White Outfielders and Hall of Fame Don't Mix

[bedir than average]

Triad, is Biggio white enough?

[Triad]

Originally Posted by bedir than average

Triad, is Biggio white enough?

You can't get much whiter than Biggio. He played over 400 games as a catcher as well, so I don't think that makes him a catcher. He's a second baseman who was a catcher for a while.

Berkman is not on track for the Hall of Fame, in my estimation. Even with All-Star players, most of them hit about 75% of their home runs by age 32. If Berkman were a good fielding outfielder for the better part of his career, then he'd have a so-so chance, but otherwise he doesn't seem to have much of a chance as a first baseman. If he were a good-fielding center fielder like Dale Murphy, whose offensive numbers are very similar to Berkman's so far, then I'd say he'd have a decent shot.

[nanwynnfan]

"Berkman is not on track for the Hall of Fame, in my estimation. Even with All-Star players, most of them hit about 75% of their home runs by age 32. If Berkman were a good fielding outfielder for the better part of his career, then he'd have a so-so chance, but otherwise he doesn't seem to have much of a chance as a first baseman. If he were a good-fielding center fielder like Dale Murphy, whose offensive numbers are very similar to Berkman's so far, then I'd say he'd have a decent shot."

Sorry, Triad; but the above statement is ludicrous. If there is one statistic that demonstrates a player's overall offensive contribution, at the elite level, it is the percentage of Runs Created to Plate Appearances.

Before listing the players that have, since 1901, achieved a level of 19% throughout a career, or career-to-date for active players, let's just say that as a part of the the overall batter population, they are rarer than hen's teeth.

A few observations are in order. A player who hangs on long years after his peak years are behind him will have a RC/PA < 19% because the trail off years have pulled him below that level. Of course, RC is generally associated with total bases and OB%, so power guys, or high average types with pop are the ones who get into this population. I don't believe I've omitted anyone.

Interesting to note: NO catcher or third baseman has ever sustained this level; and only one 2B or SS has accomplished the level. The men listed below are therefore mostly OF, 1B or both.

Player.......................RC/PA %

Babe Ruth.................25.61
Ted Williams..............24.33
Lou Gehrig.................23.12
Barry Bonds...............22.94
Jimmie Foxx...............22.04
Albert Pujols..............21.92
Todd Helton..............21.60
Rogers Hornsby..........21.58
Hank Greenberg..........21.19
Manny Ramirez............20.67
Mickey Mantle............20.57
Joe DiMaggio..............20.45
Stan Musial................20.15
Frank Thomas.............20.13
Alex Rodriguez............19.98
Mark McGwire.............19.96
LANCE BERKMAN..........19.91
Johnny Mize................19.44
Vlad Guerrero..............19.31
Ty Cobb.....................19.29
Chuck Klein.................19.02

It's a very difficult level to attain and maintain. Willie Mays ended his long career @ 18.95%; and Jeff Bagwell ended up at the same level. Tris Speaker rounds out a tro at that level; but he was less known for power than the other two.

At the other end of the spectrum, having had a slow career start on the front end, David Ortiz is at 18.90%. A few more seasons like his past 2 or 3 may get him there; but there are a great number of PA to be balanced.

Dale Murphy ends up at 14.47%, common among decent hitting infielders, but hardly in a class with HoF OF and 1B [who are better known for power, rather than high average OB% and speed-related skills].

[WilsonC]

In Berkman's case, it's really a question of how well he ages. He's had some HoF level seasons, and some great rate stats at this point in his career, but hasn't had enough yet for serious consideration, and brings little defensive value to the table. He'll need at least a few more peak-level seasons, which will mean sustaining his peak ability into his mid 30's. Possible, but it is also quite likely he'll start to decline and fall into the Hall of the Very Good category.

Jim Edmonds is an interesting candidate to me. His raw totals fall a little short of the glossy numbers associated with the HoF, in part due to a somewhat late start to his career as a regular, but in reviewing his career he's an interesting case. His career value looks awfully similar to Duke Snider's to me. Snider has a bit better totals due to an earlier start and a bit better durability, and Snider gets a slight edge as a hitter, but on the other hand I'd give Edmonds a definite advantage in defensive value. I'd 'probably' still give Snider the edge, but I don't consider it nearly as clear cut as I would have thought before checking Edmonds' numbers.

I'd rate his peak level about even with Murphy's (Edmonds gets the nod in rates, but Murphy evens it out with better durability) and he sustained peak or near-peak value for longer. Aside from Griffey, Edmonds may well be the best centerfielder since Mays.

[Triad]

[quote=WilsonC;131479]Jim Edmonds is an interesting candidate to me. His raw totals fall a little short of the glossy numbers associated with the HoF, in part due to a somewhat late start to his career as a regular,
He also only averaged 142 G per season in his ten most productive seasons. That's 200 missed games.

Originally Posted by WilsonC

I'd rate his peak level about even with Murphy's (Edmonds gets the nod in rates, but Murphy evens it out with better durability) and he sustained peak or near-peak value for longer. Aside from Griffey, Edmonds may well be the best centerfielder since Mays.

Interesting argument about Edmonds. I would question the points regarding their relative merits in peaks. There are a few things to look at there...

MVP voting: (top 20 finishes)
Edmonds - 4th, 5th, 14th, 17th
Murphy - 1st, 1st, 7th, 9th, 11th, 12th

How do you account for this wide gap? To what would you attribute Edmonds' lackluster MVP showing? (lackluster in terms of a HOF candidate)

While MVP voting is not a great representation of a player's true ability, it gives a rather decent snapsnot of his peak, both in length and depth.

Also, Murphy had 24% more career PA than Edmonds.

Murphy was the better base stealer. Edmonds never had more than 10 in a season, while Murphy had at least 10 seven times, getting into the 20s once and the 30s once.

And lastly, the offensive production must be taken in context of a player's team's offense.

Cardinals NL rank in R/G in Edmonds' first six seasons with them:
4th, 4th, 2nd, 2nd, 1st, 3rd

Edmonds was always in a fertile setting for offense, which helps his stats. Murphy's Braves were generally toward the bottom of the league in offense.

Isn't it curious to you that Edmonds never had a season above 137 OPS until age 30, at exactly the time he joined the Cards? And then he was above 140 for five straight years. There's a suggested correlation, and it can be partially explained by the team environment he was in.

With the Angels, his four full-season OPS average was 128. His first five years with the Cardinals averaged to 156.

[WilsonC]

Originally Posted by Triad

He also only averaged 142 G per season in his ten most productive seasons. That's 200 missed games.

Edmonds has never been the most durable player, but he's also a player whose reputation for being injury prone is a bit overblown. He's generally missed between 10 and 30 games a year due to injury, which is not a trivial amount, but is also not excessive. It seems about on par with Gwynn's durability, and it wasn't really a major consideration for Gwynn's candidacy. Durability is a strike against Edmonds. Had he averaged 154+ games a season I'd likely view him as a no-brainer, rather than an interesting candidate.

Note that "ten" is bolded. That is to emphasize the fact that Edmonds had ten productive seasons. That's the big strike against Murphy. He had five very productive seasons, maybe two or three other somewhat productive ones, and drops off very quickly. If a player's hitting .226/.313/.421 from a corner outfield position, is it really that valuable if he plays in 156 games? He's very light on the number of good seasons beyond his peak.

Originally Posted by Triad

Interesting argument about Edmonds. I would question the points regarding their relative merits in peaks. There are a few things to look at there...

MVP voting: (top 20 finishes)
Edmonds - 4th, 5th, 14th, 17th
Murphy - 1st, 1st, 7th, 9th, 11th, 12th

How do you account for this wide gap? To what would you attribute Edmonds' lackluster MVP showing? (lackluster in terms of a HOF candidate)

While MVP voting is not a great representation of a player's true ability, it gives a rather decent snapsnot of his peak, both in length and depth.

Griffey, Bonds, Pujols. While he wasn't quite an MVP candidate with the Angels, he was a very, very good player. Griffey was the golden standard in CF - better than any CF since Mays - and Bernie Williams was probably a better player than the younger Edmonds, but I'd consider Edmonds to be arguably the third best CF in Baseball from 95-98. He was overshadowed by two better players at the position, but he was very good.

Starting in 2000, he went from very good to great. I would rate him as the top CF in the 2000-2005 stretch. There are a few evident reasons as to why he didn't fare too well in the MVP voting during this stretch:
- Bonds, a power hitting outfielder, put up numbers ridiculously above anyone else in the league in one of the best peaks in history, basically monopolizing the award when healthy.
- Pujols, a teammate, was also ridiculously great during this stretch, overshadowing Edmonds.
- While he was recognized as a plus defender, defense is rarely a major consideration for outfielders in the MVP voting.
- He often missed just enough time to fall short of shiny RBI numbers.

Also, bear in mind that these are the very same voters who picked Tejada over A-Rod for the 2002 MVP, among other cases where a better player clearly lost out on the award simply by merit of the team he played for, his RBI total, wheter he made a good September story, or whatever other reasons for voting. Awards can be a very, very loose starting point - MVP winners generally had good years, multiple Gold Glove winners can generally field, at least early in their career - but should not be trusted over more objective metrics.

I actually don't have any issues with either of Murphy's MVP awards. I likely would lean toward Gary Carter or Mike Schmidt in 1982, but Murphy was a solid choice, and I probably would have picked Murphy in 1983. I'd also not even consider Edmonds for an MVP in any of his best years, because Bonds and Pujols had other-worldly years throughout that stretch. I think Murphy's peak was comparable to Edmonds', but without a true juggernaught having his best years in the NL during Murphy's prime. The difference is that Murphy had a very nice peak with about 5 excellent years, and dropped off the table contributing very little past his top 7 or 8 years. Edmonds had about 5 excellent years too, of similar quality, but also had about 6 very nice years not too far below his peak.

Originally Posted by Triad

Also, Murphy had 24% more career PA than Edmonds.

Sorry, can't have it both ways. Murphy had over 2300 AB from his age 32 season on, in which he hit .234/.307/.396 - mainly as a corner outfielder. Focusing on peak is fine, but if we're going to look at longevity beyond a player's peak, the quality of play needs to be considered. Murphy was a fringe player to accumulate those 2300+ AB, whereas up until this year, Edmonds has been very effective past his prime.

Originally Posted by Triad

Murphy was the better base stealer. Edmonds never had more than 10 in a season, while Murphy had at least 10 seven times, getting into the 20s once and the 30s once.

No argument there.

Originally Posted by Triad

And lastly, the offensive production must be taken in context of a player's team's offense.

Cardinals NL rank in R/G in Edmonds' first six seasons with them:
4th, 4th, 2nd, 2nd, 1st, 3rd

Edmonds was always in a fertile setting for offense, which helps his stats. Murphy's Braves were generally toward the bottom of the league in offense.

Isn't it curious to you that Edmonds never had a season above 137 OPS until age 30, at exactly the time he joined the Cards? And then he was above 140 for five straight years. There's a suggested correlation, and it can be partially explained by the team environment he was in.

With the Angels, his four full-season OPS average was 128. His first five years with the Cardinals averaged to 156.

First off, I've yet to see a study that's successfully isolated any correlation between team offense and individual statistics (beyond things like R/RBI). In extreme cases, we can see a subtle inverse relation between team offense and individual OPS, in that a premium hitter in a poor lineup will be intentionally walked more often thereby raising his OBP, but that's about all I've seen.

I don't find Edmonds' career pattern to be particularly odd. His numbers suggest a pretty clear change in approach. His power went up, but so did his K and BB rates, suggesting he became better at working deep into counts and waiting on his pitch to drive. It could be a hitting coach, it could be learning from McGwire, or it could just be maturation as a hitter. The reason's less important than the result - he simply became a better hitter.

One more point that's barely been touched on here: Defense.

Murphy won five Gold Gloves. However, looking at BP's metrics (Fielding Runs Above Average (FRAA)) he was very good for one of those years, but below average the other four. He shows as awful defensively the last two times he won, and given his decrease in SB totals those years as well as his shift to a corner the following year in 1987, my feeling is that he probably wasn't a very good defensive centerfielder by 1985. For his career, he shows at 60 runs below average defensively (including all positions, not just his time in center), and 212 runs above replacement.

Edmonds, on the other hand, has won 8. FRAA has him below average in only one of those years, and typically good to very good. Not Andruw Jones good, but a definite defensive asset. He was still in CF at 37 this year, and shows up at 125 runs above average defensively, and 378 runs above replacement. Despite having played fewer games, Edmonds has about 50% more games in CF than Murphy. To me, it certainly appears that Edmonds has a substancial edge on the defensive side of things.

[Triad]

Originally Posted by WilsonC

That's the big strike against Murphy. He had five very productive seasons, maybe two or three other somewhat productive ones, and drops off very quickly. If a player's hitting .226/.313/.421 from a corner outfield position, is it really that valuable if he plays in 156 games? He's very light on the number of good seasons beyond his peak.

I think this factor is overblown, and is not as important as at first blush. The good seasons don't tell us very much about a player's overall worth. It's just my philosophy, but the seasons where a player is in the range of 50% or more above the league average in certain respects, that's much better than two season at 25% above, or five seasons at 10% above. I'll have to show a basic diagram of what I'm talking about. This is why I think it's too easy to give too much credit to players who had many good seasons but were rarely or never dominant (Blyleven, Niekro, Ryan) in individual seasons. I do think Ryan belongs in the Hall of Fame, by the way, but that he's overrated by the casual fan.

Originally Posted by WilsonC

Starting in 2000, he went from very good to great. I would rate him as the top CF in the 2000-2005 stretch. There are a few evident reasons as to why he didn't fare too well in the MVP voting during this stretch:
- Bonds, a power hitting outfielder, put up numbers ridiculously above anyone else in the league in one of the best peaks in history, basically monopolizing the award when healthy.
- Pujols, a teammate, was also ridiculously great during this stretch, overshadowing Edmonds.
- While he was recognized as a plus defender, defense is rarely a major consideration for outfielders in the MVP voting.
- He often missed just enough time to fall short of shiny RBI numbers.

But Murphy outpolled Edmonds in their best MVP seasons by 3, 4, 7, 8, 10+, 10+. You're suggesting that there were this many otherworldly players each season in the National League who stole the limelight from Edmonds, or that the voters shafted Edmonds sometimes by 7, 8, or 10 spots below what he should have gotten.

Originally Posted by WilsonC

Also, bear in mind that these are the very same voters who picked Tejada over A-Rod for the 2002 MVP, among other cases where a better player clearly lost out on the award simply by merit of the team he played for, his RBI total, wheter he made a good September story, or whatever other reasons for voting.

I know, it does get ridiculous sometimes. Don't get me started on Maury Wills. Or Terry Pendleton. However, it's still a good approximation. While the voters can shaft a player here and there by a few slots in the voting, it doesn't explain how a player only gets in the top ten twice.

Originally Posted by WilsonC

The difference is that Murphy had a very nice peak with about 5 excellent years, and dropped off the table contributing very little past his top 7 or 8 years. Edmonds had about 5 excellent years too, of similar quality, but also had about 6 very nice years not too far below his peak.

Edmonds did have more breadth, but I'd argue that he had less depth. More on this later.

Originally Posted by WilsonC

Sorry, can't have it both ways. Murphy had over 2300 AB from his age 32 season on, in which he hit .234/.307/.396 - mainly as a corner outfielder. Focusing on peak is fine, but if we're going to look at longevity beyond a player's peak, the quality of play needs to be considered. Murphy was a fringe player to accumulate those 2300+ AB, whereas up until this year, Edmonds has been very effective past his prime.

And Edmonds has about 1400 AB of padded stats himself, so the 2300 of Murphy isn't quite so bad when viewed in that light.

Originally Posted by WilsonC

First off, I've yet to see a study that's successfully isolated any correlation between team offense and individual statistics (beyond things like R/RBI).

Comparitive OPS

_______AVG__2006..2003..2000..1997..1994..1991..1988..1985..1982..1979..1976..1973
None ---.715-----.753---.739---.769---.745---.748---.688---.676---.698---.696---.712---.664---.691
Men On-.751-----.788---.775---.798---.770---.782---.734---.721---.735---.735---.746---.703---.719

I'm not sure to what degree this translates into better production overall for a batter, but the raw value is that it gives about a 5% advantage to a batter who's batting with men on base.

I think that's just one manifestation of it. Also, a player on a better offensive team will get up to bat more, will knock the starting pitcher out of the game earlier and get to the bullpen quicker, letting him face more mediocre pitchers. In this sense, success breeds success and the rich get richer. Another factor is that if a player plays for a team with good pitching, he doesn't have to face his own pitchers, whereas a player on a team that has bad pitching does have to face those pitchers, plus that player doesn't get to face his own bad pitchers. This would be more of a factor when there were only eight teams in each league, but it can still come into play somewhat.

Originally Posted by WilsonC

I don't find Edmonds' career pattern to be particularly odd. His numbers suggest a pretty clear change in approach. His power went up, but so did his K and BB rates, suggesting he became better at working deep into counts and waiting on his pitch to drive. It could be a hitting coach, it could be learning from McGwire, or it could just be maturation as a hitter. The reason's less important than the result - he simply became a better hitter.

Or steroids! (just kidding) I think the K and BB rates often go hand in hand. He was probably pitched around a little more when he started establishing himself as an offensive force.

Originally Posted by WilsonC

To me, it certainly appears that Edmonds has a substancial edge on the defensive side of things.

Agreed. Edmonds is probably a little overrated defensively because of his acrobatic catches, but he was still very good. (Hey, Jim, maybe if you didn't play so shallow you wouldn't have to dive over the shoulder for so many balls)

[WilsonC]

Originally Posted by Triad

I think this factor is overblown, and is not as important as at first blush. The good seasons don't tell us very much about a player's overall worth. It's just my philosophy, but the seasons where a player is in the range of 50% or more above the league average in certain respects, that's much better than two season at 25% above, or five seasons at 10% above. I'll have to show a basic diagram of what I'm talking about. This is why I think it's too easy to give too much credit to players who had many good seasons but were rarely or never dominant (Blyleven, Niekro, Ryan) in individual seasons. I do think Ryan belongs in the Hall of Fame, by the way, but that he's overrated by the casual fan.)

It's fair to give more weight to a peak, but to what extent? Is Mattingly a HoFer? Will Clark? Brian Giles? Do we look at a player's five best years and disregard the rest of his career? Of course not.

Now, I'm not talking about the years tapped onto a player's career where he's just hanging on. I'm talking about a player having a sustained stretch beyond his peak where he's in the top 10-15% at his position. A player who has a relatively short productive portion of his career can be a fine selection - if the peak is above and beyond most HoFers peaks.

One way of approaching this evaluation is to use a ten year baseline. Why ten years? Because it's the minimum requirement as per the HoF rules. If you take an average of a player's ten best years and convert if to a single seasonal notation, you'll get an average that looks at a player's extended peak. If this average looks like a HoF caliber season, it's a good starting point. A player with a top-heavy peak and with less than ten particularly productive years will make the cut if his peak years are far enough above the pack to offset his weaker years, but will fail if he has too few peak years or those years aren't enough to carry a short peaks or a peak that's not high enough. A player who's consistently good but never great will have an average that looks like one of his typical years, which falls short. I'd put Cecil Fielder in the first category and Harold Baines in the second of players who very clearly fall short by this standard.

How do Murphy and Edmonds fare here?
Murphy: (1979, 1980, 1982-1988, 1990)
G: 154...AB: 570...H: 158...2B: 26...3B: 3...HR: 32...R: 95...RBI: 94...SB: 13...CS: 5...BB: 77...K: 125...BA/OBP/SLG/OPS: 0.276/0.363/0.502/0.865

Edmonds: (1995-1998, 2000-2005)
G: 142...AB: 500...H: 147...2B: 33...3B: 2...HR: 32...R: 99...RBI: 93...SB: 5...CS: 4...BB: 77...K: 128...BA/OBP/SLG/OPS: 0.295/0.389/0.560/0.949

What's interesting is how eerily similar these two players' raw counting numbers are over their best 10 years. Edmonds clearly has the better rates (these are not context adjusted), and Murphy the better durability. I'd lean toward Edmonds based on this, but would need to see the context adjustments. Defensively, there's no question Edmonds wins here.

*edit - I checked the adjusted OPS+ numbers for these years, and Edmonds comes out substancially higher on that rate too - 143 to 132 in Edmonds' favor. By rate, Edmonds 10 yead average OPS+ comes out ahead of Murphy's first MVP year. Edmonds has 3 years with an OPS+ higher than Murphy's best OPS+, and Edmonds best 5-year average is almost identical to Murphy's best OPS+ of 157 (a fact that's even more significant since Murphy wasn't even a centerfielder anymore at the time).

Originally Posted by Triad

But Murphy outpolled Edmonds in their best MVP seasons by 3, 4, 7, 8, 10+, 10+. You're suggesting that there were this many otherworldly players each season in the National League who stole the limelight from Edmonds, or that the voters shafted Edmonds sometimes by 7, 8, or 10 spots below what he should have gotten.

I'm suggesting that there were two players, Bonds and Pujols, who were better than any players in the NL during Murphy's peak - and that since one of those is a teammate of his, it eliminated Edmonds from getting many "Best player on a contending team" type votes. Beyond that, I'd suggest the following factors to be in play:
- Voters like RBI's, and Edmonds lacked gaudy RBI numbers
- Current voters seem to undervalue defensive contributions for outfielders, so there are a fair number of corner guys who rated higher than him, but likely should not have.
- It's a higher octane offensive era, meaning that there is more potential for variance from year to year. That means more guys sneak in ahead of him in a typical year - possibly even deservingly - than in a lighter offensive era.

Originally Posted by Triad

I know, it does get ridiculous sometimes. Don't get me started on Maury Wills. Or Terry Pendleton. However, it's still a good approximation. While the voters can shaft a player here and there by a few slots in the voting, it doesn't explain how a player only gets in the top ten twice.

Absent of any other evidence, you'll find a pretty strong correlation between voting results and performance. It's not a reliable measure, though, as factors such as team quality, teammate quality, RBI biases, market biases, drama, the number of players who happen to have career years in a given year, the number of times he made the highlight reel in September, etc. can all affect voting results. Whether we can determine whether a player is suject to certain biases, I think we can agree that biases do exist. As a starting point, fine, but when we have at our disposal numerous metrics with which to evaluate a player's peak, both in isolation and in comparisson to his pears, why put any significant weight on the voting process?

Originally Posted by Triad

And Edmonds has about 1400 AB of padded stats himself, so the 2300 of Murphy isn't quite so bad when viewed in that light.

Which 1400 AB are you refering to? 2007 is really the only year that I'd classify as 'padding'. He was above average still in 2006. With Murphy, he just wasn't a productive hitter after 1987, and played a corner position. His perfomance alone likely wouldn't have even meritted a starting job on many teams past 1987.

Originally Posted by Triad

Comparitive OPS

I'm not sure to what degree this translates into better production overall for a batter, but the raw value is that it gives about a 5% advantage to a batter who's batting with men on base.

I think that's just one manifestation of it. Also, a player on a better offensive team will get up to bat more, will knock the starting pitcher out of the game earlier and get to the bullpen quicker, letting him face more mediocre pitchers. In this sense, success breeds success and the rich get richer. Another factor is that if a player plays for a team with good pitching, he doesn't have to face his own pitchers, whereas a player on a team that has bad pitching does have to face those pitchers, plus that player doesn't get to face his own bad pitchers. This would be more of a factor when there were only eight teams in each league, but it can still come into play somewhat.

I'm curious as to how those numbers were generated. If the numbers are somehow adjusted to compensate for variance in opportunity, or are consistent with typical individual player differencials, it could have some merit as a subtle adjustment factor. If it's simply a league-wide comparasson of performance with and without baserunners, however, there's a hefty selection bias:
- Leadoff hitters typically have fewer opportunities with men on, and rarely have the power to put up big OPS totals.
- Teams tend to put their best hitters in the middle of the lineup, and their best OBP guys at the top of the lineup. This leads to a disproportionate number of PA with men on base going the better hitters. If those are raw numbers, we'd expect that kind of result, by design.

Now, I have no doubt that there are subtle factors of team strength that can have an effect, but the key is subtle. Far more subtle than, say, park effects. And speaking of park effects, let's take a look at how our two outfielders were effected by their parks, both for their career, and in their peaks:

Murphy:
Career:
Home - 0.281/0.368/0.499/0.867...........Road - 0.250/0.324/0.440/0.764
1982-1987: (avg park factor: 106.5)
Home - 0.310/0.410/0.572/0.982..............0.269/0.355/0.493/0.848

Edmonds:
Career:
Home - 0.291/0.386/0.541/0.927............Road - 0.283/0.372/0.521/0.893
2000-2005: (avg park factor: 100)
Home - 0.293/0.407/0.599/1.006.............Road - 0.291/0.403/0.569/0.972
1995-1998: (avg park factor: 100.25)
Home - 0.304/0.375/0.537/0.912............Road - 0.292/0.356/0.518/0.874

Wow! Murphy sure did like hitting at home! Are we sure he wasn't playing in Colorado? Murphy played in a pretty significant hitters park, and gained an enormous advantage at home. Edmonds, on the road, had a higher OPS in his Angels 'sub-peak' than Murphy did in his true peak. Now, with era adjustments and durability, I think Murphy's peak was still better than Edmonds' 'sub-peak', but given Edmonds defensive advantages, the gap is smaller than it appears at first glance. Edmonds did benefit at home, slightly, but was relatively neutral overall with his splits.

I can see giving Murphy some credit for using his home park better than most, but it's pretty clear that he played in a hitter-friendly environment, much more so than Edmonds.

Originally Posted by Triad

Or steroids! (just kidding) I think the K and BB rates often go hand in hand. He was probably pitched around a little more when he started establishing himself as an offensive force.

hehe, steroids are always a possibility! 8P There are definitely some correlations between these that make it hard to determine everything through numbers. We also get a chicken-and-the-egg scenario, where we need to ask whether a player walks more because he's being pitched around due to better production, or if he has better production because he's being more disciplined and waiting on the right pitch to hit. Ultimately, the reason would probably require a scouting analysis, whereas the value is independant of the reason.

Originally Posted by Triad

Agreed. Edmonds is probably a little overrated defensively because of his acrobatic catches, but he was still very good. (Hey, Jim, maybe if you didn't play so shallow you wouldn't have to dive over the shoulder for so many balls)

I'm not sure what other metrics say about him, but BP's metrics are pretty close to this assessment. He's shown as well above average, but well short of the Andruw Jones defensive superstar. Not sure how many GG it translates to, but it's a fair it of value.

[Triad]

Originally Posted by WilsonC

It's fair to give more weight to a peak, but to what extent? Is Mattingly a HoFer? Will Clark? Brian Giles? Do we look at a player's five best years and disregard the rest of his career? Of course not.

Mattingly's peak is too short. Clark's not a bad choice, but others not in are more deserving. Giles' peak was also too short.

It's not that we look at a player's best 7 years per se, but we look at his years where he was substantially above the league norm, however many that may be. And for most borderline Hall of Fame-type players, this would be somewhere in the range of 6-8 years. This is why Albert Pujols, after 7 years, has already done enough to make him worthy of the Hall of Fame, ahead of other HOF first basemen like Johnny Mize, Orlando Cepeda, and Tony Perez. I rate him as the 11th best first baseman already. What more would he need to do to be at the Hall of Fame level? Have 4-5 more decent seasons and 4-5 more so-so seasons? That wouldn't change his true value to any real extent.

Originally Posted by WilsonC

One way of approaching this evaluation is to use a ten year baseline. Why ten years? Because it's the minimum requirement as per the HoF rules. If you take an average of a player's ten best years and convert if to a single seasonal notation, you'll get an average that looks at a player's extended peak. If this average looks like a HoF caliber season, it's a good starting point.

That may work a lot of times, but I think that method would overlook certain types of players. It would tend to give extra credit to players who had lower peaks but sustained them longer, and that's kind of what we're trying to get away from.

I do think you have the right idea. I would slim it down from 10 to about 8 or even 7, and then I think you'd have something substantive. It would be more realistic, being in line with the actual career trends of the lower realm of Hall of Famers. There aren't a lot of borderline Hall of Famers with 10 excellent seasons. It's closer to 7.

I think the mentality we get into is that Mays and Cobb and Ruth each had 15 or so superstar seasons, and so then we start measuring a player's HOF worthiness next to that standard, thinking that a peak needs to last at least 10 years. What happens as a result is that we give too much credit to those who had 5 great seasons and 12 average seasons, while underrating those who had 7 great seasons and 6 average seasons. The great seasons should carry most of the weight, and the average seasons shouldn't be much of a factor (even if 10-15% above the league average). Also a player's position comes into play. A corner outfielder who is slightly better than the average corner outfielder in a given year is being more productive to his team than a second baseman or shortstop who might be 20% better than the average 2B/SS in the league.

So anyway, I don't think there's a good reason for the HOF to have a 10-year minimum. 10 is a nice, round number, and almost every superstar's career lasts at least 13 years, and most are closer to 17-18, so it seems like a reasonable minimum. But I think the real reason it's set at 10 is that there hasn't been a mechanism in place for judging peak vs. career, so it's basically a cop-out so they don't have to deal with that analysis. Someday a player is going to have a stellar 9-year career, and it will force them to make an exception.

Based on these observed career trends, Manny Ramirez and Gary Sheffield were good enough for the Hall of Fame two years ago, as was Jeff Kent. They had all had enough high-quality seasons to be in line with what other legitimate Hall of Famers had done.

Originally Posted by WilsonC

How do Murphy and Edmonds fare here?
Murphy: (1979, 1980, 1982-1988, 1990)
G: 154...AB: 570...H: 158...2B: 26...3B: 3...HR: 32...R: 95...RBI: 94...SB: 13...CS: 5...BB: 77...K: 125...BA/OBP/SLG/OPS: 0.276/0.363/0.502/0.865

Edmonds: (1995-1998, 2000-2005)
G: 142...AB: 500...H: 147...2B: 33...3B: 2...HR: 32...R: 99...RBI: 93...SB: 5...CS: 4...BB: 77...K: 128...BA/OBP/SLG/OPS: 0.295/0.389/0.560/0.949

What's interesting is how eerily similar these two players' raw counting numbers are over their best 10 years. Edmonds clearly has the better rates (these are not context adjusted), and Murphy the better durability. I'd lean toward Edmonds based on this, but would need to see the context adjustments. Defensively, there's no question Edmonds wins here.

*edit - I checked the adjusted OPS+ numbers for these years, and Edmonds comes out substancially higher on that rate too - 143 to 132 in Edmonds' favor. By rate, Edmonds 10 yead average OPS+ comes out ahead of Murphy's first MVP year. Edmonds has 3 years with an OPS+ higher than Murphy's best OPS+, and Edmonds best 5-year average is almost identical to Murphy's best OPS+ of 157 (a fact that's even more significant since Murphy wasn't even a centerfielder anymore at the time).

It is kind of weird how closely their raw stats mirror one another.

Keep in mind the following about context...

Comparing their seasons with at least 400 PA, here's the breakdown per season:
(not counting 1981)

__________Edmonds . . Murphy
Tm R/seas. -- 817 . . . . . 650

The highest a Murphy team ever scored was 747 runs. Edmonds' teams topped that all 11 of his measured years. They were over 800 seven times, while Murphy was only over 700 three times.

That's a 26% advantage overall for Edmonds. But by your measurements, Edmonds scored only four more runs and had 1 fewer RBI per each of those seasons. They basically produced the same number of runs even though Edmonds was in an environment where 26% more runs were scored. We must account for this discrepancy somehow. I say it's because Edmonds was not as good of a run producer.

Here's an interesting view of the same thing, looking at a player's percentage of his team's runs each season:

tmR% of tmRR+RBI1995EdmondsCAL80114%2271996EdmondsCAL7629%1391997E dmondsANA82910%1621998EdmondsANA78713%2062000Edmon dsSTL88713%2372001EdmondsSTL81413%2052002EdmondsST L78711%1792003EdmondsSTL87610%1782004EdmondsSTL855 12%2132005EdmondsSTL80511%1772006EdmondsSTL7818%12 2tmR% of tmRR+RBI1978MurphyATL60012%1451979MurphyATL6698%1101980Mur phyATL63015%1871981MurphyATL39512%931982MurphyATL7 3915%2221983MurphyATL74617%2521984MurphyATL63215%1 941985MurphyATL63218%2291986MurphyATL61514%1721987 MurphyATL74715%2201988MurphyATL55514%1541989Murphy ATL58412%1441990MurphyATL-PHI66811%1431991MurphyPHI62912%147 - 15% or more - 12-14%

Murphy has the six highest percentages, and eight of the top ten.

What's even more striking is that while Edmonds had only five seasons at at least 12% of his team's total, Murphy had 12 seasons at at least 12% of his team's total!

You had also mentioned that Murphy's performance in 1988 and 1989 wouldn't have even merited a starting job on many teams, but notice that he produced 14% and 12% of his team's run those years.

Also note that Edmonds' percentages are rather consistent going from the Angels to the Cardinals.

Originally Posted by WilsonC

I'm curious as to how those numbers were generated. If the numbers are somehow adjusted to compensate for variance in opportunity, or are consistent with typical individual player differencials, it could have some merit as a subtle adjustment factor. If it's simply a league-wide comparasson of performance with and without baserunners, however, there's a hefty selection bias:
- Leadoff hitters typically have fewer opportunities with men on, and rarely have the power to put up big OPS totals.
- Teams tend to put their best hitters in the middle of the lineup, and their best OBP guys at the top of the lineup. This leads to a disproportionate number of PA with men on base going the better hitters. If those are raw numbers, we'd expect that kind of result, by design.

I suppose we'd need to examine all parts of the lineup. I'll look at some quick examples, isolating individual players. I'll pick some players who had borderline or near-borderline HOF careers.

____________none/men on
Fred McGriff -- .873/.900
Will Clark ----- .855/.910
Jim Thome ---- .975/.973
Larry Walker -- .960/.971
Albert Belle --- .934/.933
Dwight Evans - .829/.852
Tim Raines ---- .793/.838
Jim Rice ------ .839/.868
Brett Butler --- .754/.749
Julio Franco --- .764/.802
Dale Murphy --- .784/.852
Jim Edmonds -- .920/.897

Interesting that when I took ten random players, plus Murphy and Edmonds, Murphy had the largest positive differential with men on base (9%), while Edmonds had the lowest (-3%). I don't know that we can determine anything from a specific player's rates on this, because we'd have to look closer at different types of situations with men on base.

For our sample here, the average is about +3% with men on base. And this also wouldn't completely bear out other advantages a player on a high-scoring team would have, such as facing more middle relievers.

Originally Posted by WilsonC

Now, I have no doubt that there are subtle factors of team strength that can have an effect, but the key is subtle. Far more subtle than, say, park effects.

I don't think park effects are very well defined, at least not yet. I think there's always a significant margin of error in that data, so it's hard to go by that too much.

[WilsonC]

Originally Posted by Triad

Mattingly's peak is too short. Clark's not a bad choice, but others not in are more deserving. Giles' peak was also too short.

More on this later.

Originally Posted by Triad

It's not that we look at a player's best 7 years per se, but we look at his years where he was substantially above the league norm, however many that may be. And for most borderline Hall of Fame-type players, this would be somewhere in the range of 6-8 years. This is why Albert Pujols, after 7 years, has already done enough to make him worthy of the Hall of Fame, ahead of other HOF first basemen like Johnny Mize, Orlando Cepeda, and Tony Perez. I rate him as the 11th best first baseman already. What more would he need to do to be at the Hall of Fame level? Have 4-5 more decent seasons and 4-5 more so-so seasons? That wouldn't change his true value to any real extent.

In my view, Pujols would need to play at least one more game in three more seasons to qualify, or see the HoF make an exception. Pujols is a good example of how my methodology is intended to work. So far he's had 7 fantastic seasons - most of them better than anything Murphy or Edmonds did. If I take his career to date and plug in 3 absolutely awful years to make up the 10 years (400 AB/year at .250/.300/.400) his line comes out at:
G: 149... AB: 525... H: 164... HR: 33... R: 95... RBI: 96... .313/.394/.570/.964
That's not a 10 year peak, that's a 7 year peak that was dominant enough that, even if he hangs on as a replacement-level player for 3 years to drag down his averages, his 10 year average is still HoF level. Pujols passes the litmus test even with unreallistically awful years dragging him down because his peak is exceptional.

Originally Posted by Triad

That may work a lot of times, but I think that method would overlook certain types of players. It would tend to give extra credit to players who had lower peaks but sustained them longer, and that's kind of what we're trying to get away from.

Well, if a player didn't have a HoF peak to begin with, it would be impossible for this method to make a HoFer out of someone due to a sustained lower peak, as the average can never go higher than the best season. If a player sustains a borderline HoF peak for 10 or more years, why shouldn't he be given credit for that? I am not talking about a player 20% above average for 10 year, more like 40% above average, for example.

Originally Posted by Triad

I do think you have the right idea. I would slim it down from 10 to about 8 or even 7, and then I think you'd have something substantive. It would be more realistic, being in line with the actual career trends of the lower realm of Hall of Famers. There aren't a lot of borderline Hall of Famers with 10 excellent seasons. It's closer to 7.

The reason I went with 10 was to differenciate the borderline candidates. Someone with 6 borderline HoF years as his peak needs to do a little more in his career to make the grade, whereas someone with 6 years well above borderline or 10 years marginally above borderline will likely make the grade.

Originally Posted by Triad

I think the mentality we get into is that Mays and Cobb and Ruth each had 15 or so superstar seasons, and so then we start measuring a player's HOF worthiness next to that standard, thinking that a peak needs to last at least 10 years. What happens as a result is that we give too much credit to those who had 5 great seasons and 12 average seasons, while underrating those who had 7 great seasons and 6 average seasons. The great seasons should carry most of the weight, and the average seasons shouldn't be much of a factor (even if 10-15% above the league average). Also a player's position comes into play. A corner outfielder who is slightly better than the average corner outfielder in a given year is being more productive to his team than a second baseman or shortstop who might be 20% better than the average 2B/SS in the league.

I ag