I don't know where one would find the median data for a league.

I personally prefer a career WAR, with adjustments for "pennants added" (which is sorta like peak) and postseason performance. Unlike some other people, I would be super-impressed by a player who was able to last 40 years (age 20 to age 60) as an avg regular every year, and would vote him into the HOF in a heartbeat...

But what I really want to say is that I completely support the "voting" system for the HOF, in which one guy looks more at "peak", while another guy looks more at "career", and yet another guy might over or undervalue fielding (relative to some hypothetical unbiased measure). This is the "wisdom of crowds" perspective. The only problem is that the actual people who get to vote for real do not appear to have any better qualifications than a below-avg primate poster.

The only "real" problem with the HOF is who is permitted to vote. It's not a matter of which stats to look at.

But what I really want to say is that I completely support the "voting" system for the HOF, in which one guy looks more at "peak", while another guy looks more at "career", and yet another guy might over or undervalue fielding (relative to some hypothetical unbiased measure). This is the "wisdom of crowds" perspective.

Oh, I agree, and I wouldn't suggest that this stat, or this approach, would be the be-all and end-all of the discussion. People are going to have different things they value when determining greatness. I just thought that this was a method that was worth looking at, and takes into account several different things that people look at.

Nice to see a studies article, it has been too long.

Anyone know what happened to primer numbers? It seemed to disappear.

For how many potential HOF caliber players do they have many years above replacement level but below average?
In otherwords how many seasons of data are actually being dropped?

Or is it more a case of you don't want a guy who has 20 years of 10 runs above replacement to look like a guy who has 10 years of 20 runs above replacement? In which case this is more like career WAR - (years_played * difference_between_replacement_and_average).

I agree with looking at the pennants won system which does represent a good way to go.

And I also think this does drive home just how much Bonds is one of the best players ever and one of the greatest players ever. Bonds has played 20 seasons so far (including the 14 games last year) and has averaged more than 6 wins above average per season. There are generally only about 15 position players in both leagues to make that many wins above replacement in any given season.

#33:

Anyone know what happened to primer numbers? It seemed to disappear

Despite the fact that I'm reading this thread, that's a big reason why I read BTF.org a lot less than I used to (and most times of the year, I've stopped reading it completely).

The new look (well, it's many months old now) that BTF has is disastrously bad. The content gets squeezed into a hard-to-read column in the middle. The old pages that I used to go to (namely baseballprimer.com, although that's now an alias) have been changed or merged or something -- I don't know what it is, I only know that when I go to "baseballprimer.com" I don't get the content that I used to get, nor the content that I want. Usually some articles from BaseballCentrist will be prominently featured; I have nothing against the central division or whatever the heck it is that BaseballCentrist is about, but it doesn't have articles that I want to read. Plus other bizarre stuff: "Gonfalon Cubs" articles keep showing up; again nothing against the Cubs but I don't want to read about the Cubs (if that is indeed what "Gonfalon Cubs" is about; what the heck IS "Gonfalon Cubs"?).

A few months ago I discovered the "Primery numbers" link that Michael mentions; that at least was a link that I could click on and know what I was going to get and know that it would usually be pretty good. But as he says, that link has disappeared!! On a few pages (I can't even begin to tell you which ones) there IS a link labelled "sabermetrics" which takes one to a good page, probably the same as the Primery numbers page. But most of the time, I cannot find that sabermetrics link.

Summarizing: Links which appear and disappear, and even when they exist, only exist on some pages and not others. Impossible navigation -- I don't even know what page is where. Worse quality content -- the old baseballprimer.com type of discussions and threads are probably somewhere, but I don't know where.

If, as I claim, I don't read BTF anymore, why am I here? Because for the past month or two there have been occasional "Looking Forward to" articles. As a Mariners fan, I've been waiting/looking for one about the Mariners. So that's why I've been clicking over to BTF once every couple of days lately. But: not only have I not found a "Looking forward to: 2006 Seattle Mariners" article, I don't even know if there has been one. (Doing a search for, say "forward 2006 seattle" turns up hundreds of hits, most of which are "Looking forward" articles for other teams.) So again the ungainly navigation of the current website is a huge negative. I have no problem with there being blogs or forums for Cubs fans and centrists and what not; but what I as a reader do need is a navigational structure that lets me avoid those topics and find the topics that I do want. BTF doesn't have that, and that's why I hardly read it anymore.

For how many potential HOF caliber players do they have many years above replacement level but below average?
In otherwords how many seasons of data are actually being dropped?

It's pretty variable, which is one reason why I wanted to take a look at this--it's not like you can apply some sort of constant percentage adjustment to all players to correct for their below-average years. Almost every player, even the true greats, has a few below-average years mostly at the beginning or end of their careers that pull their career rate stats down, but there are some with a lot more than that: Robin Yount had six below-average seasons, Rickey Henderson seven, Eddie Murray eight. Dennis Martinez, who has a very respectable 22.21 GWAA, drops down to 9.78 if you include all his below-average seasons. An average-based metric (like ERA+) sheds very little light on a player like that without taking a closer look.

One thing that does come through is that it's very, very unusual to find a player without at least a couple bad years. Tony Gwynn was above average every single year of his 20-year career, which is truly remarkable. Pedro Martinez, who played fourteen years before 2006, is the player with the second-longest career I've found in this group to never have a below-average year, and we haven't seen Pedro's decline phase yet. (Though Bonds comes as close as one can get to being above average every year of his career: he had a WAA of -0.01 in his rookie year, and has been above average every year since.)

Or is it more a case of you don't want a guy who has 20 years of 10 runs above replacement to look like a guy who has 10 years of 20 runs above replacement? In which case this is more like career WAR - (years_played * difference_between_replacement_and_average).

Eddie Murray vs. Rafael Palmeiro is an excellent example of this line of thinking. If you were simply to compare Murray's career stats to Palmeiro's, you'd probably conclude that Palmeiro was the superior player: he had only about 700 fewer PAs, and an OPS+ of 132 vs. Murray's 129. But Murray had a lot more really good years than Palmeiro did--he just happened to have some really crappy years at the end of his career as well. I might be wrong, but I think that most people would agree, even before the steroid revelations, would consider Murray the superior player. Yet an average-based metric won't capture this, and, while I don't have the numbers in front of me, I don't think a replacement-level one would either. Even JAWS ranks Palmeiro as the seventh-best first baseman ever--while I don't know where Murray ranks, that seems awfully high for a player who could rarely be considered among the game's best. And JAWS includes a peak adjustment--I think Palmeiro's raw WARP3 score would rank him considerably higher on the list.

As a Mariners fan, I've been waiting/looking for one about the Mariners. So that's why I've been clicking over to BTF once every couple of days lately. But: not only have I not found a "Looking forward to: 2006 Seattle Mariners" article, I don't even know if there has been one. (Doing a search for, say "forward 2006 seattle" turns up hundreds of hits, most of which are "Looking forward" articles for other teams.)

I'm not going to address the site's other issues, but the Mariners preview was supposed to be written by Dan Werr, who is a wonderful writer (my favorite on the site), but seems to have an issue with deadlines. If you'd been a fan of any other team, you'd have had some better luck finding a preview.

I might be wrong, but I think that most people <strike>would agree</strike>, even before the steroid revelations, would consider Murray the superior player.

You probably got what I meant there, though.

It's going to make me so sad when the HOF voters snub the Big Hurt.

I completely agree that Mike Schmidt is far more valuable than Edgar or even Frank Thomas. There's the defensive adjustment, and the baserunning adjustment that needs to be made when you're looking at two of the slowest, lumbering sluggers of the past thirty years.

If you pressed me, I'd argue Schmidt had a more valuable career than Roger Clemens as well.

My point still stands (in my mind at least). Edgar had a heckuva career with the bat. Most casual fans don't realize it.

I like the way this looks at it, I'm sure you'll get complaints about dropping subpar seasons or whatnot, but that is just a different way of looking at it. The only criteria I would have is that you have to have at least 10 seasons(even if a couple of them are subpar) this represents the minimum number of years you must have to qualify for the hof. (of course you would still want to include guys like Pujols on the list just to show them off)

This list seems to be somewhat intuitive with what I would think. (the palmiero murray comments being exactly what I'm talking about) of course you have to start out when reading this list is that defense isn't included at all. Obviously Schmidt, Larkin, Trammel, Sandberg, Whittaker get a mental nudge upwards, but I think the hardest one to nudge is going to be Piazza, do you nudge him upwards or downwards?

glad to see there is at least one metric out there that puts Edmonds ahead of Bernie.

Even JAWS ranks Palmeiro as the seventh-best first baseman ever--while I don't know where Murray ranks, that seems awfully high for a player who could rarely be considered among the game's best.

I have Palmeiro at #6 with a JAWS of 99.5 and Murray at #5 with 101.4. The WARP3 adjustments BP made recently raised Bagwell to #2, Thomas to #7 and kept McGwire at #11. JAWS does have a peak adjustment, but it simply isn't enough to offset someone with an extremely long career like Palmeiro or Yaz, #3.

Nudge Piazza up or down? Does it matter? You could cut his score in half and it would still be higher than the tentative cutoff mark of Guerrero. :-)