No, you do not need to get him some work

We’ve all heard managers say it before. When a good relief pitcher is brought into a blowout game, whether his team is winning or losing, it’s almost inevitable that the team’s manager will explain that the reason this good pitcher appeared was so that he could “get some work,” and the general feeling is that it is to keep the pitcher from getting rusty. Take Friday’s 11-3 win over the Detroit Tigers, where Mark Lowe threw one inning, allowing 2 hits and 1 run while recording a strikeout. He threw 18 pitches. Kirby Arnold of the Everett Herald noted on his blog that “getting some work” was Wakamatsu’s motivation for putting Lowe in the game.

Wakamatsu said Lowe is fine physically, noting that his fastball reached 97 mph tonight, but that he was rusty after having not pitched since April 10.

Well, honestly, I think that’s a load of crap, and I set out to prove it.

Generally speaking, managers will only use this excuse with good pitchers. No manager puts, say, Sean White in the game to get him some work. When a manager uses this excuse it’s generally for setup men and closers who haven’t pitched in awhile. The managers must believe, then, that a relief pitcher’s performance will suffer if he does not pitch regularly. Perhaps managers believe the layoff may make them lose the strike zone, or become more hittable. Is this true? Or are managers simply wasting an appearance from one of their better relievers in a game that doesn’t matter?

I looked at Mark Lowe and five closers’ 2009 appearances after which they hadn’t appeared in a game for at least three days. What I found was unsurprising. Though small sample size applies, these six pitchers generally performed better on longer rest than their overall season line. So it would appear that having extra days between appearances improves pitcher performance instead of diminishing it. Seems logical enough to me. Anyway, enough words, it’s time for numbers!! Hooray!

The first set of numbers: Stats accumulated when pitching on 3+ days of rest.
Numbers in parenthesis: Overall 2009 stats.
P/IP=Pitches per inning.

Too see all of figure 1 just click on it. Our blog is too narrow to see the whole thing.

Figure 1

Figure 2

As you can see from the tables above, it would appear that when pitchers are more rested, their stuff becomes harder to hit, but also slightly more difficult to control. However, their walk rate goes up less than one walk per nine, while their strikeout rate goes up by more than 2. Obviously, 90 innings is an extremely small sample size, but being that each reliever’s performance changes similarly, I’m willing to say that this is probably significant.

So, next time your manager says “he just needed to get some work in” facepalm away, because your manager is wasting one of your better pitchers’ appearances in a meaningless game. If the M’s need Mark Lowe this afternoon and can’t use him since he pitched on Friday, manager Don Wakamatsu will have essentially removed one of his best late inning weapons for a game for no reason. Take note, MLB managers. You don’t need to get your pitchers extra work to keep them from getting rusty, because they don’t get rusty. They get better.

Next Mariners Game View full schedule »
Wednesday, Aug 2020 Aug1:05at Philadelphia PhilliesBuy Tickets

Tags: Brian Wilson David Aardsma Jonathan Papelbon Mariano Rivera Mark Lowe Rafael Soriano

  • Jim

    That’s a very interesting observation. If only Wak would read this…

  • Brett Miller

    Thanks!! I don’t expect this to change how managers work, but I just wanted to know if the concept of “getting some work in” made any sense. Maybe some day managers will look at the numbers and they’ll stop throwing away appearances by good relievers.

  • Edrac

    I don’t think your evidence is enough to reach your conclusion. Lumping 3+ days into a single stat doesn’t say anything. 3-5 days rest many be ideal, however, 6 or more may be the point where relievers do indeed to need work.

  • Brett Miller

    Yeah, but there’s going to be an opportunity to pitch before they go 6 days without pitching. The only time any of these pitchers were rested for 6 or more days came over the all-star break, so since we don’t really have data of a pitcher going 6 or more days without throwing, we can’t do that study. We need pitchers to throw less often before we know that.

  • Gomez

    I suppose, with research of leverage indexes and such, maybe one could find some slight correlation to justify giving a reliever a rust-shaker outing in a blowout.

    But really, your point stands, Brett: A reliever’s performance doesn’t appreciably drop off with extended rest. The only thing everyone on either side can agree on is a reliever’s walks do go up a bit.

    Really, if Wak wanted to get Lowe some work, he’d make sure to use him regularly (every 2-3 days) regardless of the score, just like he should with all his top relievers. Then he wouldn’t have to worry in a blowout about anything other than letting the back of the bullpen mop up the final frames. Situations like this just come down to managers managing their bullpen better every day.

  • Daniel Carroll

    A couple thoughts:

    It strikes me that managers give “rust-shaking outings” when the game is basically in hand, so the comparative leverage index of these appearances is lower, which could make a potential difference. On the other hand, I think that would imply that relievers aren’t good in high leverage or clutch situations, and I’ve still yet to see a study that says that clutch performances are sustainable.

    The other thing is that I think about my experience with bowling. It is entirely possible for a good bowler to go a couple of months without throwing a ball and instantly return to a high level of ability one s/he picks up the equipment. However, if a bowler is trying to “work on something,” the only way to get the muscle memory to start remembering is practice. I’m sure relief pitchers get a lot of throwing in, but sometimes you need to put a competitor in a game situation to see if it’s working. My guess is that professionals at that level are always trying to fine-tune SOMETHING, so in that respect, maybe they DO need to see live hitting.

  • johnbai

    I always assumed the real deal was that managers want to go home. No one wants to sit through 4-5 hours of a blow out. So occasionally I think they throw out a well-rested top notch reliever in order to finish the game as quickly and painlessly as possible. They also want it to be someone that hasn’t pitched in a couple days so that they can always pitch again tomorrow if necessary.

  • Sean

    This analysis is so flawed I don’t know where to begin.

    As for Lowe, well he had a tight back. To test a guy coming off back stiffness in a blow-out game, throwing 98, is actually pretty valuable. A manager doesn’t want to take the chance in a close game, especially when Lowe is a guy used in close games.

    So in this case it was much more about testing a minor injury than concerns over rust from days off.

  • Griffin Cooper

    Sean, you start your comment by saying it’s so flawed, and then you go on to not explain at all.

  • Sean

    I just think it’s obvious:

    1. You only used 6 pitchers, all of which are above-average major league players. And over half are closers. And most throw pretty hard with reasonably good control. All established big-leaguers, all had better 2009 than Mark Lowe.

    2. If a manager feels the need to shake rust off a pitcher in a non-pressure situation, it’s because he knows something we do not, or the player has a injury concern, like Lowe did.

    3. Not every appearance after three days is a “rust-shaking” try by the manager. Not pitching for three days is not uncommon, even for closers.

    4. Since we’re using such tiny sample sizes, the game situations come into play.

    And there are more reasons. This is why I didn’t even want to begin.

    I don’t disagree that the average reliever is the same pitcher coming off 3+ days, but you picked the perfect group of guys to prove your point, over a small sample size, so that is why I said this analysis is flawed. Even though what your trying to prove may be true, your methods were very deliberate.

    Finally, your initial complaint was about Wak using Lowe during a blowout to check-in on his condition. I don’t see how your analysis ties into your criticism, because Mark Lowe was slightly injured during his time off, and your criteria for analysis has no consideration for either injuries or game score.

    Personally, I have no problem with Wak using that time to determine the condition of his pitcher. I was at the game against the A’s last Tuesday and Lowe was stretching more than anyone, starting in the 5th inning. He was hurting, and back recovery times can be unpredictable.

    Now, I UNDERSTAND why you used your methods, because a definitive analysis would be exhaustive, and who has time for that? But I disagree entirely that those numbers are useful for making a conclusion about relief pitcher’s performance on three or more days rest.

  • Pingback: Baseball Side of FanSided (4/24) | Call to the Pen