The 13th Man Out

Last time I railed a bit on the act of carrying three dedicated catchers on a roster. Tonight, with real substantial baseball news still coming in at just a trickle as we lead up to the start of the regular season, I decided to continue the trend. This time, it’s not such a cut and dry case as I think the three catchers issue is. Tonight’s issue is the 12-man pitching staff.

I believe there are a number of circumstances where it is prudent to carry 12 pitchers. My beef is that seemingly every team does it all season long and that just screams wasted roster spot to me. One only needs to look through usage patterns to see the evidence; relievers going six or longer days between appearances, that is a surefire clue that there could be better uses for that roster slot.

Namely, it boils down to this; I think the game has evolved lately into a situation where managers are paranoid about treating each individual game as if it were totally independent. But it’s not, not on a usage level. What you do in one game does affect what your options are in the following. How many pitches does a reliever throw each time he gets warmed up? 20, 30? How many times does he make an appearance which involves less than that many pitches? My hunch would be a majority.

Why not just carry 11 pitchers and if you are facing a game after a particular heavy usage day, just swap one or two out with the best available Triple-A relievers for a short while. They’re RP, and not dominant ones even, they’re almost the baseball definition of fungible assets.

Managers seem too cavalier to me in making pitching changes to try and extract the greatest possible match ups at that particular moment. Of course, that sounds good, right? And it is good, if usage was independent from game to game, but it’s not. And endless pitching changes to eek out an extra percent or two of favorability on a single pitcher-batter battle is focusing on the micro level and missing that if you relaxed your standards in that regard, perhaps you would be able to go with an 11-man staff and thus carry another bench bat which can provide you with a greater total contribution.

This specific topic, and roster maximization in general, makes for a fantastic area of research that in time I plan to get around to analyzing with hard numbers, but for now I just wanted to present something that sticks out to me every March as Opening Day rosters get set. What are your thoughts?

Print This Post

Matthew Carruth is a software engineer who has been fascinated with baseball statistics since age five. When not dissecting baseball, he is watching hockey or playing soccer.

20 Responses to “The 13th Man Out”

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

    Of course, the question that goes hand-in-hand with this is how often teams need an extra bat off the bench – keeping in mind that they’re already probably keeping at least their top two or three options on the major league squad, so whoever you bring up is probably at best going to be your third best pinch hitter. In addition, you’re limited in terms of who you’re going to call up – you don’t want to pull up young prospects who you’re trying to get regular playing time in AA or AAA just for the occasional pinch hitting duty; instead, you’re looking at the more veteran AAA players – ones you don’t mind jerking back and forth between the big league club and AAA. You’re also limited in terms of who you can send down – it has to be someone with options left, which means this probably only works if one of the non-crucial elements of your bullpen (one of your worst 3 pitchers, say) is young and has options. That may not eliminate many teams, but it will certainly affect some.

    The last element to consider is the psyche of those involved. I tend to think that stuff like defining roles for players is overrated, but some players simply might not deal with a situation like this as well as others. If you’re shuffling the same players too regularly, there may even be a concern of tiring them out from frequent travel back and forth between teams in addition to travel with the major or minor league club. The baseball season is a tiring experience – that’s why greenies were so prevalent – and this is just adding to that.

    Overall, I think it’s an interesting idea, but it’s something that, practically speaking, is limited to a small number of teams with the appropriate personnel situation. Along the same lines, I hope you’ll look at the roster issue from the position player side as well – I’d be interested to see how frequent it is for bench players to go 6 or 7 games without being used.

    Vote -1 Vote +1

  2. rwperu34 says:

    In my Dice League, I have to have several teams carry 12 pitchers. The biggest single factor is quality of starting pitching. The second biggest is probably park.

    I’d much prefer to carry 11 pitchers and have an extra platoon or quality bench player, but when I’ve got a crappy rotation in a hitters park, sometimes it’s unavoidable. I do have one team that’s able to go with a ten man staff, especially if the off days line up. They’ve got the best starters, quality bullpen depth, and play in the old Royals Stadium with turf that depressed offense significantly. Ironically, this team carries three catchers:)

    Vote -1 Vote +1

  3. Nathan says:

    I hope your only talking about NL teams. Bats come off the bench so infrequently in the AL that I would think the extra pitcher would be the better call.

    Vote -1 Vote +1

    • Basil Ganglia says:

      That means that for an AL squad the tradeoff will skew more towards carrying a 12th pitcher vs. creating an additional platoon in your starting team. In the NL, the analysis will skew a bit more towards the value of having an additional bat on the bench for late inning pinch hitting.

      Vote -1 Vote +1

    • Marlowe says:

      But even in the AL there is a chance that carrying an extra bench player to use in a platoon might be more valuable than an extra arm in the pen.

      Vote -1 Vote +1

    • Kincaid says:

      It is true that AL teams pinch hit far less frequently than NL teams. It’s something like 3.4-3.6 pinch hitters per game in NL stadia and 0.9-1.1 per game in AL stadia if I’m totaling the data right, so the difference is substantial. There are still more factors to consider, however. AL teams also have an extra starter in the DH, so they have one fewer bench position to begin with. Adding the extra pitcher is like taking two bats off the bench compared to an NL team. There is also Marlowe’s point that bench players can be used in other ways, like in platoon situations. Is the ability to exploit more platoon advantages in your starting lineup or off the bench at the plate worth more or less than the extra arm? How about the use of pinch runners or defensive replacements? Pitchers can generally go slightly longer in the AL than in the NL since they aren’t removed for pinch hitters as well, so the need for an extra arm may not be as strong.

      I would definitely agree that each league is different enough here that each would need to be approached independently, and that there are factors that could shift one or the other in a different direction, but I do think it would still take careful consideration and a look at the hard numbers to be able to say definitively either way.

      Vote -1 Vote +1

      • Eric says:

        As far as how long the pitcher can generally go-

        Turns out it is VERY close [2000-2004 data], DH vs no DH.

        Is that really true?

        Starters in DH games went 66915.7 innings in 11330. Average =5.91
        Starters in non-DH games went 76426.7 innings in 12954 games. Average =5.90

        The gap of .00621 IP/GS amounts to one total extra inning per 161 games. I think it is a safe to call it about even.

        While not having to hit, they also don’t get to face the opposing pitcher [or the generally below average PHers], they face generally above average DHs.

        Vote -1 Vote +1

      • Kincaid says:

        Starting pitchers are not the only pitchers removed for pinch hitters, though. Relievers almost always get pulled when their spot in the lineup is due up, regardless of whether the manager would prefer they keep pitching. Over the past 5 years, the average NL team has made 3.0 pitching changes per game. The average AL team has made 2.7. The difference of about .26 pitching changes per game works out to about 43 extra relief appearances per season per team in the NL. That’s easily a full season’s worth of appearances for a 12th pitcher, so it’s a significant difference. Essentially, teams in the AL are getting away with using the equivalent of one low-end reliever fewer on average than NL teams.

        Vote -1 Vote +1

  4. brent says:

    I think the problem is usually that you want a pinch hitter for your SS or 2B. That hitter usually can’t play that position, so you have to use another player to substitute for the pinch hitter to get 1 AB. Teams just try to stick a player a SS or 2B that can hit and play passable defense.
    Yes, it is a waste, but I think the real solution is just making a 24 man roster to get rid of the extra guy in the bullpen. Getting rid of one extra guy down there would probably make the manager consider more carefully removing pitchers too quickly.

    Vote -1 Vote +1

  5. Kevin S. says:

    I agree, and this is a point I often make (specifically for the Yankees, but can be generalized to any team). Your team carries five starting pitchers, plus one swing man. That’s six. Throw in your closer and two other pitchers you trust in leveraged situations, and that brings you to nine. Those last two spots are your MR guys, suited for mop-up duty and other non-leveraged situations. You aren’t spending any more than the minimum on them anyway, so they’re most likely guys with only a year or two of experience or pitchers on minor league contracts anyway. They aren’t almost the definition of fungible, they are fungible. Even if you want a total of four leveraged guys, you still have one spot that isn’t on lock down. If that guy is getting overworked, swap him out for a few days, no harm done. In the AL, your normal bench is probably backup catcher, backup corner infielder, backup middle infielder, fourth outfielder. Having that fifth spot to house a platoon, or hell, even a third catcher, if you’re into that, is more valuable than burning it on an utterly replaceable reliever.

    Vote -1 Vote +1

  6. brent says:
    The score is 10-2, and the Korean manager pulls his pitcher after getting one out in the ninth (no one on base either). This is an instance where he should have switched the pitcher between innings or just left him in to get the last two outs. This is a totally unnecessary delay of the game. When they make rule changes, I think they should try to keep in mind that baseball should usually take about two hours. There is no need for games to be dragged on and on into three, three and a half or four hours.

    Vote -1 Vote +1

    • Fresh Hops says:

      It’s all well and good to tell a manager that they should try to keep the game fun to watch, but the manager is never going to face the heat for taking too long. They will face the heat for losing.

      The big issue, of course, is that this is wasted usage. If you leave that pitcher in for three outs, that’s one fresh pitcher for your next game that you don’t have because you pitched two guys in that inning.

      Vote -1 Vote +1

  7. JC says:

    While I like the idea of a 11 man pitching staff, the thought of swapping tired/heavily used pitchers with AAA guys may be impractical for some teams. Say the Mariners are out east playing in Boston or New York. That is one long flight from Tacoma, unless they just bring some extra guys along for the trip.

    Vote -1 Vote +1

  8. Paul Scott says:

    I think the three biggest areas of improvement for a ball club are, in order:

    1. Platooning (the failure to do it enough)
    2. Relief Pitcher Usage (the “closer” issue)
    3. Quality of rotation (4 man rotations improve the overall quality of innings pitched)

    The 12 pitcher slots going to 11 would allow more platooning, so it would be a good thing if the two could go hand-in-hand (and going to 2 catchers would improve that one further). I don’t see much of an advantage just to have a PH available. That, it seems to me, is working at the same goal and the question just becomes “would you rather have the handedness advantage to prevent scoring or to create scoring?”

    Outside of platooning more, the 25th slot being assigned to a pitcher or batter it seems to me should be controlled with an eye to protecting prospects. A look at the Dodgers, for example, sees a strong possibility of having to release Delwyn Young to satisfy Torre’s “need” for 12 pitchers. To me, the obvious solution to the Dodgers situation is to go with 11 pitchers and give the 25th spot to Young. Other years and other teams, that same analysis could tell you to give the 25th spot to a pitcher. Either way, being flexible on this point AND going to two catchers means being able to protect prospects for a longer time. That is more likely to lead to a real payoff than other considerations.

    Vote -1 Vote +1

  9. sen-baldacci says:

    There is a great argument for the extra bat in the NL. I would pinch hit for the pitcher almost every time possible, and you can upgrade your defense somewhere for the last few innings or have a pinch runner to help score the late game-tying run or go-ahead.

    But in the AL it makes no sense. Ok, there are some teams with guys that are so poor they can’t do enough right and you need to platoon with them, and pinch run for them and get their poor defense out, but most decent teams aren’t in that situation. If you keep the extra bat to get the lefty righty matchup, they can just change the pitcher to offset it, because they are keeping the extra arm in the bullpen. It comes down to one main thing for me. The extra batter can only be used when one of your poorest hitters is up in the right situation towards the end of a game. The extra pitcher can come in at any time in the last few innings to give yourself a better matchup at an important point in the game. The flexibility of using the pitcher to maximize his usefulness outweighs any other skill sets a single end of the bench player can bring (defense). There are obvious exceptions and those are determined by the players you have available. If you have blazing speed sitting on your AAA team, I’d consider teaching him to bunt and bringing him to the big squad. Speed on the bases can always be substituted in key situations….Dave Roberts anyone?

    Vote -1 Vote +1

  10. Brett says:

    I’m not a big number cruncher (though I love to sit back and watch others do the dirty work), but I’d love to see a breakdown of average non-closer relief pitcher win values and the average breakdown of win values for non-starting bench players. Corrected forwhatever additional noise might be appropriate, of course.

    Vote -1 Vote +1

  11. CaR says:

    I believe that the revolving door method suggested in order to maximize position player availability would not work as a practical matter. First, I think that the general psyche of the players involved would have to be considered. The reality of working in the show as compared with AAA is an important factor. Important enough, I feel, so as not to be brushed aside as statistically irrelevant . Second, and related, is the notion that the relative production of AAA fodder is as good as MLB talent to justify the move to begin with. Bullpen arms are the easiest resource to reproduce, I understand that. But if one is willing to take the step down in production that would be required in order to justify the strategy, I suppose that there had better be a tangible (not theoretical) difference shown by the bench player who would take the 12th pitchers’ spot.

    Vote -1 Vote +1

  12. RedSox2004 says:

    Another aspect to consider is the broader roster strategy of the club. If you believe that the 25th spot is largely fungible whether it goes to a pitcher or a bench player, then one should also consider whether there are players out of options, Rule 5 draftees, etc. If there’s a player like that who the team likes but must put on the 25 man roster or risk losing from the organization, then it makes sense to stick that player in the 25th spot, regardless of whether it’s a pitcher or bench player.

    Vote -1 Vote +1