19 October 2013

Making the Orioles a Champion in 2014: Designated Hitter

This post is part of the Making the Orioles a Champion in 2014 Series.  Below you will find links to the other articles.  We will do our best to make sure the links go live with each new update.
C | 1B | 2B (12) | 3B | SS | LF | CF | RF | DH | Bench | SP (1, 2) | RHRP | LHRP | Conclusion

.236/.290/.418.

That's the Batting Average / On-Base Percentage / Slugging Percentage of the Orioles' 2013 Designated Hitters. The OPS is .708.

.256/.320/.404.

That's the Batting Average / On-Base Percentage / Slugging Percentage of the American League. The OPS is .724.

That's right. The Orioles got worse performance out of their designated hitters - a position with zero defensive requirements, a position at which the only responsibility is offense - than the entire American League. They did get a little bit more power than the League, but were substantially lower in on-base percentage.

But wait. It gets worse, from a generic baseball fan perspective, or better, from the Orioles fan perspective. The Orioles .708 OPS from their designated hitters ranked eighth in the American League.

The production of American League teams' designated hitters can be put into five groups. Far and away the best was the Boston Red Sox (mostly David Ortiz), at .957. Here's the breakdown, with teams in the groupings ranked from highest to lowest:

.957 - 1 team (Boston)
.754 - .797 - 6 teams (Kansas City, Toronto, Detroit, Seattle, Cleveland, Los Angeles)
.660 - .708 - 5 teams (Baltimore, Texas, Oakland, Tampa Bay, Chicago)
.615 - .632 - 2 teams (Minnesota, Houston)
.567 - 1 team (New York)

I need to get this on the record. The 2013 New York Yankees at .184/.270/.297, got worse production from their designated hitters, relative to the league, than the 1971 San Diego Padres got from Enzo Hernandez when he notoriously drove in 12 runs in 618 plate appearances.

However, to get back to the Orioles. It's difficult to get a handle on all the statistical measurements out there, but it seems that David Ortiz and the Red Sox designated hitters was worth about four wins. It's even harder to estimate the contribution of the Orioles' designated hitters to their success, but I don't think you can claim that it was worth more than one win. So, given that the 2013 Orioles needed to improve by seven games to make the postseason, perhaps half that could have been made up if the Orioles actually had a good hitter be their designated hitter as opposed to the below-average troupe they used.

Before we look at how the Orioles could improve their 2014 designated hitters, we need to look at the individuals who served in 2013. The Orioles had 576 plate appearances from their designated hitters (which is lowest in the American League by 30. That, at least, reveals that the Orioles knew they had bad designated hitters; they batted their DHs lower in the order than every other team.) 143 came from players who normally played a defensive position but were designated hitting on a partial day off - Chris Davis, Brian Roberts, Adam Jones, Nate McLouth, Nick Markakis and Matt Wieters. If we subtract those players' contributions, we get a different view:

Day-off: .254/.329/.492
Quasi-Regular DH's: .231/.277/.395

Most of the DH productivity came from Chris Davis' 21 plate appearances; he hit .471/.571/1.235. Just removing Davis' 21 appearances leaves the rest of the DHs - both the starters having a day off and the "regular" DHs - at .228/.286/.390.

Historically, some teams have not used a regular DH, instead using the DH position to give regulars partial days off. There are two reasons why that was not a good option for the 2013 Orioles. First, most of the regulars didn't hit well as a DH. Second, if you do give the regular a break from defensive responsibilities, then you're playing a bench player. And, as we'll see when we look at the bench, the Orioles didn't have a good bench. It's doubtful that the benefits of giving players more defense-free days outweigh the losses from their lack of production as a DH and from playing a bench player.

Danny Valencia leaving the batter's box in Norfolk's Harbor Park. (Elaina Ellis / Norfolk Tides)

So, I believe that if the Orioles are to improve their DH production, they'll need to find a good hitter or hitters to be their DH. Fortunately, one of the 2013 Orioles did hit very well as a DH. Danny Valencia had 145 plate appearances as a DH, and produced at a .307/.342/.912 clip. It seems likely that Valencia did most of his damage against left-handed pitchers; overall (both as a DH and as a non-DH) Valencia hit .371/.392/.639 against lefthanders but .203/.250/.422 against right-handed pitchers. During his major-league career, Valencia has hit .329/.367/.513 for an .880 OPS in 431 plate appearances versus left-handed pitchers, which includes his disappointing 2012. Valencia's career production versus left-handed pitchers would be better the production from all but one team's DHs.

So that's the right-handed half of a DH platoon. And the Orioles potentially have the left-handed half already in their organization in Henry Urrutia. Urrutia, of course, is a Cuban defector who played his first season in the U.S. in 2013. He played at AA, AAA, and in the Majors in 2013, and hit right-handers at least fairly well at all three levels - .374/.446/.585 at AA, .338/.376/.500 at AAA, and .294/.294/.333 in the Majors. While it's premature to project Urrutia to be as good against right-handers as Valencia is against left-handers, he has demonstrated the potential to hit .300 against right-handers with line-drive power. He's earned a full chance in the major leagues. A platoon of Urrutia and Valencia has the chance to be one of the most productive DHs in the League.

Henry Urrutia hitting from the left side. (Elaina Ellis/Norfolk Tides)


One of the problems in using a DH platoon is that it takes up two roster spots to fill one position. If a team carries twelve pitchers, and there are eight players in the starting lineup, that leaves five spots for reserve non-pitchers. There must be a backup catcher, a backup infielder, and a backup outfielder. So if you plan to use a DH platoon, you have no more available roster spots.

If the Orioles intend to compete with the financial resources of the Red Sox and Yankees and the seemingly endless parade of prospects of the Rays, they're going to need to think outside the box. And it's my belief that there is absolutely no need to carry twelve pitchers; that the Orioles will gain more from a DH platoon than from a twelfth pitcher.

It's always hard to determine who the twelfth pitcher on a staff is, much less a team like the Orioles which shuttles pitchers up, down, and around frequently. At any given time, though, there are five pitchers in the starting rotation, whoever they may be. The first six members of the bullpen were Jim Johnson, Darren O'Day, Tommy Hunter, Brian Matusz, Troy Patton, and T.J. McFarland. Pitchers occupying the seventh spot in the bullpen were Pedro Strop, Francisco Rodriguez, Jairo Ascensio, Kevin Gausman, etc. etc. None of these guys were notably effective; some were completely ineffective.

And, as best as I can determine, the pitchers in the seventh bullpen spot combined to pitch 72 2/3 innings before August 31, after which rosters expand and roster spots are plentiful. A baseball season lasts almost exactly six months, or 26 weeks; there are 22 weeks when the roster is limited to 25 men. So, the seventh spot in the bullpen pitched about 3 1/3 innings per week. Surely those six other relief pitchers can absorb an extra inning per week.every two weeks.

So, one of the easiest and cheapest moves the Orioles can make toward becoming a champion in 2014 is to drop the twelfth pitcher and platoon Henry Urrutia and Danny Valencia at DH.

12 comments:

  1. I really like this idea but there's no way DD/Buck do this unless they land a top of the rotation starter. I don't see them using a 6man bullpen with a Tillman, Chen, Gonzalez, Norris, and Gausman rotation.

    2013 would have been the year to try this. Britton and Stinson are now out of options leaving the 40 man with only TJ, Mike Belfiore and Steve Johnson to shuttle back and forth.

    ReplyDelete
  2. The trouble I potentially have with an 11 man pen would come to three things:

    1) Why is no one else doing this? Is anyone doing this?

    2) The Orioles' starters are not exactly known for going deep into games.

    3) That 7th man in the pen can be a useful junk arm that enables you to keep your other relievers fresh for meaningful innings.

    Maybe things have shifted too far and bullpens are overstaffed. It may be with the swing back from the offensive era we witnessed in the late 90s and early 00s that a new perspective with the pen needs to be taken. I am not exactly there yet. I'd like to see some thesis on that.

    It is a pretty fascinating question. At what point can you get rid of the 12th man on your staff.

    Personally, I'd like to see veteran starters be available for an inning of junk time in blow outs on their pen days. That probably won't happen anytime soon.

    ReplyDelete
  3. Just discovered your blog. Good stuff, for the most part. Except for one statement, where you sounded like Gary Thorne..."So, I believe that if the Orioles are to improve their DH production, they'll need to find a good hitter or hitters to be their DH. :-)

    ReplyDelete
  4. Patrick - I think with that sentence following the proceeding paragraphs, Joe is saying that the Orioles would be better suited to find a plus bat or two for DH instead of using it as a place to rest players. In that way...it really is not a "the team who scores more will win this game" kind of comment.

    Or, I could just be confused.

    Anyway, thanks for dropping by.

    ReplyDelete
  5. #Jon - I think that teams have gravitated to the 12- or 13- man bullpen because (1) they've gone too far in protecting bullpen arms; (2) managers don't want to be criticized if they lose a game because they can't manipulate the lefty-righty platoon advantage; and (3) they really haven't thought in depth about it.

    But more to the point, I don't think the Orioles can compete by copying everyone else. I'd rather them give the existing bullpen arms another twelve innings a season and get some positive production from DH.

    ReplyDelete
  6. Jon has my point exactly. Mediocre hitters aren't going to cut it.

    ReplyDelete
  7. "Maybe things have shifted too far and bullpens are overstaffed. It may be with the swing back from the offensive era we witnessed in the late 90s and early 00s that a new perspective with the pen needs to be taken. I am not exactly there yet. I'd like to see some thesis on that.

    It is a pretty fascinating question. At what point can you get rid of the 12th man on your staff?"

    Heh, funny that you ask. Suppose we try to split up relievers based on three categories; saves, holds and innings. Namely, #1 relievers would be the guys with the most saves. #2, #3 and #4 relievers would be the most pitchers with the most saves and holds. #5 to etc relievers would be the pitchers with the most innings. Once a reliever is put into a category, he can't be put into another one.

    The reason why I use this methodology is because #1 relievers are usually closers who get saves. #2, #3 and #4 guys are usually relievers who get holds. Typically, a team has a setup man who pitches the eighth and two guys that pitch the seventh. In a given year (from 2002-2013), there are usually about 3.8 relievers on a team that have a total of ten or more saves, holds. If you include blown saves in that total then it's at 4.5.

    After this, we can use innings pitched to determine which relievers are #5s, #6s, #7s, etc. Given that I'm using a dataset consisting of 12 years and there are 30 teams playing each year, that means that there should be roughly 360 relievers per slot. That means of the remaining relievers, the next 360 are #5s, the ones after that are #6s, the ones after that are #7s etc.

    If we split the relievers into these groups and look at the average, first quantile and third quantile of ERA and WAR then we see the following:

    ReplyDelete
  8. #5 relievers have an average ERA of 3.98 and an average WAR of .2. They range from a third quartile of 3.23 and .5 WAR to a first quartile of 4.58 and -.1 WAR.

    #6 relievers have an average ERA of 4.28 and an average WAR of .09. They range from a third quartile of 3.41 and .4 WAR to a first quartile of 5.08 and -.2 WAR.

    #7 relievers have an average ERA of 4.47 and an average WAR of .01. They range from a third quartile of 3.29 and .25 WAR to a first quartile of 5.38 and -.2 WAR.

    #8 relievers have an average ERA of 4.53 and an average WAR of 0. They range from a third quartile of 3.33 and .2 WAR to a first quartile of 5.53 and -.2 WAR.

    #9 relievers have an average ERA of 4.79 and an average WAR of 0. They range from a third quartile of 3.29 and .2 WAR to a first quartile of 6.15 and -.2 WAR.

    #10 relievers have an average ERA of 5.28 and an average WAR of -.05. They range from a third quartile of 3.70 and .1 WAR to a first quartile of 6.75 and -.2 WAR.

    Given that the average #5 pitcher in the bullpen gets fewer then 10 holds and saves combined we can conclude that he's not pitching in crucial situations. Given that he averages about 66 innings a game, that would be that using an average #9 reliever instead of an average #5 reliever would cost a team roughly .8 runs per 9 innings or about 5.9 runs a season. Since the runs aren't given up in crucial points, the impact is probably less then that. The average #6 pitcher in the bullpen only throws 48 innings on average and averages half a run fewer per 9 innings. That would mean the difference between him and the #9 pitcher would be about 2.6 runs a season.

    We see that provided that a team doesn't have a good #6 reliever, there's little difference in performance between him and an average #9 reliever. This means that the only benefits that the average #6 and #7 relievers have is that they throw low leverage innings so that the top guys don't have to do so. Likewise, if a team also has a below average #5 reliever then his only benefit is throwing low leverage innings.

    If a team wants to have a six man bullpen then all you'd need to do is make sure that your low leverage relievers didn't get worn out. You'd do that by making sure that your #6 and possibly #5 guy has options and then simply sending them down to the minors frequently to give them rests while calling up reserves.

    This would probably kill your AAA team and would mean that you'd need to consider keeping a bunch of marginal arms on the forty man roster. The major difficulty would be managing the forty man roster.

    But having an extra player on the bench does give you the ability to have another platoon. If that platoon player was worth an extra win then it would be worth it.

    ReplyDelete
  9. Two other points - (1) Instead of shuttling pitchers to the minors, you can always stash them on the disabled list. As far as I know, no one is going to care if the seventh guy in the bullpen is on the DL for a couple of weeks with a minor "injury." And second, we also don't know how many more innings the higher-leverage relief pitchers can pitch without damage.

    ReplyDelete
  10. You definitely wouldn't want to make your high leverage relievers throw more innings. I don't think that's inevitable for a six man bullpen.

    The top four guys are already pitching a large majority of the high leverage innings. The top five guys are pitching nearly all of them.

    As long as you can leverage the #5 and #6 guy to pitch the same percentage of low-impact innings as the #5, #6 and #7 guy do right now then the numbers of innings that the #1-#4 guys pitch shouldn't change. And if you swap out the #5 and #6 guy regularly then that means you can use them often over a short span and then switch them out for another guy or two.

    It should be possible.

    ReplyDelete
  11. Regardless of the merits of going to a 6 man pen, Showalter is pretty old school and frequently preaches about the importance of protecting the bullpen, so its probably not going to happen. Improving the bench is certainly a major priority, though, and I think it will happen.

    If Betemit had played and been productive again vs RH pitching we might not even be having this discussion, so I'm sure they'll look at bringing him back for 2014. There's generally some flexibility between corner outfield spots and the DH position, so bringing in a high-obp corner outfielder with subpar defense could also make sense. Maybe Carlos Beltran if the price isn't too high?

    ReplyDelete
  12. this just came up on a BP effectively wild podcast a couple of days ago and thought it related well to this discussion of an 11 man bullpen. the thought that starting pitchers with a scheduled bullpen session should pitch those pitches in a game, allowing the team to get more value from them over the course of the year. after discussing some pros and cons, they didn't come to a conclusion as to whether or not it would be a good idea, but i thought it was an interesting idea.

    ReplyDelete