After reading through those stats, I’m ready to listen to the argument that Roberto Clemente over 17 seasons was better than either Mays or Musial.
I’m not convinced he was better than Mays. As I alluded to, the list there seems to overstate the importance of being better than replacement level players at your position, but doesn’t account for the difference in value between positions. So that list assumes that a right fielder is inherently as valuable as a center fielder (sort of) which is not necessarily true.
But it’s possible. And Roberto Clemente better than Musial? Well, sure, what’s surprising about that? Clemente was an absolutely brilliant right fielder, often cited as the greatest throwing outfielder who ever lived. He won 11 Gold Gloves for good reason; he was an amazing outfielder. All contemporary observers thought Clemente was a brilliant fielder.
The fact Clemente played 17 season to Musial’s 24 is not really relevant or evidence there’s a problem with the methodology becuase the seven missing seasons are the ones at the end. Players tend to amass all their defensive value at the beginning of their careers; according to WAR, Musial compiled as much defensive value in the first four seasons of his career (against war-depleted opposition) than he did the rest of his career. That’s actually pretty typical; players often peak as fielders early.
Here is another way to look at defensive value. I broke out my copy of Win Shares by Bill James and looked at fielding win shares by position (pages 602 to 622 in my copy). I looked for the highest win shares per 1000 innings played for each position. I’ll warn you that the lists are ordered by total innings played, so I could easily have missed something.
I have some caveats about interpreting the data. First, the book was published in 2002, so the data aren’t up to date on players who have played since then (e.g. Andruw Jones’ fielding took a dive since this book was published). Second, the numbers represent a player’s career, not his peak value (which may make Willie Mays look worse than he really was because he continued to play past his prime, and at the same time it may make Curt Flood look better than he really was because his career had no decline phase). Third, as was noted in previous postings, at some point in history third base became less important defensively - most of the top-fielding third basemen in this list were from the dead ball era. I put a “*” in front of each third baseman from the dead ball era. For the sake of understanding I skipped over some of the top-fielding third basemen in the list in order to get to the ones we know from the live ball era (I put in a “…” as a dividing line). Finally, Bill James lumped all three outfield positions together, and I didn’t try to separate them.
Here they are:
Catcher
Ivan Rodriguez 8.29
Ossee Schreckengost 7.93
Paul Richards 7.76
Pop Snyder 7.63
Buck Ewing 7.33
Mike Powers 7.31
Gabby Hartnett 7.28
Bill Killefer 7.24
Charles Johnson 7.23
Ron Karkovice 7.13
Hal Smith 7.01
First Base
Tino Martinez 2.60
Pete O’Brien 2.58
Todd Zeile 2.52
Milt Scott 2.51
Todd Helton 2.39
Lamar Johnson 2.38
Wes Parker 2.36
Patsy Tebeau 2.36
John Ganzel 2.29
Bill Terry 2.28
Vic Power 2.28
John Olerud 2.26
Mike Squires 2.26
Dave Foutz 2.25
Frank Isbell 2.24
Roger Connor 2.20
Frank McCormick 2.20
Second Base
Bill Mazeroski 6.13
Hal Lanier 6.11
Mark Lemke 6.04
Jody Reed 6.01
Jackie Robinson 5.94
Glenn Hubbard 5.92
Eddie Mayo 5.89
Burgess Whitehead 5.73
Bobby Grich 5.68
Morrie Rath 5.63
Red Schoendienst 5.61
Frank White 5.58
Jerry Coleman 5.55
Nellie Fox 5.54
Jerry Adair 5.53
Third Base
- Bobby Wallace 6.73
- Billy Clingman 6.59
- Lee Tannehill 6.58
- Lave Cross 6.03
- Jimmy Collins 5.93
- Buck Herzog 5.86
- Tommy Leach 5.76
- Charlie Reilly 5.69
- Art Devlin 5.67
Mark Christman 5.64 - Billy Nash 5.51
- Ned Williamson 5.49
- Ossie Vitt 5.45
Babe Pinelli 5.37
Bernie Friberg 5.32
Willie Kamm 5.30 - Billy Shindle 5.30
- Tom Burns 5.30
Heinie Groh 5.29 (about half his career was in the dead ball era) - Terry Turner 5.29
- Arlie Latham 5.28
- Art Whitney 5.27
- Charlie Deal 5.26
…
Clete Boyer 4.97
Pie Traynor 4.65
Mike Schmidt 4.51
Graig Nettles 4.40
Gary Gaetti 4.40
Darrell Evans 4.37
Robin Ventura 4.31
Brooks Robinson 4.24
Shortstop
Bob Allen 7.73
Hughie Jennings 7.68
Marty Marion 7.32
Joe Tinker 7.28
Dal Maxvill 7.24
Rick Burleson 7.17
Phil Rizzuto 7.14
Art Fletcher 7.04
Everett Scott 6.98
Honus Wagner 6.89
Mickey Doolan 6.86
Germany Smith 6.86
Bill Dahlen 6.82
Monte Ward 6.74
Mark Belanger 6.72
Lee Tannehill 6.71
Neifi Perez 6.69
Charlie Gelbert 6.61
Orlando Cabrera 6.59
Lou Boudreau 6.49
Ozzie Smith 6.42
Rabbit Maranville 6.42
Herman Long 6.40
Billy Rogell 6.39
Dave Concepcion 6.37
Rey Ordonez 6.32
Tommy Corcoran 6.27
Dave Bancroft 6.20
Outfield
Andruw Jones 6.47
Curt Flood 5.32
Jimmy McAleer 5.20
Curt Welch 5.11
Mike Kreevich 5.06
Tris Speaker 4.93
Jim Busby 4.71
Marquis Grisson 4.70
Dom DiMaggio 4.66
Tommy Leach 4.64
Pop Corkhill 4.56
Paul Blair 4.55
Jimmy Piersall 4.50
Terry Moore 4.41
Hugh Nicol 4.38
Max Carey 4.37
Tommy Agee 4.37
George Gore 4.35
Hi Myers 4.35
Lloyd Waner 4.33
Fielder Jones 4.29
Otis Nixon 4.29
Amos Otis 4.25
Taylor Douthit 4.25
Sammy West 4.20
Garry Maddox 4.18
Willie Wilson 4.13
Steve Brodie 4.12
Willie Mays 4.11
Darren Lewis 4.07
Omar Moreno 4.05
Jim Edmonds 4.02
Joe DiMaggio 4.01
Bill Virdon 4.00
Isn’t win shares pretty much an abandoned methodology at this point?
Giving credit to the team based on won/loss record is wide open to overcrediting or undercrediting players. If a team gets outscored on the season, yet wins 90 games, its players are going to be given a lot more credit, in terms of win shares, than they actually deserve.
Has a team ever won 90 games while being outscored for a season? While the relationship between winning percentage and run differential isn’t perfect, this would be an extreme outlier.
Generally, a team’s win-loss record can be predicted with pretty good accuracy from the number of runs scored and allowed. There are many things in sabermetrics that depend on this relationship - not just win shares. WAR depends on this, for instance.
The 1984 New York Mets quite famously did this. Quite an unusual outlier.
This is an interesting philosophical question though, and is why you can’t just go by any one number. The logic behind Win Shares is that the aggregate amount of value given to players must logically be equal to the team’s ability to win games. WAR dismisses that entirely, and if a team has more or less WAR than their actual number of wins then that’s just ignored. I suspect, personally, that the problem is fielding WAR, but that’s based on a few brief observations, not any sort of detailed study.
Personally, I think the logic of Win Shares is unassailable. If the 1984 Mets went 90-72, a methodology that pretends they went 78-84 (their projected record) is just plainly wrong. They did not win 78 games; they won 90. Pretending they won 78 games is absurd.
On the other, for the purpose of evaluating a player’s ability and how he is likely to do in the future, knowing the 1984 Mets might have been really lucky is of great value, and when examining them in retrospect it’s worth it to know that, hey, maybe buddy with 15 Win Shares but a shockingly low WAR wasn’t really the guy who was helping the Mets win those extra games.
I like looking at all the evidence; I don’t buy any one number.
I’ll not take anything away from CF, but 1B seems to get a bad wrap, at least on the thread. 1B is the 3rd busiest position on the field. While the OF players can stand around a scratch themselves between pitches the 1B 'man is often in constant motion holding a runner on…well, I shouldn’t have to explain. 1B is the “hot corner” on the right side of the diamond that requires the same reaction time as 3rd base (and more than any OF position). I know sabermetrics considers 1B like the ugly step-child…their bias.
Jack Buck (R.I.P.) was around as Cardinal broadcaster to see Mays’ career from 1954
on and got to witness Mays’ 4 WS seasons, his 4 HR titles, 1 BA title, 4 SB titles and the 29 times Mays lead the **National League **in MLB’s 13 hitting categories.
Buck did not broadcast Cardinal games for 13 of Musial’s 21 full seasons, missing the first 11 seasons of Musial’s career. Buck did not get to see Musial:
- Play his 4 WS seasons
- Win 6 of his 7 batting titles
- Collect 82% of his career 3B
- Record 71% of his career 2B
- Register 78% of his career HITS
- lead the Major League 46 times in MLB’s 13 hitting categories
- Record the highest TOTAL BASES in a season (429) in over 80 years (thru 2010)
- Become the only player in history to record at least 20 - 3B and 50 - 2B in the same season
- Registar most of the 16 consecutive seasons he had BA .300 or higher - Major League record - 3.1 PA
- Registar most of the 15 consecutive seasons he SLG .500 or higher - Major League Record - 3.1 PA
The problem with Win Shares is that you could have two players, who played for different teams, and put up exactly the same stat lines, including defense, clutch hitting, on-base, etc.
However, because one of the player’s teams got “lucky” in terms of out-performing their pythagorean, or the other player’s team underachieved, there would be a disparity in their Win Shares, despite having played exactly as well as one another.
This is the silliest thing I’ve read all week.
I don’t know. It’s right up there with asserting that Jack Buck never saw Musial play before the first time he broadcast a Cardinals game.
Charlie, there’s a reason Frank Thomas played first base and not center field. All the positions have some importance, but come on. We all know first is not generally where the top defensive talents go.
Just out of curiosity, how does MLB have only “13 hitting categories”?
[QUOTE=markdash]
The problem with Win Shares is that you could have two players, who played for different teams, and put up exactly the same stat lines, including defense, clutch hitting, on-base, etc.
However, because one of the player’s teams got “lucky” in terms of out-performing their pythagorean, or the other player’s team underachieved, there would be a disparity in their Win Shares, despite having played exactly as well as one another.
[/QUOTE]
That’s entirely possible. But it’s not a bug, it’s a feature.
Again, the point of Win Shares is to allocate credit for games that are actually won, not games that could have been won if a team had been luckier. It’s a different measurement from WAR. Neither is conceptually wrong; they measure different concepts.
And there may be a reason Mays was just a CF. They tried him at 1B late in his career and he performed poorly…the greatest outfielder of all time couldn’t handle the “easiest” position on the field. How could this be? Many an OF, both LH and RH throwing, moved to 1B late in their career and managed quite well: Stargell, Mantle, Aaron, Yaz, to name a few. He might have extended his playing time and possibly chased down Ruth’s record could he have adapted. Wouldn’t it be something if Mays were a one-trick pony on defense and, coupled with a lack of versatility, it cost him a chance to break Ruth’s record? Maybe it was Willie’s “replacement player” that was trying to play at 1B. Just wondering…
HIT, RUNS, 2B 3B, HR, RBI, TB, BB, BA OBP, SLG, XBH AND OPS. There are actually 4 more categories: GAMES, PA, AB AND SO but the don’t contribute much in determing a player’s “batting value”. Check out at MLB.com.
[QUOTE]
MLB.com lists many more stats than that. And you just topped your previous silliest comment of the week - seriously, plate appearances and at bats don’t do much in determining a player’s value?!? Please tell me you’ll post more of these gems in the near future.
Are you calling Mark silly?
“It made me not lazy with my hands. First base is the hot corner on the other side.” - Mark Teixeira
(Google Junior Baseball,5 Tips to Master First Base.)
Okay, go ahead and include them in the calculation, it isn’t going to make any difference to the outcome of the value reached.
Read what I said in my post. I agree, he surely must have seen Musial play before 1954, somewhere. After WWII he enrolled at his home state university, Ohio State (where he broadcast OSU basketball games). Upon graduating in 1949 or 1950 he spent three years broadcasting minor league BB games in Ohio and New York before being hired by the Cardinals to team with Harry Carey. [Harry (the homer) Carey saw Musial play more games than any other broadcaster.]
His trick was being the greatest ever at one of the three most difficult positions to play on the field. Musial was not the greatest at any position. OK? That’s the first thing, and really the only one that matters. But you’ve already ignored this point, so let’s move on.
Second: he handled first base just fine. You’ve pointed on multiple occasions to Musial’s fielding percentage, so how exactly do you deal with the fact that Willie Mays had a better fielding percentage at first base than Stan Musial did, and was better than Musial in every way that can be quantified, when the whole career is considered? “Couldn’t handle it”? Sure, if you’re flatly dishonest about it. This crosses the line from weird opinion to outright misrepresentation, I think, so I’m not even pointing it out for your response, to be honest. Just saving everybody else the trouble of investigating it.
I’d call Mark “self-interested.” You don’t want to cite a 1B-man on how hard it is to do what he does for a living, any more than you want to ask if he feels properly appreciated.
Well, you do. But most people don’t.
What most intelligent people would do is maybe figure out which position is the one most frequently played by players who DH a lot. DHs generally can’t field at all, and would play a position if they could, but sometimes are compelled to play one in a pinch. That might give you some idea if 1B is difficult or relatively easy. Go ahead. Try it. I’ll wait.
Makes my head hurt, sorry.
(Turns out I was entirely wrong in the post above about fielding percentages; read the wrong data. My apologies)