Harmon Killebrew is in our backlog. You’re likely surprised to see someone in every other hall of fame, someone with 573 homers, outside our proverbial candy store looking in. One of the biggest reasons we haven’t voted for him is his defense. It’s really bad.

It’s too bad for him that the designated hitter wouldn’t be adopted until 1973. He might have avoided all that bad fielding and let his bat do all the talking for him. Right? Maybe, maybe not.

BBREF’s WAR breaks out its components so we can play around with them. One of those components, abbreviated Rpos, is an adjustment made to account for the relative difficulty of each position. For example, in today’s game, BBREF adds about 8 runs to a shortstop’s ledger. First basemen have about 10 runs subtracted from theirs. To put it simply, the difference in fielding difficulty between shortstop and first base is around 17 runs, or about 1.5 wins. That’s a lot when an average player is worth two wins.

What this means is that we can answer a question such as “*What if Harmon Killebrew had always played DH?*” For Killebrew we know the following information:

Batting runs: 487

Base running runs: -24

Grounding into double play runs: -27

Fielding runs: -78

Position adjustment runs: -77

Total runs above average: 284

All we have to do is replace Killer’s position adjustment runs with the adjustment he’d get as a DH in the same playing time and then remove his fielding runs.

Unfortunately, BBREF doesn’t give us the by-position breakdown of Rpos for a player, so we need a hack. Killebrew is a great example to work with because he straddled the DH era. We can figure out a player’s Rpos at every fielding position using the table on BBREF’s WAR Explained for hitters page. We know Killebrew’s innings at each position, and the values on the table are all rates per 1350 innings, so some simple algebra will do the job. If we sum our results at each position he played in a given year and subtract from his total Rpos for the year, the difference is his runs as a pinch hitter, pinch runner, or DH. We then replace his by-position Rpos with the value if he’d been a DH (using the defensive innings) and add it to his non-fielding Rpos to get our estimate of what his DH-only Rpos would be. Then we can plug it into the Runs Above Average above.

Here’s an example from Killebrew’s 1974 season. He appeared at DH 57 times and played 223.67 innings at first base. First basemen in 1974 are debited 9 runs per 1350 innings, which means Killer is adjusted downward by about 1.5 runs. His total Rpos for the year was -7, which means his DH and pinch hitting appearances totaled -5.5 runs. Switch that first base Rpos to DH Rpos, and he would be debited -2.5 runs for those appearances. Which added to the -5.5 DH Rpos is a total of -8 runs. Now we return to his runs above average:

Batting runs: -4

Base running runs: -1

Grounded into double plays runs: -1

Fielding runs: -1 0 (notice, we have removed his fielding)

Position adjustment runs: -7 -8 (notice we have substituted the Rpos we calculated above)

Total runs above average: -1.5 (which is identical to what he produced as a part-time fielder)

In this case, it’s a wash. But when I took Killebrew’s career as a whole, season by season, the story was very different:

Batting runs: 487

Baserunning runs: -24

Grounding into double play runs: -27

Fielding runs: -78 0 (removing fielding runs)

Position adjustment runs: -77 -215 (swapping in DH-only runs)

Total runs above average: 284 221

Just being able to stand at a position without fielding as poorly as I would was worth about 60 runs to Killebrew and his teams! Unquestionably, Killebrew was better off in the field. Probably at first base, but the Twins had Don Mincher at the same time. That’s how much being a DH takes a bit out of a player’s value. Why? Because, generally speaking, a replacement level DH is a league-average hitter. That’s why David Ortiz, despite his 385 batting runs and 139 OPS+ in 8851 PAs, has only mustered 16.9 wins above average and 47.7 WAR in his career. His debit for DHing is 183 runs.

But all this brings up two questions. All other things being equal, how bad a fielder would a player have to be for his team to be better off DHing him? Well it depends on his position, of course. Derek Jeter’s defense is much maligned, but he’s nowhere near DH-only. During his career, shortstops had a positional adjustment of +7.5 runs per 1350 innings. So the break even appears to be -7.5 fielding runs. Except that DH is 22.5 runs easier per 1350 innings than shortstop is. Jeter is boosted +135 positional runs as a shortstop, even though he’s worth -220 runs in the field. If he were a DH, he’d have no fielding value, but his positional adjustment would be more than -250 runs. Let’s lay this out:

Batting runs: 353

Base running runs: 56

Grounding into double play runs: 8

Fielding runs: -246 0

Position adjustment runs: 138 -251

Total runs above average: 305 166

See, the positional adjustment totally swamps the gain in fielding runs. Now as to whether Jeter should have stayed at shortstop his entire career….

The question, then, is whether *anyone *is worth making a DH-only because he is that bad with a glove on? The answer is yes. Gary Sheffield, for example, has the second worst career fielding runs (after Jeter) of anyone in history with -195. If we swap in DH positional adjustments and remove the fielding, look what happens.

Batting runs: 561

Base running runs: -1

Grounding into double play runs: -11

Fielding runs: -195 0

Position adjustment runs: -84 -227

Total runs above average: 270 322

Sheffield’s teams would have been better off were he a full-time DH. By like five wins’ worth of runs.

Danny Tartabull is a similar story. Another infielder turned outfielder who was good at neither. When I put him through the process above, he leapt up by 70 runs. Jeff Burroughs would have been about five-wins better as a DH-only.

But rather than go through this lengthy process for a whole bunch of plodding defenders, we can use a quick and dirty inequality to see who other candidates might be

(Rfield + Rpos) < (fldINN * (-15/1350))

This isn’t perfect for someone like Adam Dunn who spent a good chunk of time at DH, but it identifies guys well enough. Here are some with long careers:

- Adam Dunn
- Frank Howard
- Matt Stairs
- Greg Luzinski
- Preston Wilson
- Jason Giambi
- Raul Ibanez
- Ken Singleton
- Dante Bichette
- Jorge Orta

Realistically, Jason Giambi is the only one for whom this might be an interesting what-if scenario as regards his case for the Hall of Miller and Eric. I used the long method to see if that’s so:

Batting runs: 444

Base running runs: -15

Grounding into double play runs: -10

Fielding runs: -83 0

Position adjustment runs: -136 -191

Total runs above average: 199 228

When I plug DH-only Giambi into my personal evaluation system for the HoME, the extra 30 runs would likely slide him just beneath Mark McGwire. He would end up just above Will Clark and John Olerud who are very close to the line indeed but just below it. He’d be a really tough call, especially for someone who prefers a candidate with strong peak performance.

Overall, what we find is that many players whose value we think might benefit by being lifetime DHes actually wouldn’t. At least not in a theoretical vacuum. Of course, teams push their chess pieces around to best meet their real-life strategic needs—platooning, rotating regulars through the slot to keep their legs fresh, putting a brittle hitter there to keep him whole, or simply playing the lesser fielder at DH even if his fielding may not be *that *bad. We don’t really do what-ifs at the HoME, but we do enjoy thinking about them. Luckily, we don’t have to worry about them like real teams do.

—Eric

## Discussion

## No comments yet.