That is, not guys who were moved one position up or down the defensive spectrum at normal ages, but guys who played at a position for which they were vastly over- or under-qualified?

Jose Reyes played half a season as a second baseman so that the Mets could keep Kaz Matsui at shortstop. Gary Carter spent most of his rookie season in rightfield, so that the Expos could play Barry Foote behind the plate.

I don't think anyone has mentioned the legendary Giants run. in 64 they had Jim Ray Hart at 3rd, McCovey in the outfield. Later in 72-73, Dave Kingman played a total of 119 games at third.

I remembered that Yount had the shoulder trouble, so I agree that the data is problematic. Still, he lost 70 runs over the next three years vs. the prior three.

So you have two players with modest declines, which you would expect simply due to aging, and one guy who had a serious shoulder injury, was forced to make a position move he wouldn't otherwise have made, and declined significantly in defensive value. Seems like your evidence, limited though it is, suggests that position changes don't impact value very much.

*

Jeffries was -19 per 150 games as a 2B, and league average at 1B. Rose is an interesting case because he had a lot of innings at a lot of positions. He was -5 at 2B, then +5 in the OF, then -10 at 3B, and then -8 at 1B. Only the last of those changes is really different from what we'd expect, and of course Rose was quite old when he played 1B.

That said, you have to take Total Zone ratings before about 1990 with a grain -- make that a shaker -- of salt. TZ really provides a very crude estimate of defensive value, and one that is way over-regressed (really good and really bad fielders will appear to be much closer to average than they really are). You would need a large sample covering a lot of players from earlier years before drawing even tentative conclusions.

So you have two players with modest declines, which you would expect simply due to aging, and one guy who had a serious shoulder injury, was forced to make a position move he wouldn't otherwise have made, and declined significantly in defensive value. Seems like your evidence, limited though it is, suggests that position changes don't impact value very much.

It would be interesting to know how much defensive value is typically lost via the aging process by all players, and whether elite players fit that profile. If A-Rod and Carew lost 6 runs/year defensively from age 27 to age 30, is that typical? It sounds a bit high to me, but I don't know, and it is likely within the SD. My sample is too tiny to be very useful, alas.

Not that it matters to the overall point, but didn't Yount win a GG in CF? I have a vague recollection that he's the only GG winner at both SS and CF, but I might be thinking of being an allstar at both (not that getting an all-star nod means anything more than a GG).

BTW, I thought I read somewhere that B-Ref was going to start using Dewan's DRS in calculating WAR, instead of Total Zone (for years that DRS is available). But when I look at players' value calculations, they appear to still use TZ. Is this change going to happen in the future? Was this a hallucination on my part? Anyone know the story?

BTW, I thought I read somewhere that B-Ref was going to start using Dewan's DRS in calculating WAR, instead of Total Zone (for years that DRS is available). But when I look at players' value calculations, they appear to still use TZ. Is this change going to happen in the future? Was this a hallucination on my part? Anyone know the story?

I think they are using DRS. Look at Andruw Jones's page. Down in the fielding table, for 2007, he's at +6 in what BB-Ref calls "Rtot" (which, if you hover over the title, is defined as "Total Zone Total Fielding Runs Above Average") and +19 in "Rdrs". If you scroll back up to the WAR table for Jones, they're showing his "Rfield" in 2007 as +19.

Not that it matters to the overall point, but didn't Yount win a GG in CF?

No, Yount only won one Gold Glove in his career, in his MVP season at shortstop (1982). Yount won MVP awards as both a shortstop (1982) and a CF (1989), which is probably what you're thinking of.

Kiki/107: thanks. I was fooled by the fact that when you hover over "Rfield" it is still credited to Sean (which is true for the pre-DRS years).

Kiki/107: thanks. I was fooled by the fact that when you hover over "Rfield" it is still credited to Sean (which is true for the pre-DRS years)

Yeah, it's very confusing; too many fielding numbers that seem like they should be measuring the same thing (I understand why BB-Ref shows them all; it's just confusing).

Returning to the myth of the "2B curse," Willie Randolph is a very good counter-example. He held up very well through his mid-30s, and appears to have still been at least a league-average player when he packed it in. (Maybe he developed a chronic injury?)

More generally, Randolph seems to be a surprisingly under-appreciated player. He had more career value than Jeter, and played on 3 post-season NYY squads (though only 1 champion), yet his reputation seems to have faded a fair amount and few seem to feel he was cheated by the HOF. Could he be that rarest of flowers, an underrated Yankee?

No, Yount only won one Gold Glove in his career, in his MVP season at shortstop (1982). Yount won MVP awards as both a shortstop (1982) and a CF (1989), which is probably what you're thinking of.

D'oh!

Yup - you're right... I was confusing three equally meaningless measures ;-)

Randolph was on 4 postseason Yankee teams, 76, 77, 80, and 81. In addition he was the starting 2B for the 1978 championship team, but was hurt for the playoffs. He's every bit as deserving for the HOF as Phil Rizzuto.

They have a lot of similarities, high OBP, great defense, no power, played up the middle for Yankee champions. Randolph needs to guest star in a Meatloaf song or something.

Randolph needs to guest star in a Meatloaf song or something.

I'm not sure that would help his reputation. Meatloaf didn't do much for Mitt Romney.

I have an acquaintance who is a die-hard Yankee fan, and despite this we're good friends. Randolph is his all-time favorite player.

I checked BB-Ref for 20 all-star levelnon-catcher position players from the last couple of decades who maintained the same defensive position from ages 26 through 31. They are Robbie Alomar, Scott Rolen, Torii Hunter, Barry Larkin, Larry Walker, Tino Martinez, Bernie Williams, Craig Biggio, Aramis Ramirez, Miguel Tejada, Derek Jeter, Cal Ripken, Chase Utley, Barry Bonds, Albert Pujols, Dan Uggla, Matt Holliday, Jimmy Rollins, Andruw Jones and Shawn Green. I chose them as they came to me, no method -- I justed wanted players roughly similar to Carew and ARod to see how much defensive decline one would have expected from them if they had stayed at their original positions (see 104).

I compared their defensive value ages 26-28 with their defensive value age 29-31. I expected to see a decline, but didn't find it. In total, this group was 75 runs better during ages 29-31 than ages 26-30. Players who had reduced defensive value as they aged were Jones, Pujols, Green, Hunter, Larkin, Walker, Williams, Bonds, Uggla, and Rollins. Holliday was the only outfielder who got better, but a majority of infielders showed improvement, especially Alomar, Ripken, Rolen and Biggio. The 13 infielders as a group improved by 140 runs, or almost 4 runs per season per player.

I intentionally chose only guys who were healthy and not changing positions, so I recognize that I biased the sample. I'm not claiming that this shows that players as a whole get better defensively as they age. But I was still surprised with the result.

Tippecanoe:
Those are interesting results. But if you're going to try to measure defensive aging, I think you really need to stick with one metric to make sure you aren't capturing methodological differences. If you are using dWAR, any player who plays before and after 2003 will have a mix of TZ and DRS ratings, which might create the illusion of decline or improvement. And any player whose career spans 1989 (I think) will be rated by two different versions of TZ, which can make a big difference. Ripken's dramatically improved ratings starting at age 28 (1989), for example, is mostly a function of changed methodology not talent.

Honestly, you're probably best off using something crude and simple, like assists for IF and PO for OF, or RF9. None of these are very good measures at the individual player level, but they have the virtue of consistency, and averaged across a bunch of players they should tell you how many fewer balls the older players are getting to.

There are very few great players who didn't hit for any power. Randolph is one of the few.

I ran this query: Born after 1920, less than 100 career HR, sorted by WAR. I like doing birth year instead of a year range, so that you always get a player's complete stat line in there. But I needed to exclude the pre-Ruthian players, as otherwise Home Run Baker would have made the list. I'm looking for the best players to not hit for any power in a power hitter's game. The results:

1. Rod Carew (77)
2. Ozzie Smith (73)
3. Randolph (63)
4. Richie Ashburn (60)
5. Luis Aparicio (51)

To make the list you need to hit for ridiculously high averages, or else get on base while playing elite defense at an up the middle position, and to play for a long time. The other 4 are in the HOF. Randolph really should be as well.

Looking at this list for active players, Dustin Pedroia, Jose Reyes, and Joe Mauer qualify for now. But each has over 90 homers and will pass the 100 mark before they accumulate too much WAR.

The only one who has a shot to be among the elite light hitters is Michael Bourn, with 19 WAR and 22 homers.

Richie Ashburn hit 29 HR. In his career! How is that possible for a MLB outfielder? He obviously had some speed, but only stole 30 bases twice. What an amazing/weird profile.

If we define "no power" as hitting fewer than 10 HR per season (per 650 PA), then that would add Boggs (118 HR), Gwynn (135), and Lofton (130) to your list. Yes, they had a little more pop than the guys on your list, but not much. And there are some guys who played entirely post-Ruth but were born before 1920 and who are under 100 HR, like Vaughan, Boudreau, and Appling.

How much data is there on true out-of-position players? That is, not guys who were moved one position up or down the defensive spectrum at normal ages, but guys who played at a position for which they were vastly over- or under-qualified?

Keith Hernandez? He was by all accounts a natural 3B or 2B, except for the pesky detail of throwing lefthanded.

I'm surprised that's never been a thing, other such players who were good fielders but stuck at 1B by lefthandedness. I guess they just all play the outfield, or become pitchers early in life.

Pujols in his twenties supposedly had 1B fielding stats on par with converted 2Bmen, implying Pujols could have played a passable 2B and been even more spectacular in overall value. (What was the story, that he preferred 1B because it took less of a physical toll?) Although that probably doesn't/won't hold true for his decline phase.

guy

richie could really run but nobody stole bases back then except willie

when he stole 32 bases in 1948 he led the nl and when he stole 30 he finished second in the league

he beat out numerous bunt singles every season

I would hypothesize that the normal values of the defensive spectrum wouldn't work for shifts of that extent. But how much data is there?

Some guys who spring to mind but maybe aren't great because they're mostly OF:

Willie Wilson played mostly LF early on (Cowens in CF). He was +78 runs in 569 starts in LF; for 1200 starts in CF (mostly older) he was +28. That's roughly consistent with what WAR might expect (all TZ I think).

Damon moved to mostly LF in 99 to make way for Beltran. He was below-average in CF before that, put up a +11 in mostly LF, moved back to mostly CF but was above-average after that. Again looks to be about right.

Howard Johnson played some SS and CF. Relative to his below-average 3B numbers, his SS numbers look too good and his OF numbers look much worse than they should be but almost all of that is CF where he understandably was a disaster.

Jose Valentin was a good fielding SS who made a lot of errors so he got shifted to 3B for ages 31-32 then back to SS for 33-34. He had 1.1 dWAR at 29-30, 1.7 dWAR at 31-32 and 2.5 dWAR at 33-34. Make of that what you will.

Pedro Guerrero was an average-ish OF in his younger days who spent most of a laughable season at 3B and parts of others (about 2 seasons worth of starts overall). TZ puts him at -29 which is possibly kind. He went on to be a below-average 1B. That 1B/3B numbers look pretty consistent, the OF numbers look much better. Maybe just a guy who couldn't transition from OF to IF.

In 1987, the Cubs decided to move Keith Moreland to 3B. It actually kinda worked so, being the Cubs, they gave up on the idea. I remember this. Moreland was a disaster as an OF, especially by that stage -- he had no range. It may actually have been the rare case where you might expect an improvement moving to the tougher position -- his lack of range didn't hurt him much at 3B. I recall him being substantially worse than TZ says but I'm willing to believe it might be right -- he was very good at sticking his big chest in front of the ball and had a good arm.

There apparently was something a bit irresistible about the idea of Bobby Bonilla at 3B because he spent more innings there than in the OF (I was surprised to discover). He was a bit better at 3B than in the OF corners in TZ so that's inconsistent with dWAR. Perhaps also a limited range "benefit" of 3B.

Miguel Cabrera was a bit below average in (mostly) LF then was a predictable disaster at 3B (about in line with dWAR). His dWAR stayed pretty much unchanged with the move to 1B. Then, exactly as I predicted, he was an unmitigated disaster at 3B last season. Oh wait.

Tippercanoe -- just use dWAR -- it adds Rfield and Rpos for you.

Like I said, we should be smart enough not to take dWAR literally. Sure, there's no reason to think that Ozzie would have been +42 at 1B -- as Guy notes, it's probably not even possible to be +42 there. Similarly there's no reason to think that Derrek Lee at SS is a particularly good idea ... although surely a much better idea than Prince or Manny at SS.* Yes, it probably will underestimate how horrible a HoJo or Sheffield at SS will be. But it should be a pretty good guide for standard, sensible defensive moves.

*Somebody here did run a DM sim of Bonds at every (non-pitcher?) position. The defense was predictably horrible, I think Bonds the C had over 200 passed balls. The fact that the team was scoring about 18 runs a game helped them to a pretty good record as I recall. Step 1 of "find 8 Barry Bonds" is perhaps the challenging part of that strategy.

*Somebody here did run a DM sim of Bonds at every (non-pitcher?) position. The defense was predictably horrible, I think Bonds the C had over 200 passed balls. The fact that the team was scoring about 18 runs a game helped them to a pretty good record as I recall. Step 1 of "find 8 Barry Bonds" is perhaps the challenging part of that strategy.

I take it this was late career Bonds? It would be interesting to compare the results of using 2002 or 2004 Bonds vs. using 1992-3 Bonds who wasn't as good of a hitter but still played defense well.

I'm surprised that's never been a thing, other such players who were good fielders but stuck at 1B by lefthandedness. I guess they just all play the outfield, or become pitchers early in life.

I doubt Hernandez is the only one. I would guess that most of the slick fielding lefty 1B (Grace, Joyner, Olerud) could have played a decent 3B had they been righthanded. Don Mattingly even played a few games at third in the majors.

Pujols in his twenties supposedly had 1B fielding stats on par with converted 2Bmen, implying Pujols could have played a passable 2B and been even more spectacular in overall value. (What was the story, that he preferred 1B because it took less of a physical toll?) Although that probably doesn't/won't hold true for his decline phase.

Pujols came up as a third baseman. He was pretty good at it. Third and second are close to equivalent positions. He played some at first in his rookie year, as it was also Mark McGwire's final year and Mark was hurt most of the season. The next year Albert played mostly OF and some 3B, as the team signed Tino Martinez for reasons I can't fathom. Tino couldn't play anywhere but first, Albert could, so that settled the first base issue. Albert was a decent outfileder. After the midpoint of the season they traded for Scott Rolen, and Albert wasn't going to play 3rd base anymore.

In 2004 he found a permanant home at first. He wasn't bad at other positions. It seems like the decision was made to prevent wear and tear more than anything else.

Brett Butler had 54 career HRs in 17 seasons as an OF. .290/.377/.376
OPS+ 110. 47 bWAR (-7 defense) Not sure what his SLG+ was but I'll bet it's lower than Ashburn's.

Harvey: Thanks. Ashburn's SB totals are indeed high in context. Still, it surprises me a guy could swing the bat well enough to hit .308 over 15 seasons and not hit more than 29 HR by accident. Did he swing down on the ball? Was Phil. a tough park for power hitters? It's also amazing that he could lead the league in BBs four times while offering no power threat at all -- seems like pitchers were more careful than they should have been.

Jackie Robinson played good defense all over the place. You could probably do a WAR study just with him.

Juan Pierre has played about as long as Ashburn did, and only has 17 career homers. Think of Richie as a Juan Pierre with better defense and 100 walks a year.

His ballpark had a CF fence that was 468 feet away some years, 447 others. Not good for hitting homers but very good for accumulating large putout totals, which Richie certainly did. He wasn't going to hit for power anyway though, he hit 12 at home and 17 on the road. 6 of those were inside the park.

There were a number of guys in the 50's who took walks without being real threats at the plate. Several of them named Eddie. This is very rare today. My guess is either the strike zone is tighter, or pitchers have better control today. Or a mix of both.

There were a number of guys in the 50's who took walks without being real threats at the plate. Several of them named Eddie. This is very rare today. My guess is either the strike zone is tighter, or pitchers have better control today. Or a mix of both.

Interesting. I suppose the closest parallels in last 30 years are Boggs and Henderson. Although both has a few seasons with some power, both also drew huge numbers of walks long before they ever proved they could hit the ball out of the park. Rose also drew a decent number of walks for a guy with no power (6.5 HR/650 PA). In the case of Rose and Boggs, the threat of a non-HR hit was high; in Rickey's case, it was mostly the crouch I think.

There was a large, very high scoreboard in RF in Connie Mack, too, but I don't think that would have cost Ashburn much. Wes Covington lost a few HRs to it while leaving some dents.

I would love to read some contemporary explanation of the Walking Eddies phenomenon. Not why they were frequently named Eddie, just how they existed, how their walking talents were appreciated, why pitchers didn't just challenge them, etc.

Interesting. I suppose the closest parallels in last 30 years are Boggs and Henderson. Although both has a few seasons with some power, both also drew huge numbers of walks long before they ever proved they could hit the ball out of the park. Rose also drew a decent number of walks for a guy with no power (6.5 HR/650 PA). In the case of Rose and Boggs, the threat of a non-HR hit was high; in Rickey's case, it was mostly the crouch I think.

Brett Gardner?

Late career Joe Morgan hung on at the end of his career with a Eddie-style batting line of 251/375/388 from 1979-1984. Other middle infielders did that in this period. Willie Randolph as mentioned. Tony Phillips with a career line of 266/374/389. Craig Counsell in recent years at 255/342/344. Mark McLemore, sort of the poor man's Tony Phillips, hit 259/349/341 in a better offensive environment.

The purest all-OBP / little else player of recent years was probably Dave Magadan. 288/390/377 from a corner player with a mediocre glove and no speed. He at least hit for a solid average, though. Mike Hargrove was Dave Magadan before Dave Magadan.

Small foul territory helps with the walks. Keep fouling off the close ones until they run out of patience. Large foul territory means some of those fouls get caught.

MCOA: One thing these guys have in common is they tend to be on the short side, giving them a smaller zone (and smaller still if they crouched). Morgan is 5-7, Phillips 5-10, McLemore and Randolph 5-11. But Magadan, at 6-3 but with no power, really is a mystery. Why did pitchers walk him?

Minor point in agreement - McLemore's listed height of 5-11 is crazy generous. I'd have guessed 5-8, tops.

I remember Bill James saying that listed player heights and weights were really unreliable. Has anyone done a project where they look at photos of ballplayers standing next to other players in order to cross-check listed heights?

Project, no, but plenty of observations. Hank Greenberg is listed at 6'3, Lou Gehrig 6'0. One of those is off by a few inches. I've seen a photo with Greenberg towering over Lou and other players.

[134] That reminds me that Ashburn was famous for fouling off pitches. There is a story of him hitting a woman with a foul, and later in the same AB, he hit her as they were carrying her out on a stretcher. From Wikipedia:

During an August 17, 1957, game, Ashburn hit a foul ball into the stands that struck spectator Alice Roth, wife of Philadelphia Bulletin sports editor, Earl Roth, breaking her nose. When play resumed Ashburn fouled off another ball that struck her while she was being carried off in a stretcher.[3] Ashburn and Ms. Roth maintained a friendship for many years, and the Roth's son later served as a Phillies batboy.

#38 James' study on aging showed that only catchers aged less well than second-basemen. But the difference between catcher and 2B (in terms of aging) is greater than the difference between 2B and any other position. IOW it's something that seems to be worth noting but not worth sweating over.

Now what James was looking at was more along the lines of expected career value for a young player rather than what will happen with a 31 year old. It seems plausible that this study has no impact on the topic at hand. I honestly don't know.

My guess is either the strike zone is tighter, or pitchers have better control today.

More of the second. Pitchers have better command both in the zone and outside of it, with the upshot being that if you have a small hitting zone and survive by simply taking pitches until the pitcher either walks you or throws you a pitch in the area you can handle, you're probably going to wait a very long time.

-- MWE

#86 Nettles himself started as an infielder. Boog Powell started in the outfield because Jim Gentile was even slower.

George Kelly was actually able to handle second base. Did a capable job there as the primary starter in 1925 and was an elite defender at first.

Honus Wagner was an elite defender at short and played other positions regularly (and well, but not well enough to cover the positional advantage of playing him at SS)

George Davis was another relatively late conversion to SS. He had more than 1,000 major league games before becoming a regular SS.

Bill James had a mini-study of attempts to convert outfielders to third-base. His conclusion: roughly an 80% failure rate.

Speaking of which: Dave Kingman spent significant time at third.

And one of my all-time faves: Bill James on the Bill Madlock not being happy with the Giants attempt to shift him to second:

"This is understandable; bears don't like to roller skate, cows don't like to dance and the Pope rarely appears on game shows. Playing second base was not among Madlock's considerable talents."

the Pope rarely appears on game shows

The Dating Game beckons!

#129 Dave Magadan is is an example of a guy who couldn't do all that much but draw walks -- and was 6'3".

EDIT: cokes. Missed his mention.

I don't think pitchers wanted to walk him. Walks are always available if you're willing to take pitches.

Also, something pretty clearly changed to end the great AL walk fest. Bill James suggested it was a change in the type of pitcher being selected. I think Don Malcolm pretty clearly demonstrated that this doesn't work as the explanation.

Far more likely that there was a change in the effective strike zone.

I don't think pitchers wanted to walk him. Walks are always available if you're willing to take pitches.

Not really. Magadan had a BB rate of 14.5%. That's far higher than you'd get just by taking a lot of pitches. I'm too lazy to check how many pitches he saw per PA, but he had a pretty low K rate (11%), so this can't be as simple as keeping the bat on his shoulder.

Pitchers as hitters only walk about 3% of the time. And plenty of them have OBP below .145. If simply not swinging could get you walked 14% of the time, a lot of pitchers would simply never swing.

#145 I think Daric Barton shows what happens if all you do is look for a walk -- and you can't punish the resulting BP fastball. So yeah, I'm over-simplifying. There's more to it than just standing there at the plate.

Pitchers rarely see anything that the guy on the mound can't reliably throw for a strike.

Magadan had terrific bat control - consistently good batting average - so I'm pretty sure he could foul off pitches not in his zone and extend at-bats. He was a good guy to have as a pinch-hitter because there was a good chance he'd keep the inning going.

This conversation reminded me of a Fangraphs article on Ian Kennedy.

http://www.fangraphs.com/blogs/index.php/ian-kennedy-the-hitter-who-doesnt/

Makes you wonder if keeping the bat on your shoulder is an optimal strategy for a bad hitting pitcher.