<$BlogRSDUrl$>

Talking Baseball

Your weekday baseball fix. Some days.



Posted by Ben K. on Sunday, March 07, 2004

Statistical Differences in DHing

Back on Thursday, February 26, following Bernie Williams' emergency appendectomy, I wrote about the Yankees' strengths and weaknesses. In this post, I suggested, among other things the following about Jason Giambi's production when he's playing first base as opposed to when he's the designated hitter:
"Looking at last season, during which Giambi had 50 more at-bats while playing first than while DHing, I noticed that Giambi's offensive production increases dramatically when he's playing the field. In 292 at-bats as a first basemen, he hit .270-.441-.592. He launched 25 homers and drove in 66 runners. While DHing, in 241 ABs, he hit .220-.377-.452. He homered only 16 times and drove in 41 times. Clearly, when Giambi plays the field, he is more focused on all aspects of his game, including his potent offense."
Almost immediately, our readers (and my fellow bloggers) started calling me out on this assumption. Many of you believed that Giambi's injuries prevented him from producing and not his stints as the DH. An e-mail from David Blackburn sums up the critiques of my argument. He wrote to me:
"Isn't it just a little possible that Giambi's numbers are better at 1B than as DH because he was put in as a DH on days that his knees, back, whatever were bothering him more than normally? So, he's playing 1B when relatively healthy and DH when relatively injured and thus has good numbers as a first baseman not because he concentrates better when playing the field (or something like that) but rather because he's healthier?"
Quite frankly, I had thought about this possibility only briefly and dismissed it out of hand. Giambi's knees were bothering him the entire season, and his eye was troublesome for a long time as well. Yet, there was no real consistency to when he played first and when he DHed. Maybe he told Joe Torre he was feeling good on the days he played the field, but I never though that injuries really had that much say in his final statistics. An experienced baseball player, I thought Giambi produced better when he played the field because fielders are more focused on the game. A DH comes to bat every few innings and doesn't have the same levels of adrenaline or stimulation that those playing in the field enjoy. If my ideas were correct, I imagined that a look at Designated Hitters across the league would turn up similar results to Giambi's 2003 splits. Testing my new theory theory, I started with a look at some Giambi's numbers, and what I found not only amazed me but validated by assumption as well.

In 2002, Giambi was unhindered by knee or eye problems. During his first year in pinstripes, he hit .314 with 41 home runs and 122 RBIs. He had a slugging percentage of .598 and an OBP of .435. He walked 109 times and struck out 112 times, for a K to BB ratio of 1.03. Statistically, Giambi's 2002 campaign was one of the best in the league. Looking at Jason's 2002 splits, I found that he had 331 at bats as a first baseman and 229 at bats as the Designated Hitter. Yankees manager Joe Torre used Giambi as DH in order to give the better-fielding Nick Johnson a chance to man first base. Giambi's numbers, however, suggest this might not have been the best idea, and they also suggest that Giambi's limited production last year was due, in part, to his not playing the field.

In those 331 at bats as a first baseman, he hit .344 with an astounding .674 slugging and an OBP of .461. He homered 29 times and his K to BB ratio was 0.88. In the 229 at bats as the DH, he hit only .271 with a .489 slugging and an OBP of .397. He homered only 12 times and his BB to K ratio was 1.27. Those numbers, in a season where Jason stayed healthy, show a clear drop in production on days that Giambi was DHing instead of playing the field. So, in summary:

ABAVGOBPSLGK:BBHR:AB
2003First Base292.277.441.5921.0111.68
Designated Hitter241.220.377.4521.1815.06
2002First Base331.344.461.6740.8811.41
Designated Hitter229.271.397.4891.2719.08


From this table, it's easy to see that as a DH, Jason Giambi produced far less than he did when playing first base. To further test my assumption that hitters hit better on days they played the field, I decided to examine numbers from other players who spent a considerable amount of time DHing and playing the field. (Note: Not all teams are represented here because these teams did not have established DHs or players who split enough time at DH and in the field for me to adequately test my assumptions. For the sake of this post, I picked only the guys who were clearly split between the field and DH and not just those guys who happened to be filling in as DH on a certain day like Matt Lawton, Ellis Burks, and Travis Hafner did last season for the Indians.)

Frank Thomas -- Last season, the White Sox slugger finally had a healthy season. In fact, he held up enough for the Sox to stick him at first base for 29 games last season. For the other 124, he was the DH. His aggregate numbers were quite impressive. He hit .267 with a .562 slugging and a .390 OBP. He drove in 100 runs while slamming 42 home runs. His K to BB ratio was 1.15. These numbers were three-year highs for a the aging slugger. Looking at Thomas' splits, however, suggest that he could have been even better last year.

ABAVGOBPSLGK:BBHR:AB
First Base91.352.487.7251.0011.38
Designated Hitter453.252.371.5321.1813.32


Thomas seems to be one hitter who definitely benefits from playing in the field. Yet, because the White Sox have Paul Konerko, Bill James' number 1 fielding first baseman in the majors, to anchor the infield, Thomas doesn't get too many opportunities to play the field. When he does though, the White Sox offense certainly enjoys a substantial boost. As you may have guessed, the obvious question arises then: why don't these DHs play the field more? The answer I will get to later on.

Rafael Palmeiro -- Raffy, a former Gold Glove winner, DHed 97 times last season and played first 55 times, spelling Mark Teixeira for about a third of the season. Despite a slower bat, Palmeiro put up solid numbers last season; he hit .260, slugged .508, and had a .359 OBP. He homered 38 times or once every 14.76 ABs. His K to BB ratio was 0.92. Palmeiro's splits, once again, support my assumption.

ABAVGOBPSLGK:BBHR:AB
First Base200.275.386.5600.7212.5
Designated Hitter359.251.343.4791.0616.32


Unlike Thomas, Palmeiro is relatively injury-free. Also unlike Thomas, Palmeiro does not have one of the top defenders backing him up; in fact, he is one of the top defenders at first. If his production is so much better when he plays the field, what explains Buck Showalter's decisions to slot Palmeiro at DH and Teixeira at first?

Erubiel Durazo -- By all accounts, Durazo had a disappointing first full year on the Athletics. Billy Beane and the A's expected more than a .259 average, a .430 slugging, an .374 OBP, with 21 home runs (one every 25.57 at bats), 77 RBIs and a K to BB ratio of 1.05. Maybe the problem was that Durazo spent much of the season as the DH, leaving first base open to Scott Hatteberg. Take a look at his splits:

ABAVGOBPSLGK:BBAB per RBI
First Base113.283.403.4251.005.38
Designated Hitter424.252.366.4321.067.57


Durazo's numbers as a first baseman were more along the lines of what the A's had expected, and maybe the problem was that Durazo, who came from the Diamondbacks, was not used to DHing every day. During 2002, when he played in the NL, Durazo had only 20 at bats as the DH and almost 200 at first base. As the Diamondbacks' first baseman, he was more involved in the game and was thus more focused when he came up to the plate.

Tim Salmon -- Although Salmon is one, like Giambi, who has faced numerous injuries recently, his numbers at DH show a dramatic decline compared to those when he plays the field. Overall, last season, Salmon hit .275/.374/.464. He had a 1.21 K to BB ratio, and he drove in 72 runs. (Since his power numbers are fairly underwhelming, I'm using RBIs, as I did with Durazo, as an indication of Salmon's production. Most would argue that these are more important than home runs anyway.) Salmon had a solid season on an underachieving team, but could he have done better?

ABAVGOBPSLGK:BBAB per RBI
Outfield278.299.396.5141.136.62
Designated Hitter248.250.351.4111.308.27


In Salmon's case, his injuries probably contributed to the difference in his splits as much as playing in the field did. However, it's hard to ignore that Salmon's production declined as it took him 1.65 ABs longer to drive in a run when he DHed.

Aubrey Huff -- Now, I would like to take a look at one more hitter; this one, however, is not an aging star like Palmeiro or Thomas and he did not suffer from injuries like Salmon or Giambi. Aubrey Huff, the 27 year old outfielder on the Devil Rays, last season played in all 162 games, and he is the only player in my study to do so. During those games, he hit .311/.367/.555 with 34 home runs, 107 RBIs, and a K to BB ratio of 1.51. Of those 162 games, he spent 30 in the infield, 102 in the outfield, and 33 at DH. (Obviously, there's some overlap between the infield and the outfield.) Let's look at his splits.

ABAVGOBPSLGK:BBAB per RBIAB:HR
Outfield393.328.375.5901.865.0117.09
Infield107.243.336.4301.0810.726.75
Designated Hitter136.316.367.5511.097.1619.42


Huff's numbers show us what happens when a young player DHs. His numbers as a DH slumped compared to when he played the outfield. When he played the infield, however, is a different story. Maybe Huff really gets nervous playing third or first, and this tension carriers over to his at bats. He's more comfortable in the outfield, and thus, when he plays in right, he is better able to get on base, drive runners in, and hit for power and average.

David Ortiz -- The Red Sox slugger had a break out year last season; while he opened this season at first base, by the end of the year, he was crushing the ball as DH. In fact, Ortiz's splits are the only ones that do not adhere to my theory. Overall, Ortiz hit .288/.369/.592 with 31 home runs, 101 RBIs, and a K to BB ratio of 1.43. Let's see his splits:

ABAVGOBPSLGK:BBAB per RBIAB:HR
First Base161.255.346.4221.436.1940.25 (!)
Designated Hitter277.310.389.7001.313.7910.25


Ortiz is the anomaly of the group; he actually benefitted from not playing the field. Additionally, it's no secret that Ortiz doesn't like playing the field. Under no circumstances should the Red Sox have Ortiz play the field with Kevin Millar on the bench. Bill James' defensive stats show Millar to be a significantly better defensive first baseman than Ortiz. Ortiz's range factor was a paltry 8.86 while Millar's was 9.90. Even a hobbled Giambi (9.30 RF) makes more plays than Ortiz does in the field. This, I think, explains the difference in his stats. His reluctance and inability to play first leads him to be hesitant at the plate. Plus, when he was playing first early in the season, Grady Little was unsure how to use Ortiz. He was sitting many days while Jeremy Giambi was playing, and it was only Giambi's injury that propelled Ortiz to the spotlight as the DH.

Having completed this examination, I would like to make note of one thing analytically. One season is not a very good sample size. Yet, looking at career splits, the same patterns emerge. I did not include them hear in the tables because then this long post would simple become unwieldy. it's time now to look at the aggregate results. Throwing out Huff's differences between the infield and DH (but keeping the outfield-DH differences; after all, the OF is his natural position) and Ortiz's numbers, based on the 2003 data, players improve on their batting average by .051. OBPs rise .054 percent, and slugging rises by .105, on average. When you include Ortiz, BAs go up around .038, OBP by 0.42, and slugging by 0.58. The large drop in slugging occurs when you realize that Ortiz, somehow, slugged almost .300 higher as a DH than as a first baseman. I am skeptical to include Ortiz in this study because it looks like something other than time in the field was at stake. I think overall playing time and consistency had more of an effect on Ortiz than anything else. In every case, hitters drove in more runs and hit home runs more frequently when they were playing the field. Interestingly, hitters, except for Huff and Ortiz, also struck out less and walked more often when they had regular at bats while playing the field. Huff's numbers improve when he plays the field except for this stat, and I can't quite figure out why. If anyone has any suggestions, send them my way.

So it seems that, except for David Ortiz in general and Aubrey Huff's specific K to BB ratio, all other stats improved when a player spends more time in the field. In my mind, two important questions arise out of this conclusion. First, what accounts for this rise in productivity? As I explained briefly before, when you are in the field, you are much more focused on the game. Your adrenaline is higher; your engine is higher; and your preparedness is higher. A DH who gets four or five at-bats a game won't be as involved in the game. To stay loose, he'll probably return to the clubhouse in between at bats and have the trainers stretch him or simply ride the bike to keep the blood flowing. He won't be gearing up to make a play or have the ball hit at him with every pitch. He won't already be tracking his own pitcher throw hard. Instead, he'll jump in there every 9 batters and swing away off of live pitching. Even if he's in the clubhouse taking BP swings, the difference in speed and location between batting practice and in-game pitching is simply incredible. I believe that DHs are simply not as adjusted to playing the game and are therefore less focused on their at bats than those men in the lineup who are out in the field for half of the game.

The second question that naturally follows from the first then is why don't these players just play the field instead of DHing? It would seem that they would be more productive and a better help to their teams if they did this. The answer lies in part in David Blackburn's e-mail. I don't believe that these guys could play the field for an entire season. Palmeiro's old; Tim Salmon's fragile; Giambi's knees were killing him by the end of last season; and Frank Thomas is blocked by the top fielding first baseman in the league who isn't a bad hitter himself. (Don't look at Konerko's 2003 stats; look at the years preceding that. If Konerko doesn't return to 2001-2002 form, I think Thomas should just play first everyday.) For these players to produce at all, the DH spot gives them the opportunity to rest aging bodies. While they may build up more strength and stamina by playing in the field, by August, these players would be worn down. They DH simply because their bodies cannot take playing the field every day, not because they want to.

Finally, I would like to pose a different question: What should we expect from some of these players this year? In the case of Ortiz, I would say more of the same. He'll kill the ball, again, though not level with his DH stats from last year. I don't think he'll slug .700 and a drive in a man more frequently than every four times to the plate. If Giambi's knees are healthy, he'll have a monster season. Playing in the field seems to benefit him quite a lot. I liked watching him hit a Grand Slam yesterday. That's always reassuring. Thomas won't play the field, unfortunately for him, and Palmeiro will see some action at first and some at DH. He is pretty old after all.

So there you have it. While David Blackburn brought up an interesting point about Giambi's injury, I think injuries only account for some of the difference. Injuries are why I omitted Mike Sweeney. He refused to go on the DL when injuries forced him out of the field last season. As the DH, he was horrible. Yet, injuries do not account for the differences we see in Huff's, Thomas' or Palmeiro's numbers. I believe that something else affects their production, and that something else is a lack of focus from not being totally involved in the game. Players on the bench are never as focused on the game as those in the field, and the Designated Hitter is no exception to this rule.


### So what do you think? We want to know. | | E-mail us ###