Batter's Box Interactive Magazine Batter's Box Interactive Magazine Batter's Box Interactive Magazine
The DH was implemented in the American League in 1973 with the stated intention of increasing offence. Unlike most of major-league baseball’s innovations, this one actually worked, kind of. Scoring, which had been in the doldrums, perked up: the AL’s collective batting average rose 20 points, and the AL has outscored the NL in runs per game in 31 of the 32 seasons that followed (interestingly, the NL actually outscored the AL by 0.05 runs per game in 1974, the year after the DH’s introduction).

The DH, who functions as a kind of permanent pinch-hitter for the pitcher in the AL, is one of the more important offensive positions. DH production has been well above the league average (measured here, and elsewhere in this article, by unadjusted OPS; my statistical arsenal is limited for this study) in every season going back as far as 1990. Here are the above-average offensive positions, by OPS, over the most recent five seasons, with the league-average OPS supplied as well:
LF	795
1B	793
3B	792
DH	784
RF	780
AVG	770

1B	813
LF	801
RF	790
DH	777
AVG	760

1B	834
DH	783
RF	782
LF	756
AVG	755

1B	859
DH	776
RF	766
ABG	761

1B	879
DH	820
RF	810
LF	810
AVG	792
Arguments in favour of and against the DH have been continuing on and off since at least 1973, and they occasionally approach holy-war status. Speaking for myself, I don’t really care. I just want to see consistency between the leagues. Thanks to free agency, inter-league play, a joint umpiring crew, and the dissolution of the AL and NL Presidents’ offices, the two leagues are about as similar now as the AFC and NFC. Either adopt the DH universally or get rid of it altogether.

Anyway, that’s not why I’m here today. My interest is in the nature of the DH position itself. It’s obviously a unique creature, unlike any other position in the game. Since the amount of offence it generates is substantial, it makes sense for teams to figure out the best way to maximize production from their DH spot at minimal cost.

So I began doing some digging, nosing around in the stats to see if I could find anything interesting. Well, I didn’t. It was a boring hodgepodge and a waste of my time.

But I did find that three distinct approaches to manning the DH spot can be divined among AL teams. And I realized something I had only been dimly aware of before: the concept of a single, individual “designated hitter” is largely a myth.

The DH is, fundamentally, a position, not a “man.” He’s a composite player, a collective offensive force. Most players who DH during the course of a given season also play numerous games at various positions in the field – often, they’re in the field more often than they DH.

Most fans, when conducting off-season assessments of their team’s offence, will invariably say something like, “X will be the DH,” or ocasionally, “A and B will platoon at DH.” They miss the fact that few teams have “a” single or even primary DH – many teams send a collection of players through the revolving door of one of the least-understood roles in the game.

DH History

The earliest DHs were not regulars in the position, either; teams would often rotate numerous hitters into and out of the DH slot. Indeed, Ron Blomberg, the very first DH (he walked with the bases loaded against Luis Tiant on April 6, 1973) played 41 of his 95 games that year at first base. But the myth of the “individual DH” is strong in baseball culture, even though the facts don’t fully support it.

The lineage of the modern “single DH” idea probably extends back to the wave of high-profile players who starred in the role throughout the mid-1980s and into the early ‘90s. In almost all those cases, these players were aging sluggers whose defence had fallen off (or really, was never that great to begin with). In almost all cases, DHing allowed them to extend their careers, often into Cooperstown territory. And in almost all cases, they’ve now been retired for several years.

Consider: only 17 different players have been named DH of the Year (now the Edgar Martinez Award) since the position was introduced 32 seasons ago. Among the multiple winners are Willie Horton (2), Greg Luzinski (2), Harold Baines (2), Dave Parker (2), Don Baylor (2) Paul Molitor (2), David Ortiz (2), Hal McRae (3) and Martinez himself (5).

Baines leads all DHs in games played, AB and hits (keep in mind, this refers to performance produced specifically as DH, not throughout one’s overall career). Martinez leads in the Triple Crown categories of BA, HR and RBI (and is among the top three in the other columns). The "DHs of the Year" listed above occupy almost all the spots in the Top Ten in each of these offensive categories. Basically, the “individual DH” club is an exclusive one, and has been for at least a decade. (Props to the site, from which these and many subsequent stats were obtained.)

Let’s take a look at perhaps the most telling statistical category for DHs: Games Played (GP). Here’s the all-time top ten list, with some other stats added:

Player	         All GP	      GP@DH	%AllGP@DH	DH Age

Harold Baines    2712	    1652	61%		28
Hal McRae	 1926	    1427	73%		30
Edgar Martinez	 2003	    1412	70%		31
Don Baylor	 2255	    1285	57%		32
Paul Molitor	 2309	    1174	51%		34
Chili Davis	 2355	    1184	50%		30
Frank Thomas	 1903	     932	49%		30
Jose Canseco	 1917	     837	44%		29
Brian Downing	 2284	     824	36%		36
Willie Horton	 1943	     753	39%		32
There are three things worth considering here. First, most of these hitters played up to 50%-60% of their career games at DH (that’s the third column – what percentage of their total career appearances were at DH). The rest of their appearances were at a given position, usually earlier in their careers.

Secondly, most of these players were 30 or older when they began making the majority of their appearances at DH (the fourth column – the player’s age in his first season when DH at-bats began outnumbering position at-bats). In other words, none of these players started off their careers as a DH. In fact, for all the current controversy over whether a “career DH” like Martinez should be admitted to Cooperstown, hardly anyone acknowledges that there have only ever been two “career DHs” worthy of the name (70% or more of their career at-bats at DH): Edgar Martinez and Hal McRae.

But the third, and most remarkable, aspect of this list of DH appearances is that Willie Horton, a player who retired 25 years ago, is still in the top ten. Brian Downing, who last played in 1992, is right ahead of him. If DH were a position given over to individual players on a regular basis, don’t you think someone would have bumped these two guys off the top-ten list a while ago?

There is only one active player among the top ten DH appearances, and that’s Frank Thomas. If the Big Hurt plays until he’s 40 (four more seasons) and averages 120 DH appearances per season in that time, he’ll end tied for third place with Martinez. That’s how easy it would be for an active DH to scramble up the DH-GP list.

But it’s not happening, because there have been very few players in the past decade or two who turned to full-time DH duty in their 20s and were good enough hitters to last. Put it this way: if David Ortiz were to play 120 games a season at DH for the next ten years, then he’d surpass Baines’ DH-GP record.

What does all this mean? It means that the full-time exclusive DH is a rare bird, and that he might be nearing extinction. But there are a few active players now who might be able to change that.

DHing today

Where have all the great DHs gone? Who’s manning the position today? Some more recent stats tell the story.

I dug up the complete list of DH appearances for each AL team in 2003 and 2004 – all players who filled the DH role, even for one game. Most teams sent about a dozen men to bat as DH; the Blue Jays, as Magpie pointed out in another thread this year, used an astounding 18 players in the position in 2004. To choose three other teams at random: the Indians used 11 players at DH, the Royals 12 and the Twins 15.

Then I reviewed the production of each player sent to bat by his team at DH, and noted each player who recorded more than 100 AB specifically in that role (there weren’t many). One hundred is admittedly an arbitrary number, but it seemed to make for an effective threshold separating the more serious DHs from the mere dabblers and September callups.

The next step was to zero in on the player who took most of his team’s DH at-bats and calculate those appearances as a percentage of his team’s total. (To take a basic example, if the Jays had 600 total DH ABs and Josh Phelps took 400 of them, his percentage would be 66%.)

What was all this statistical labour in aid of? I wanted to determine three things:

1. How many teams had one player take more than half of all DH at-bats?

2. Of the other teams, how did DH usage break down?

3. Which method of DH construction was the most effective?

Here are the results for 2004:

2004 DH Production

Team		OPS	Top DH	  %DH AB  Salary  Type
Boston		953	Ortiz	  75%	  $4.6M      A
Cleveland	953	Hafner	  79%	  $0.3M      A
Oakland	        884	Durazo	  83%	  $2.1M      A
Kansas City	856	Sweeney	  33%	  $11M       B
ChiSox	        831	Thomas	  40%	  $6.0M      B
Minnesota 	786	Offerman  21%	  $0.5M      C
New York	778	Sierra	  38%	  $1.0M      C
Texas		776	Fullmer	  42%	  $1.0M      C
Detroit		775	Young	  49%	  $7.7M      C
Seattle		754	Martinez  78%	  $3.0M      A
Anaheim	        725	Salmon	  26%	  $9.9M      C
Toronto	        693	Phelps	  45%	  $0.3M      C
Tampa Bay 	677	Huff	  23%	  $2.7M      C
Baltimore	649	Newhan	  22%	  $0.3K      C

OPS = team’s total OPS from the DH position

Top DH = player who took the greatest number of team DH AB

% DH AB = percentage of the team’s total DH AB taken by Top DH

A = One hitter takes more than half the DH at-bats; nobody else has >100

B = Two hitters combine to take more than half the DH at-bats between them; each with > 100

C = No hitter takes half the DH at-bats; only one with >100

A few notes:

1. The three teams with the best DH production all gave one hitter the great majority of DH at-bats. This is hardly surprising: if you have a terrific hitter to slot into the DH spot, you’ll use him as often as possible. Only the Mariners, in Edgar Martinez’s final year, got poor production out of a full-time DH.

2. Of the ten remaining teams, eight (the C teams) distributed DH at-bats among a plurality of hitters, none taking more than half. The two exceptions were Mike Sweeney and Ken Harvey in Kansas City, and Frank Thomas and Carl Everett in Chicago. However, the ChiSox and Royals benefited from great OPS production from Thomas (972) and Sweeney (918); their "platoon partners" Everett (752) and Harvey (846) dragged down the team total. So B teams, in a sense, are actually A teams whose best hitters just didn’t get enough at-bats.

3. C teams, as is clear from the chart, did the most poorly from the DH slot. You could argue that they simply didn’t have the horses to run out there, so they plugged in whomever they could find in hopes that it would work. But I think these teams could have done better. Josh Phelps, for instance, had an OPS of 770 at DH; the Blue Jays’ total DH OPS number (693) was much poorer because other players did much worse in the position. Troy Glaus gave the Angels an 889 OPS in 133 ABs, but the Halos were done in by Tim Salmon and his 539 OPS in 145 AB. Dmitri Young posted a solid 837 OPS in 283 DH AB; Rondell White’s DH appearances (781 OPS in 163 AB) hurt the Tigers’ overall DH production.

4. Only one team gave three players 100 or more at-bats at DH in 2004. Would you believe it was the Yankees? Ruben Sierra got 201 DH AB, Bernie Williams 189 and Jason Giambi (okay, he’s borderline) 99. Even more remarkably, these three hitters’ respective DH OPS (721, 726 and 765) were each less than the Yankees’ overall total of 778. In a sense, the Yankees kept sending the wrong guys out there to DH.

5. The salary column is interesting, insofar as some teams really didn’t get much bang for their buck. The Tigers and Angels paid out $17.6M for second-division DH production (in case you’re wondering, Young and Salmon posted better OPS totals as DHs than in the field). The best investment clearly was Cleveland’s, paying the league minimum for Travis Hafner’s tremendous output. A lot of other teams nickel-and-dimed the DH slot, and got what they paid for.

2003 DH Production

Team		OPS	Top DH	%DH AB	  Salary  Type

Boston		908	Ortiz	  49%	  $1.2M      B
Seattle 	886	Martinez  90%	  $4.0M      A
Chicago	        860	Thomas	  82%	  $5.0M      A
Texas		811	Palmeiro  61%	  $9.0M      B
Cleveland	808	Burks	  33%	  $7.2M      C
Toronto	        800	Phelps	  63%	  $0.3M      A
Detroit 	782	Young	  59%	  $6.7M      B
Minnesota	779	LeCroy	  36%	  $0.3M      C
New York	771	Giambi	  44%	  $11M       C
Anaheim	        753	Salmon	  44%	  $9.9M      C
Oakland	        749	Durazo	  76%	  $1.1M      A
Kansas City     736	Sweeney	  41%	  $11M       B
Tampa Bay	867	Martin	  30%	  $0.3M      C
Baltimore	685	Segui	  34%	  $7.0M      C
Some more notes:

1. Boston and Texas came within inches of being A-teams (so to speak). David Ortiz single-handedly lifted the Red Sox to the top of the DH pile; even though he took only 49% of the team’s DH at-bats, he posted a tremendous 1075 OPS therein. Juan Gonzalez just barely tipped over the margin of 100 DH AB for the Rangers (103, to be precise), qualifying them as a B-team. A few more DH AB for Ortiz and a few less DH AB for Gonzalez, and you’re looking at 4 A-teams at the top of the DH OPS list.

2. What’s Cleveland, a C-team, doing this high on the list? Simple: they gave too many DH AB to Ellis Burks (784 OPS in 190 AB) and not enough to Travis Hafner (867 OPS in 149 AB). The Indians rectified that situation in 2004.

3. Would you believe that the Yankees, again, gave three players more than 100 DH AB? Jason Giambi posted an 829 mark in 241 AB, and Nick Johnson weighed in with 861 OPS in 107 AB. Dragging down the team’s overall mark was, once again, Ruben Sierra (709 OPS in 103 AB). Only Sierra and Tim Salmon have been so consitently poor for their teams at DH the last two seasons. The Angels were stuck with Salmon’s contract, but I don't know why the Yankees keep bringing Sierra back (again in 2005) for a role in which he appears to be hurting his team.

4. Minnesota consistently throws a grab bag of players into the DH slot, yet they consistently end up about league-average at DH, whereas most C-teams are well below the median. Not only that, most of their DHs worked for the league minimum or close to it. I can’t find any other reason for this success except good fortune, and accordingly, I wouldn’t advise Ron Gardenhire to try it a third time.

5. In his two years at DH for Toronto, Josh Phelps was at least league-average, if not more: in 2003, his 863 OPS was 63 points higher than the Blue Jays team mark. And here’s something to consider for those folks (including me) who've been stumping for Frank Catalanotto to be Toronto’s primary DH against righties: in 2003 and 2004, Cat’s DH OPS was 633 in 82 AB and a stunning 503 in 96 AB, respectively.

6. Oakland started in 2003 giving most of their DH at-bats to Erubiel Durazo. It didn’t pay off in ’03; it paid off handsomely in ’04, and will probably continue to do so for a few years yet.

7. The Royals and Mariners have depended heavily the last two years on the health and productivity of expensive vets like Mike Sweeney and Edgar Martinez at DH. Some years, that works; some years, it doesn’t.

8. Yes, that’s Al Martin leading the 2003 Devil Rays to 2nd-last place in DH production. Yes, that’s $7M David Segui leading the Orioles to last place in DH production. The Orioles and Devil Rays haven’t a sweet clue what they’re doing at DH.


Can we draw any lessons for building a good DH? Here are some possibilities:

1. Treating your DH position as a revolving door is not a method for success. If you don’t have one dominating player to fill the DH spot, the odds are stacked against you.

2. Lacking a Hafner or an Ortiz, your best bet seems to be to assign two hitters to fill the primary DH roles (a L-R platoon would be nice, but seemingly not necessary). Keep in mind, however, that almost every B team had a primary DH pulling down $6M to $11M.

3. Populating the DH with a cast of thousands doesn’t work. Fourteen teams tried it in 2003-04, and 12 ended up with below-the-median DH production.

4. The median team DH output, in case you’re wondering is about a 780 OPS. Shea Hillenbrand’s OPS was 789 in 2002, 782 in 2003, and 812 in 2004. Draw your own conclusions.

5. There is no correlation between primary-DH salary and DH output. There are as many teams with low-cost superstars at DH as there are teams with vastly overpaid DHs in the bottom half of league production. Put differently: out of the top ten $6M-plus primary-DH salaries in ’03-‘04, six finished at or below the league median.

The easiest route to success appears to be: get yourself a late-20s slugger and stick him in the DH spot. The top three DH teams in 2004 each feature a relatively young hitter (David Ortiz, Travis Hafner and Erubiel Durazo) ripping the cover off the ball at a reasonable price, and each can be expected to continue to produce big numbers for at least a few more years. Will they be this generation’s Martinez and McRae?

Those three players have one other thing in common, too: they were all acquired from other organizations, at virtually no cost. Which leads to about the only real conclusion you can draw from all of this: smart teams that can judge talent can find affordable, productive DHs.

The Myth of the Designated Hitter | 5 comments | Create New Account
The following comments are owned by whomever posted them. This site is not responsible for what they say.
Mick Doherty - Monday, March 14 2005 @ 10:40 AM EST (#106036) #
Awesome. Best of show for the new Batter's Box so far. The Yankees have always screwed up their DH use (more on that in my preview in 10 days).

I've always wondered if managers don't use the DH sometimes to "reward" players they feel bad about not starting in the field for whatever reason. What a bad idea!

Here's a thought ... if the NL adopted the DH, do you think the "strategy league" would quickly be "better" at it than the AL? Would the NL teams, in the post-Moneyball era, take a study like this one -- and presumably something far more historically exhaustive -- and "do it right"? Or would Tony Womack have been the Cardinals' primary DH last year, with So Taguchi getting the rest of the AB?

Mike Green - Monday, March 14 2005 @ 11:22 AM EST (#106039) #
Interesting, Jordan.

In the AL, the ability to place a player at either first base or DH creates some interesting management decisions. At the end of 2003, the Indians had Hafner and Broussard. They could very well have played Hafner at first, and looked for a veteran DH. They chose to go for the low-cost option and it paid off, as Broussard had a great age-27 season. Whether this decision is in the long-term best interests of the club remains to be seen.
Gerry - Monday, March 14 2005 @ 11:32 AM EST (#106040) #
Roster construction plays a role in this too. If a team carries 12 pitchers they have thirteen spots available for hitters. Eight hitters are playing positions so you have five players to use as your DH. You need a backup catcher, a utility infielder, and a backup outfielder, leaving two spots. If you platoon anywhere you need extra players for those positions and they are unlikely to DH becuase if they are sitting its because they bat on the wrong side. If you carry a full time DH, that is one less fielder available to you.

Some of the teams that do not have a full time DH might have too many platoon players, or too many no hit, all field backups. The bigger budget teams can afford to have backups who can hit and field, the poorer teams might be able to afford only one such skill, John McDonald anyone?

Very interesting read Jordan.
Anders - Monday, March 14 2005 @ 11:35 AM EST (#106041) #
I think that the NL's reputation as a strategy league has to be derived almost completely from their not having a DH. It would seem that the major difference between the NL and the AL is that in the NL there are more players switched in and out (And a greater defensive persuasion, which may or be true.)

Now I dont have anything to back this up, but it would seem that the majority of switches would undoubtedley come from taking out pitchers for pinch hitters. In the AL, where you practically have to have an ops of 1.0000 to be a middle infielder, most switches come on late defensive replacement/pinch runner, and the occasional pinch hitter.

Clearly the difference between a mediocre 2b and a pitcher at bat with the bases loaded in the 7th in a close game is large. Forced to make these moves, it would seem that this would necessitate more intense game management. In the NL, players pinch hit more, thus more change.

What I wonder is if pitcher usage patterns are any different in the NL. The Jays are proably going to carry 12 pitchers (which just seems to be really dumb.)In the NL, they dont have as much luxury in that regard. It would be interesting to find out.

On a side (and unrelated note) why dont teams realize carrying more pitchers does more harm than good?
Magpie - Monday, March 14 2005 @ 06:20 PM EST (#106110) #
Clearly the difference between a mediocre 2b and a pitcher at bat with the bases loaded in the 7th in a close game is large.

In fact, the difference is so large that there's very little strategy involved. You don't let the .140 hitter bat with the game on the line. It's not a real option.

There has historically also been more - what shall I call it - "discretionary" bunting in the AL as well. Because generally 50-55 % of sac bunts in the NL are made by pitchers.

The Myth of the Designated Hitter | 5 comments | Create New Account
The following comments are owned by whomever posted them. This site is not responsible for what they say.