Platoons and Bullpens

Gabe Gross, Ryan Garko, Reed Johnson, and Eric Byrnes have all signed contracts in the past week or so, each settling for under $1 million on a one year contract after shopping their skills around. Each of these guys have some value in part-time platoon roles, specializing in hitting pitchers who throw with a certain hand, but lack a necessary skill to play full time.

Because of the ever increasing size of bullpens, these types of hitting specialists have fallen out of favor. The meager salaries that these guys have had to accept highlights the lack of value that teams are now placing on platoons. With only 13 of 25 roster spots dedicated to position players on most teams, it is becoming far more difficult for teams to accommodate left-right platoons and still have the necessary reserves for their starters in case of injury.

Is that a wise use of roster allocation? I’m honestly not sure. I know there’s been a backlash against the ever increasing bullpen sizes among the sabermetric community, but I haven’t seen much in the way of evidence that specializing your bench is more efficient than specializing your bullpen.

Yes, the 12th pitcher on any given team is usually not very good, often producing at a near replacement level. Even if they pitch decently, the leverage of the innings they are given is usually so low that their overall value is quite low. So a straight comparison between value of platoon guy versus value of the 12th reliever will naturally lead one to conclude that teams would be better off with a larger bench and smaller bullpen.

I think there’s more to it than that, however. While mop-up relievers may not pitch well or handle many important innings, having them around allows managers to use their better relievers in different ways. As we’ve seen in the last decade, the larger bullpens allow managers to mix and match based on handedness in higher leverage situations, using pitchers who are far more effective against same handed hitters.

The goal of platooning a pair of hitters or using bullpen specialists is really the same – get as much value from exploiting left/right splits as possible. So why are managers going more towards pitchers when attempting to exploit those advantages?

I think it’s because they inherently understand the pinch-hitter penalty. As Matthew noted on ESPN a few weeks ago, guys perform at a level nearly 10 percent below their true talent level when being used as a pinch hitter. There is significant evidence that the act of sitting around for a few hours, grabbing a bat, and trying to get a hit is just very hard.

If you try to exploit platoon advantages from the offensive side, you get less than the full value of the hitters you are using when pinch-hitting. Thus, any hitter with a large platoon split can be devalued in high leverage situations in a way that is tough to counter. Even if you pinch hit for the left-handed half of your platoon when the opposing manager brings in a LOOGY, you’re still at a disadvantage, because your right-handed bat now has to overcome the penalty of inactivity.

While it may be annoying to watch the parade of relievers bog down game from the 7th inning on, I think there’s a pretty decent chance that managers are making the more efficient decision. It’s an area that needs more study, certainly, but I think we should acknowledge that there is value to platooning your relievers that is not found in platooning your hitters.




Print This Post



Dave is a co-founder of USSMariner.com and contributes to the Wall Street Journal.


37 Responses to “Platoons and Bullpens”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Felonius_Monk says:

    “Yes, the 12th reliever on any given team is usually not very good”

    Probably what you’d expect, given that they’re probably only the 5th best relievers on the team’s triple-A affiliate…

    Vote -1 Vote +1

    • Steve says:

      Not sure what this means.

      Vote -1 Vote +1

      • arch support says:

        Dave’s line should read “12th pitcher on any given team” rather than 12th reliever.

        Having twelve relievers means you’ve already got the standard seven, plus five more … probably from your AAA team. Thus 12th reliever=5th AAA reliever.

        Vote -1 Vote +1

  2. Grant says:

    Aside from leveraging splits it would seem that the other value of the 12th reliever is simply reducing the workload of every other pitcher on the staff. Even if he’s barely above replacement level that’s fine for the last two innings of a 9-1 game. At least it reduces the stress on the rest of the bullpen and by extension even the starters. I’m not sure how much it has been studied, but it appears clear that pitchers pitch more effectively when they’re reasonably well rested.

    Vote -1 Vote +1

    • neuter_your_dogma says:

      Agreed. Even a 7.5FIPer can give you an inning or 2 in a 10-1 ballgame for either club, thus saving wear on the remainder of the pen.

      Vote -1 Vote +1

      • Paul B says:

        That assumes that the team would benefit from reducing wear on the pen.

        On teams I follow, it generally seems that there is more than one reliever who is getting almost no work. Those mop up innings could go to the 6th reliever instead of the 7th. Since the 6th reliever isn’t pitching too many times a week anyway,.

        Vote -1 Vote +1

  3. Kevin S. says:

    With a seven-man bullpen, what do you have? A closer and two set-up men (your relief aces), a long guy, a LOOGY or two and a mop-up guy. If you’re in the seventh inning or later, you’re already into the part of the game where the aces should be seeing action, so if you do use a LOOGY, you pair him with one of your setup guys. Given that there’s a lack of differentiation between back-end relievers, they generally should be minimum-salary guys, often with options still available. The bullpen can be expanded by dumping the “mop-up” man and rotating guys through the sixth bullpen slot and AAA.

    Vote -1 Vote +1

  4. yossarian says:

    Not that I think this would make a difference, but is it against MLB rules for a pinch hitter to go “warm up” in a batting cage before he’s called into the game? Has anyone ever done this?

    Vote -1 Vote +1

    • Choo says:

      During the game, players go under the stands to watch film, take BP, eat a banana, lift weights, ride a stationary bike, throw in a dip, take a dump, etc. Some DH’s exercise or play catch when their defense is on the field and most (if not all) pinch hitters will stretch and take at least dozen hacks in the cage or off a tee prior to their AB. Matt Stairs is sweaty when he comes to the plate for a reason. Actually, that’s probably a bad example. I’m sure there are many reasons to explain why Matt Stairs is sweaty.

      Vote -1 Vote +1

    • Scooter says:

      I’m quite certain that they can. I mean, they don’t have to stay in the dugout, or even in the stadium, so what rule could there be?

      You know, reading the article put me in mind of a story about an Oriole — might have been Lowenstein — who came to know when Earl Weaver was likely to call on him to pinch-hit, so he would start stretching and warming up maybe an inning before. He seemed to think it made him more ready. Whether it showed up in his stats, I haven’t the foggiest. (Lowenstein’s splits sure look like the answer is no, but then, I’m not even sure he was the guy I’m thinking of.)

      Vote -1 Vote +1

    • Mark S says:

      I toured Safeco Field (Seattle Mariners) this last weekend. If you walk from the dugout into the hall that leads to the clubhouse on your right there are bathrooms and on the left is a full batting cage and then a video room. I was told there is every possible angle of the game being recorded. Guys come in from an PA and can see what they did in the last PA 30 seconds after it has happened.

      I have no idea if any of it helps.

      Vote -1 Vote +1

    • Attractive Nuisance says:

      This warm-up issue is discussed a bit further here: http://www.lookoutlanding.com/2010/2/2/1289125/now-warming-up-in-the-matador-pen

      Vote -1 Vote +1

  5. BrettFan1 says:

    I wonder if the pinch-hitter penalty has diminished in recent years. The newer ballparks have hitting facilities that are more accessible to the players during the game as well as video that can be used to facilitate keeping the players on the bench “in the game”. It would be interesting to see if there is a timeline change or a ballpark affect to the pinch-hitter penalty.

    Vote -1 Vote +1

    • Doug Lampert says:

      I should think the way to check that would be to check for park dependance of the pinch hitter penalty, and then see if the parks with the smallest penalty are those with the best fitness and video facilities.

      Vote -1 Vote +1

      • DL80 says:

        I’d also be interested in seeing if there is any correlation between starting players’ first at bat and a pinch hit at bat. Is there a 10% penalty on every first at bat of a game, regardless of when it is? Does it apply only to the visitors’ team, whose players haven’t had a chance to really get in any game action?

        Vote -1 Vote +1

  6. geo says:

    Interesting, Dave, in that, at the rate Jack Zduriencik is collecting position players to platoon, your team the Mariners looks likely to employ an 11-man pitching staff.

    Vote -1 Vote +1

  7. James F says:

    Is the 10% drop in performance for pinch hitters balanced against the quality of competition?

    To clarify what I mean, here is Mr. Cameron in January 2009:

    “Running the numbers through the formula gives us a 4.68 FIP (traditional, not scaled to RA) for an AL reliever and 5.63 for an AL starter. So, if you’re looking at a pitcher’s FIP here on his FanGraphs page, and that pitcher happens to be in the American League, those are the numbers you’d want to compare him to in order to see how far away from replacement level he is.

    For the NL in 2008, the numbers are 4.45 for a reliever and 5.37 for a starter – the lack of a DH drives down the league’s offensive level, and so the performance of a replacement level pitcher will appear better in the NL than in the AL. ”

    Now, almost by definition, nearly all of the at bats by pinch hitters will be against relief pitchers rather than starters, and by your own calculation, relievers in both leagues have a FIP nearly a full run better than starters. Meaning that relievers are more than 10% more effective than starters on average. So my instinct is to say that the drop off in performance is due to pinch hitters facing better pitching on average than starters. In addition, pinch hitters are used primarily in close games rather than in blowouts against the worst pitchers in the bullpen, which would suggest that they are facing more quality relievers than average.

    Vote -1 Vote +1

    • Luke in MN says:

      I raised a similar issue on another site recently and was told the work on pinch-hitter disadvantage accounts for this. But I’d really like to hear the author address it.

      Vote -1 Vote +1

  8. Matt S says:

    I am inclined to disagree with Dave on this point, though I do not have any hard evidence to point to. My feeling though is that intelligent imployment of a platoon to cover a position instead of a single player is often smarter use of a roster spot then the 12th or even 11th reliever spots. The most recent example I can recall was last season’s Brewer’s gabe platoon in RF (Gross/Kapler) as was highlighted in a fangraphs article last season. THT did a few articles on the genius of Casey Stegel’s platooning. The pinch hitter problem is not so drastic in this case because the players are split based on the starters handedness, not as pinch hitters only. The production of two well platooned players might easily beat the production of one starter plus an almost useless reliever if each player in the platoon is equal defensively and the handness strictly observed. The added depth to the bench is a plus as well as a platoon player may be a better player than a strict backup.

    Consider which would add more value to the current Red Sox lineup, trading Mike Lowell for a replacement level player or using him in a strict DH platoon with David Ortiz (locker room problems aside for the sake of the example). It seems that such roster construction is more likely to hold an advantage over the use of an addition low leverage reliever.

    Vote -1 Vote +1

  9. Tim says:

    Sure, platoon hitters may not have as much value in pinch hitting opportunities, but this misses the point.

    Rosters should carry more platoon players and use their split advantages in the beginning of games where they will get more than one at-bat. I think there is an enormous inefficiency to be exploited by platooning players on daily basis depending on the starter. You might pay $8M for an outfielder w/ a 850 OPS when you can get that same production by paying $1.5M to two different players who have platoon splits of 850 against either left-handed or right-handed pitchers.

    Sure, you lose a roster spot, but you gain $5M. That money more than compensates for the lost roster spot that you would otherwise be giving to a replacement level reliever.

    Vote -1 Vote +1

  10. Mike Green says:

    There are a number of factors which govern typical current managerial use of the pen, and a couple of them have nothing to do with efficiency:

    -the save rule, and usage patterns built around giving the “closer” low leverage save opportunities and not giving him high leverage non-save opportunities.
    -the win rule for starting pitchers, and the incentive to allow starters the opportunity to win the game.

    This leads to shorter relief appearances, regardless of the platoon matching aspect. I am quite confident that if the save rule were dispensed with, we would see longer relief appearances and (some) 6 man pens. In the meanwhile, it will take an innovative management team to do it.

    Vote -1 Vote +1

  11. Joe says:

    Dave,
    Patrick Newman mentioned that Japanese baseball teams tend to run out 6 man rotations. Is something you could see in the MLB? Also, what, in your opinion, is the ideal allocation of players?

    Vote -1 Vote +1

    • philosofool says:

      I suppose some team could do this, but it seems like a horrible idea to me. It takes a 200 inning ace pitcher to a 166 inning ace pitcher. Those 34 innings from your ace are worth about one win over the season, and in place of those you end up starting a guy who probably should be in AAA.

      Vote -1 Vote +1

  12. Choo says:

    Beauty is in the IQ of the beholder. For some, a platoon split is worth approximately Bill Hall and $1.5 million in cash. For others, it could be as valuable as, oh, I don’t know, Shin-Soo Choo and Asdrubal Cabrera.

    Vote -1 Vote +1

  13. aj says:

    I agree with Dave that a 12th pitcher can be more valuable than another pinch hitter, and I’ll throw in another reason.

    Having a 12th RP means that a manager can give his good RPs what he deems “optimal usage.” That means not having to use them in mop-up situations, early in games, or stretching them out in extra innings. Of course managers have different ideas of what “optimal usage” might be, as do all of us on the sidelines. That could mean not overusing your good RPs, or it could mean keeping them sharp with a regular schedule, or saving them for high-leverage situation. Who knows. But whether or not he manager does use his RPs optimally is not the point. The point is that the 12 pitcher is a buffer that protects his big bullets for effective use. Add that to Dave’s point that RPs can be more effective than pinch hitters in those late game matchups, and I’d say there’s pretty good reason to leave your starting position player in the lineup and carry the extra arm instead.

    Vote -1 Vote +1

  14. kokushishin says:

    It seems more likely that the 12th pitcher who is effective probably should be higher up on the totem pole.

    A deeper bench could also make the pitching staff’s jobs easier via that slight increase in offense and/or defense..

    Vote -1 Vote +1

  15. Mike H says:

    Teams will routinely use 3 or 4 good relievers for an inning each in a game that is for all intensive purposes out of reach. This is the wasteful usage that could be cut down.

    Vote -1 Vote +1

  16. Linuxit says:

    It’s not even a matter of roster size. During roster expansion in September, managers still don’t play strict platoons much. I think it’s just a matter of talent available. It’s easier to find an effective RP from the minors, than it is to find a steady part-time player who can both hit and play good defense.

    Vote -1 Vote +1

    • Jacob Jackson says:

      “A steady part-time player who can both hit and play good defense.”

      AND…who won’t bitch excessively about being a part-time player. I think a small, small additonal factor in the shift to 7-man bullpens is that a hypothetical “5th/6th bench guy” really wouldn’t play that much, and that most players do not prefer to be a part of a platoon. Many who are put in that role complain about it. Everyone who is physically capable would prefer to play everyday rather than be a platoon guy, even if someone pointed out their bad splits to them.

      That issue of being underused and the locker room problems it could create doesn’t exist as much with relievers. They are used to spending a lot of their baseball life just watching the action. Plus, it’s easier to use a 7th reliever to his full physical capability – to make him feel “useful” – you’ve only to pitch him a few innings a week to do that.

      Vote -1 Vote +1

  17. Jacob Jackson says:

    Other questions this raises for me:

    *If you had a terrific pinch-runner – a guy who could steal practically at will with an 85% success rate, and he did nothing else for your team…how much would he be worth over the course of a season? One win? 1.5? Enough to justify carrying one all year?

    *How many at-bats does a player need, on average, per week in order to remain alert and effective? This is surely a consideration for the teams that opt to carry four-man benches in the AL. Pinch-hitting opportunities are almost nil; if your 5th bench guy was only going to get a start every two weeks, he’d probably be terrible in the role anyway, because he’s so rusty when he finally plays.

    *When we’re talking about only four precious bench spots for AL teams, it strikes me that “backup catcher” ends up being a roster slot that produces very little value. Thinking about alternatives: When was the last time a team carried a backup catcher with true versatility? Like, if Jake Fox were to be the A’s backup catcher for 2010, or if Pablo Sandoval was the Giants backup catcher entering last season, Brandon Inge, etc. A guy who hits or fields well enough to justify playing him signficantly at another position also moonlights as your 30-games-a-year, once-a-week backup catcher. Does this ever happen anymore? If not, is it because catcher defense is too important and difficult to entrust it in the hands of someone who is also putting in a lot of work at 3b/1b? If teams could get away with not carrying a light-hitting backup exclusively devoted to playing catcher, it would free up an additional roster slot for platooning.

    Vote -1 Vote +1

    • Sandy Kazmir says:

      We’ve been talking about how catcher might be the perfect spot for a natural platoon. You’re going to have two no matter what, might as well find a guy that mashes lefties to pair up with a guy that destroys righthanders. Supply isn’t exactly deep, but if you could have Gregg Zaun paired up with Kelly Shoppach you’ll have two catchers that complement each other perfectly.

      Vote -1 Vote +1

  18. aj says:

    Just like Matthew Carruth’s 10% pinch-hitting toll has on bench players, it would be interesting to also know what effect the presence of a 12th pitcher has on the other 11 pitchers. Does the extra rest help or not? Does not having to pitch in the low-leverage situations keep them sharp or not? Unfortunately I doubt there’s enough data draw any meaningful conclusions. And what we have would be skewed by managers’ varying usage of the 12th man.

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>