Weighted WAR: Positional Classification and Left Field Rankings

So far in the discussion of weighted WAR, we’ve explained the weighting system itself, as well as the first adjustment to the base WAR totals, schedule length. We’ve also examined the system’s ranking of the best third basemen ever. But as we continue exploring both the system itself and the rankings it produces, there’s a question that needs to be answered: What does it mean to rank, say, the best left fielders of all time? More specifically, are we ranking players based on their value in left field alone (which I’ll call the “positional purist” approach), or are we evaluating their entire careers and using some criteria to classify that career into one position (the “whole career” approach)?

In this case, we’re doing the latter. The whole career approach is commonly used in positional rankings, and has always been my preference. It is, however, worth exploring the merits of both options. And since we’re in left field, we can do that through the prism of everyone’s favorite multi-positional star: Pete Rose.

(Spoiler alert: Rose is classified in left field, and does in fact make the top 100.)

Pete Rose debuted in 1963, and spent the first four years of his career primarily at second base. In 1967, he shifted predominantly to the outfield, starting in left field for a season, then moving to right for four years (with a bit of center thrown in, but never as a primary position), then back to left from ’72-’74. In 1975, Rose opened the season in left, but shifted to third base partway through to allow the Reds to get George Foster into the lineup. He remained at third through 1978, at which point he made two moves: to the Phillies, and to first base. He would make occasional forays into the outfield, but stayed primarily at first for the rest of his career.

In order to determine positional classification, I split up each player’s seasonal adjusted WAR totals by the position at which the player spent the most time during each individual year. For Rose, those totals are as follows:

LF	25.2
RF	20.8
3B	17.1
2B	12.9
1B	7.1

Rose is an edge case for any method of positional ranking. By either raw bWAR or weighted WAR, he is a top-50 position player of all time. And yet, if you consider only his time at any one position when preparing a positional ranking, he would likely fail to make the top 50 at any of them.

A positional purist might reasonably take either of two stances with regard to this result. First, create a separate list for players who spent significant time at multiple positions. There are two directions this can go: either a large general list for multipositional players, or multiple smaller lists for players with more particular positional breakdowns. Either of these options runs into issues. The problem goes back to the basic purpose of positional rankings, which in my opinion is as follows: separate the very large set of “all baseball players ever” into more manageable but still substantial (and preferably similarly sized) subsets, and do so in a way that allows players to be compared to others with whom they share at least one major commonality. If you lump all multipositional players together, you end up with a list that includes both Gil McDougald and Brian Giles, who have no time spent at overlapping positions; does versatile infielder McDougald really have more in common with wandering outfielder Giles than he does with a full-time third baseman like Kyle Seager? If you avoid this overlap by making the lists more specific, the size of the subsets can easily become a problem. Yes, you can be fairly confident that Pete Rose is the best LF/RF/3B/2B/1B in baseball history, but how many other players are on that list to begin with?

The second positional purist stance is much simpler: “so what?” Rose played where he played; he would likely make the top 100 list at 2-4 positions once everyone goes through the same value separation process, and he didn’t really spend enough time at any position to expect better. And anyway, it’s Pete Rose; nobody really likes that guy.

The standard comparison in support of this stance is to team-based rankings. A list of the greatest Cincinnati Reds of all time will prominently feature Rose, but will do so based only on his time in Cincinnati; his tenures with the Phillies and Expos will not generally be factored in. Even as a whole career advocate in positional rankings, I agree that the purist approach to a team-specific ranking is appropriate. If one is ranking the greatest Reds, one should focus on the players who made the greatest contribution to the team in question – helping them win games and compete for championships and otherwise fulfill the goals that baseball teams have.

That is where the difference between team and positional rankings comes in. Positions are not teams. They are not built as cohesive wholes and they have no aims. When Pete Rose left the Reds for the Phillies after the 1978 season, he went from valued ally to direct competitor from the perspective of the Cincinnati club. His simultaneous move across the diamond from the hot corner to first base posed no such issues for the other third basemen in the league. Rose’s 1973 MVP season helped his Reds teammates a great deal, as his 8 WAR pushed them to a 3.5-game triumph in the NL West. It did not benefit his fellow left fielders whatsoever.

Ultimately, though, it is a matter of taste. If someone is fascinated by the history of a particular position for its own sake and wants to create a ranking of how much various players contributed to that position, they are free to do so. My interest, rather, is in the players. As such, the rankings presented here will be of the players’ entire careers (or careers to date, in the case of active players), and the positions will be used as a way of sorting them into subsets that have a common characteristic. If you think that means this list should be called “best players who accrued a plurality of their value in seasons in which a plurality of their playing time came in left field,” fair enough; I just don’t have the level of dedication required to type PWAAPOTVISIWAPOTPTCILF all the time, and therefore hope you’ll allow “left fielders” as a convenient shorthand.

I’ll quickly discuss the specifics of the method used to assign each season to a single position. My giant spreadsheet of WAR tables and various other things also includes each player’s Retrosheet table of defensive statistics. For seasons from 1912 forward, Retrosheet has defensive innings available, which would seem like an obvious choice when assessing playing time… except that Retrosheet does not provide innings spent at DH (which is reasonable, because that would be nonsense). To account for DH (and allow myself to be a bit more precise in seasons before inning totals are available), I used the following innings estimator, developed via regression a decade or so ago: 

Innings played = 9*(complete games) + 7*(games started – complete games) + 2*(games – games started)

The position which has the highest value on this estimator (or, pre-1901, the highest games played total, since fielding GS and CG aren’t available that far back) is designated as the player's primary position for that season. As seen in the Pete Rose example above, adjusted (but not weighted) WAR totals are calculated for each position, and the one with the highest total is the player’s primary position overall. (Yes, there is some imprecision here, which may lead to some less-than-perfect classification. I can live with this; if a player’s designation is close enough between LF and RF that half a season of WAR would change it, I think you can reasonably place him in either outfield corner at that point.)

All that said, let’s bring this around to left fielders who aren’t permanently banned from the game. Here are the positional splits of adjusted (not weighted) WAR for the top 15 left fielders:

Barry Bonds: LF 161.4, CF 3.5

Ted Williams: LF 113.2, RF 6.3

Rickey Henderson: LF 94.5, CF 21.0

Carl Yastrzemski: LF 61.8, 1B 24.1, DH 5.2, CF 3.8

Pete Rose: LF 25.2, RF 20.8, 3B 17.1, 2B 12.9, 1B 7.1

Ed Delahanty: LF 59.1, CF 3.7, 1B 2.6

Tim Raines: LF 64.1, CF 6.5, DH 0.8

Manny Ramirez: LF 34.5, RF 31.1, DH 5.2

Al Simmons: LF 42.0, CF 19.8

Billy Williams: LF 45.6, RF 12.1, 1B 2.0, DH 1.9

Minnie Minoso: LF 42.0, 3B 12.6, CF 3.8

Goose Goslin: LF 54.5, RF 2.8

Willie Stargell: LF 40.6, 1B 16.1, RF 0.1

Jose Cruz: LF 39.9, RF 13.9, CF 2.4

Brian Giles: LF 19.7, RF 18.4, CF 12.9, DH 1.4

Taking them as a group, here are the secondary positions of these 15 left fielders, sorted by WAR earned in seasons classified at the position:

RF	105.5
CF	77.4
1B	51.9
3B	29.7
DH	14.5
2B	12.9

That looks like about what you’d expect – other outfield positions, then first base. 3B over DH is maybe a bit surprising, but all of that third base value comes from two players (and, of course, DH got a late start and is playing catchup).

You can get a sense for how these classifications play out in this sample. A few players (Rose, Ramirez, and Giles) are close enough that you could reasonably place them elsewhere (right field in all three cases), but a solid majority of the players listed have an obvious primary position.

Also, a bonus table – here are the active players in the database who have been in the majors for at least 5 years and whose overall primary position changed in 2024:

Player	Years	Old Pos	New Pos
Marcus Semien	2013-24	SS	2B
Juan Soto	2018-24	LF	RF
Marcell Ozuna	2013-24	LF	DH
Michael Conforto	2015-24	RF	LF
Bryan Reynolds	2019-24	CF	LF
Yandy Diaz	2017-24	3B	1B
Daulton Varsho	2020-24	RF	CF
Randal Grichuk	2014-23	LF	RF

Not too surprisingly, the list is largely composed of outfielders, along with a couple of infielders who moved to easier positions. There’s also a minor exodus of notable left fielders, which bodes ill for the quality of our next table: active left fielders (as of 2024) who are either in the top 100 or might reasonably hope to join the top 100 soon:

Player	Rank	Years	WAR	aWAR	wWAR	2024 WAR	Rank Change
Christian Yelich	34	2013-24	42.0	42.6	34.7	2.2	+2
Starling Marte	38	2012-24	38.6	41.3	33.0	0.7	+2
Bryan Reynolds	112	2019-34	19.6	20.2	18.4
Michael Conforto	117	2015-24	17.6	20.0	17.6
Ian Happ	120	2017-24	17.8	19.4	17.1
Tommy Pham	122	2014-24	17.6	18.4	16.8
Mark Canha	129	2015-24	15.4	18.6	16.3

Third base, if you’ll recall, has twelve active players in the top 100. As of 2023, left field had five – but the aforementioned Soto and Ozuna changed spots, and Michael Brantley retired, leaving two active top-100 left fielders, neither of whom made much progress this year. Here’s hoping for some of the guys on the edges and/or some younger left fielders to make progress in the coming years. (Soto, of course, could end up back in left as well depending on where the Mets play him moving forward.)

And finally, here are the top 25 left fielders by weighted WAR, plus the players ranked at the subsequent multiples of 10:

Player	Rank	Years	WAR	aWAR	wWAR
Barry Bonds	1	1986-2007	162.8	164.9	97.6
Ted Williams	2	1939-60	122.0	119.5	82.0
Rickey Henderson	3	1979-2003	111.1	115.5	74.9
Carl Yastrzemski	4	1961-83	96.5	94.9	64.9
Pete Rose	5	1963-86	79.8	83.0	55.3
Ed Delahanty	6	1888-1903	69.8	65.4	51.1
Tim Raines	7	1979-2002	69.5	71.4	50.2
Manny Ramirez	8	1993-2011	69.2	70.8	49.3
Al Simmons	9	1924-44	68.4	61.8	47.3
Billy Williams	10	1959-76	63.8	61.5	46.0
Minnie Minoso	11	1946-80	54.0	58.4	44.5
Goose Goslin	12	1921-38	66.2	57.4	43.1
Willie Stargell	13	1962-82	57.5	56.8	42.1
Jose Cruz	14	1970-88	54.6	56.1	41.2
Brian Giles	15	1995-2009	50.7	52.5	40.6
Jim Rice	16	1974-89	47.8	49.2	39.1
Fred Clarke	17	1894-1915	68.2	56.6	38.8
Ralph Kiner	18	1946-55	48.0	45.1	38.6
Ryan Braun	19	2007-20	47.2	47.3	38.6
Luis Gonzalez	20	1990-2008	51.6	52.3	38.6
Jesse Burkett	21	1890-1905	62.9	51.6	38.4
Roy White	22	1965-79	46.8	47.5	38.1
Joe Medwick	23	1932-48	54.4	47.6	38.0
Willie Wilson	24	1976-94	46.2	48.1	38.0
George Foster	25	1969-86	44.2	46.3	37.7
Brian Downing	30	1973-92	51.6	52.5	35.7
Lonnie Smith	40	1978-94	38.7	40.2	32.7
Dusty Baker	50	1968-86	37.1	38.9	29.9
George Burns	60	1911-25	39.9	30.0	25.6
Cliff Floyd	70	1993-2009	25.9	27.6	23.9
Larry Hisle	80	1968-82	25.1	25.8	22.5
Mike Donlin	90	1899-1914	29.1	23.4	20.7
Topsy Hartsel	100	1898-1911	31.6	23.4	19.8

George Foster making a return appearance in reference to the original Pete Rose discussion is a nice bit of symmetry to close us out.

Left field skews a bit older than third base at the top of the list (despite the absence of Stan Musial, who is classified elsewhere), with nine of the top 25 players debuting pre-1950. But there’s still plenty of modern representation, even if none of it is currently active.

And that brings us to the topic of our next entry. We’ll be returning to both the infield and the discussion of WAR adjustments as we explore the ranking of the best shortstops through the lens of timelining.