Whose best position is DH?

Want to play in the THT 2011 Experts League? Click here for details.

Editor Note: the numbers in the post have been updated.

Last year, I pontificated as to whether Adam Dunn was as “underrated” as sabermetrics types make him out to be. In what I labeled “Valuable Post Script,” I noted the following:

As a right fielder, Dunn’s cumulative batting and fielding production gets a -7.5 positional adjustment (UZR measures all defense equally; Fangraphs accounts for differences in fielding difficulty between positions in WAR calculations through positional adjustments). As a DH, Dunn would get a flat -17.5 positional adjustment and a zero fielding rating. In other words, as a DH, Dunn gets just -17.5 runs subtracted from his batting line. As a right fielder (or left fielder, for that matter), Dunn gets -7.5 subtracted from his batting line in addition to his lackluster fielding. Thus Dunn, like anyone with a consistent -10 or worse fielding glove at RF/LF, belongs in a DH role.

As noted then and expand upon now, at some point, a player’s defense is so poor, so sub-par, so Adam-Dunn-like, that a player just simply should have his glove Mark Kotsayed.

Fangraphs, which uses the UZR system to account for a batter’s WAR (except catchers, for whom the plus/minus system is used), gives the following pro-rated defensive adjustments per 150 games by position to account for the ease/difficulty of the position:
Catcher: +11.6 runs (again, plus/minus system is used for the fielding)
First base: -11.6 runs
Second base: +2.3 runs
Third base: +2.3 runs
Shortstop: +6.9 runs
Left field: -6.9 runs
Center field: +2.3 runs
Right field: -6.9 runs
Designated hitter: -16.2 runs

Clearly, all “average fielders” (UZR/150=0) belong in the field. At the worst, an “average” fielder playing first will cost you 12.5 runs per season, whereas that player, at DH, would burden you with 17.5 runs. Even most below-average fielders belong in the field, at least depending on what position they play. A poor first basemen who costs you less than half a win per 150 games played is clearly worth having (or at least not worth not having) at the slow corner.

While clearly someone needs to play each of the nine field positions for eight or nine half-innings per game, there are just flat out some players who should be plopped at DH. What this threshold is varies by position, as follows:

Catcher: -27.8 fielding runs
First base: -4.6 fielding runs
Second base: -18.5 fielding runs
Third base: -18.5 fielding runs
Shortstop: -23.1 fielding runs
Left field: -9.3 fielding runs
Center field: -18.5 fielding runs
Right field: -9.3 fielding runs

Using these thresholds, here are three-year data (to help smooth out small sample fielding volatility) of all fielders who have played a minimum of 1,500 innings at a given position, put to the “should you be a DH” test via their three-year UZR/150 ratings. (Because Fangraphs’ WAR system doesn’t use UZR data to determine the value of catchers, they are omitted here.) The results are presented below. Click to enlarge. You can download my DH data sheet by clicking here.

Hitters are categorized into three categories: should be DHed (UZR/150 is below the threshold); maybe should be DHed (UZR/150 is less than five runs above the threshold), should not be DHed (UZR/150 is above the threshold). Within the later category is a sub-group of players who contribute +3.0 WAR to their team by playing the field. The residuals (guys who should not be DHed, but provide less than +3.0 WAR defensively) have been omitted from my presentation below. You can find them in the data sheet (or by the process of elimination).

First, let’s look at the players whose gloves should absolutely be burned in a bonfire:

Here, we find players with the worst defensive reputations in baseball. Dunn is not listed as an outfielder (he has played fewer than 1,500 innings in the outfield over the past three seasons, but his career OF_UZR/150 of -13.3 would have landed him a spot), but he’s nonetheless representing first basemen everywhere. All but three of the players on this list are corner outfielders, the traditional “put your slugger here when first base is blocked” vacuum. The other three are first basemen. This should not come as a surprise. Given the comparative “ease of fielding” in the corners (a 10-run swing in difficulty compared to center field), I wonder if there’s really an excuse for being that bad. We also see the reason why teams were reluctant to sign Jermaine Dye, who allegedly refused to play first base/DH for a team.

Next, the borderline DH types:

Here again, we almost exclusively find first basemen and corner outfielders. Almost, because Jorge Cantu and Vernon Wells tried really hard to make this list. Note the following: Ryan Braun, who was untenable at third, is almost unplayable even in left. Ditto with Miguel Cabrera, substituting first base for left field. Chris Davis, Jason Bay, Corey Hart and Aubrey Huff also are borderline unplayable in the field, assuming that you have another option who can play at replacement level. We also find old/aging/injury-prone players like Magglio Ordonez and Carlos Pena. These hitters should probably be DH-ing just to keep their health/longevity in check.

And finally, those guys who you play at DH only if Ozzie Guillen is your manager

A Hardball Times Update
Goodbye for now.

Here we find many of the true “Gold Glove” fielders of baseball: Jack Wilson, Franklin Gutierrez, Evan Longoria, Adrian Beltre, Ryan Zimmerman, Adrian Gonzalez, and Carl Crawford. The list is sprinkled with premium position players (positional adjustments greater than zero), though one sub-premium position player, free agent Crawford, made the list as well.

Omar Vizquel, whose career wOBA is .313 and whose wRC+ is below 90, is not listed here, due in part to his two-year stint as a utility fielder. Nonetheless it is worth noting that his three-year UZR/150 at shortstop, the hardest non-catcher position to play defensively, was +17.4. At least Ozzie started him at DH in only one game.

Jeffrey Gross is an attorney who periodically moonlights as a (fantasy) baseball analyst. He also responsibly enjoys tasty adult beverages. You can read about those adventures at his blog and/or follow him on Twitter @saBEERmetrics.
Newest Most Voted
Inline Feedbacks
View all comments
13 years ago

Hmm, Chase Headley the outfielder should never be in the field, while Chase Headley the third baseman should never be a DH.  Maybe the Nats should move Dunn to third base.

dave silverwood
13 years ago

Adam Dunn is the best dh in baseball who has not done it yet,as a Red fan his outfield play was little league and even tho I hate the dh I realize that I love to see Adam Dunn hit, next is Vlad Guerrero who as an Expo could field quite well,lost a little with leg problems but still wow what a hitter.

Detroit Michael
13 years ago

The analysis assumes that players’ offensive production will be unchanged if they were to become full-time DHs.  I believe the evidence for that is false.  Playing DH is a bit like pinch-hitting, that a guy coming off the bench tends not to do as well.  Separating that effect from the fact that guys who are DHs might be playing through injuries is difficult to do.

Jeffrey Gross
13 years ago

Wait…arg…the Positional Adjustments are per 162…I will update later with the pro-rated /150 adjustments:

Catcher         11.6
Shortstop       6.9
2B, 3B, CF       2.3
Corner Outfield   -6.9
First base     -11.6
DH           -16.2

Which alters the thresholds:
POSITION       UZR/150 DH Threshold
Catcher         -27.8
Shortstop       -23.1
2B, 3B, CF       -18.5
Corner Outfield   -9.3
First base       -4.6

I will update the spreadsheets and images accordingly….

Jeffrey Gross
13 years ago


Jeffrey Gross
13 years ago


I am simply looking at the DH-types getting playing time in the field. While clearly the DH (and 1B) spot is limited in quantity, you see that guys like Quentin are on teams where guys like Mark Kotsay are DHing.

What one can take away from this is a question of whether the DH spot is being properly utilized. Why re-sign Matsui in LA, for example, when you have Bobby Abreu ready to slide in? Or in KC, why go after a DH body when you have Butler for the role and plenty of young 1B/DH types in the system…

I realize the list has more than 14 names for only 14 positions (many of which are currently occupied). Just trying to provide some insight…

Jeffrey Gross
13 years ago

I would personally argue the following DH’s should be moved off the DH spot:

-Russell Branyan: can man 1B effectively (though not 3B anymore in his injury-prone aging years)
-Johnny Damon: has defensive value. The tigers use of him in the DH role was a moronic move, plain and simple.
-Luke Scott: can hold his own in the corners and has some marginal add-value at first
-Milton Bradley: if he’s going to play, put him in the corners or first.

Then you have the “open” DH spots:
-White Sox
-Angel (unless they resign Matsui)
-Rangers (unless they resign vlad)
-Oakland (assuming they re-re-non-tender Cust)

Brad Johnson
13 years ago


You’re missing the point entirely, what this shows us is who loses value by playing in the field rather than just batting. That’s not to say that they shouldn’t play the field under any circumstance, Prince Fielder obviously has value over Lyle Overbay as a 1b thanks to his bat. However, by WAR standards, Fielder is hypothetically more valuable to a team as a DH than as a 1b.


Let’s model it real quick. First, I’ll keep this in WAR terms while assuming a +20 run replacement level (equal playing time).

Now let’s say free agent Adam Dunn is a 40 RAR batter, -15 UZR, -12.5 Pos adj, +20 replacement, good for +32.5 runs as a 1b. He’s worth +42.5 as a DH.

And let’s say an AL team is going to sign Dunn. They want to use Madeup Player and Dunn as their 1b/DH combo. Madeup Player (as a 1b) is a 0 RAR, +10 UZR, -12.5 pos, +20 rep -> +17.5 runs. As a DH he’s +2.5 runs.

In this example, if the team uses Adam Dunn as the DH, they net +60 runs. If they use him as the 1b, they net +35 runs. Turn that into WAR and we’re talking about a 2.5 WAR difference just by playing guys at their ideal positions. Of course, this example is rather extreme, especially the 0 RAR/+10 UZR 1b part, but I hope it illustrates the point of Jeff’s exercise. By using these guys in the field (AL) teams are potentially leaving value on the table. In Dunn’s case, all you need is a league average, glovely 1b to make him much more valuable as a DH.

Last disclaimer, the example also assumes that no other non-replacement level options exist. In reality, a team will have other 1b/DH options to consider that makes the calculus slightly more involved.

13 years ago

Detroit Michael said, “The analysis assumes that players’ offensive production will be unchanged if they were to become full-time DHs.  I believe the evidence for that is false.”

FanGraphs’s and Rally’s WAR already accounts for the difficulty in DHing. In short the DH is actually a -20 runs defensive position but they get a 2.5 to 5 run “difficulty of hitting as a DH boost”; so their overall positional adjustment is -15 to -17.5.

13 years ago

This is an interesting example of playing with the numbers, but what exactly is the point?

For example, there are three Brewers between your really bad/pretty bad lists.  According to Fangraphs WAR, they were three of the five most valuable Brewers in 2010, despite their defensive shortcomings (yes, the Brewers sucked, but their hitters weren’t the problem).  I agree that Prince Fielder is terrible at first, and if there’s any justice in the world he’ll be traded to an AL team and DH next year.  But over the three years covered in your analysis, what were the Brewers supposed to have done?  Ditch Fielder for a replacement-level defensive 1B who can give them four batting wins a year?  1) Those guys don’t grow on trees, 2) You’re missing the point of WAR, which takes into account both offense and defense and still says Fielder was worth an everyday spot in the field, 3) Apparently Hart and Braun should also DH, despite contributing 3.4 and 4.2 overall WAR respectively, and 3a) THE BREWERS PLAY IN THE NATIONAL LEAGUE.

On any given day, there are 14 DH spots to go around.  Granted, they aren’t always efficiently filled.  But you have 32 guys in your first two tables, and most of them have a plus overall WAR.  You’re using new numbers to prove a very old point: some guys are awful with the glove, but hit well enough to play every day anyway.  We’ve known that since the days of Dan Brouthers.

Jeffrey Gross
13 years ago


Not all infielders can play outfield and vice versa. They involve arguably separate skill sets.

Hence, that Headley is bad in the OF, but decent at 3B is not surprising.

dave silverwood
13 years ago

I agree,but with Dunn I Dont believe that is the case.

13 years ago

I think the point is to determine what is the ideal position is for these players. If that position is taken, or non-existant, then someone will have to shift.

Paul E.
13 years ago

Great piece but,
  1) The unique thing about baseball (pre-1973) was everybody played offense and defense (except full-time pinch hitters).

  2) The DH, on average, is the highest paid player in the union (and probably not deservedly so).


dave silverwood
13 years ago

I am 67 years old 60 years a baseball fan not only do I hate the dh,but my favorite al team ,which is the Tigers dont get played on apba against anyone but nl teams-Iam fortunate to be e true Reds fan and a definate old prude who also is discussed about the state of 30 mlb teams—-thanks to you I got to say that.