FanGraphs Baseball


RSS feed for comments on this post.

  1. It’s not really that closers are avoided in extra inning road games so much as managers, wrongly, wait for the save situation to show up. As maddening as Manuel is, he’s just making the mistakes that almost all of baseball makes. How many skippers get it with bullpen usage? Very, very few.

    Comment by lester bangs — May 15, 2012 @ 5:37 pm

  2. I think most managers now use their closer once he warms up. So if his offense pushes across enough runs in the last half-inning to void the save, in he goes anyway since he’s already warmed up.

    So is that good practice or not? I suspect so, because I also suspect most pitchers hate warming up but then not pitching.

    Comment by Richie — May 15, 2012 @ 5:43 pm

  3. Bruce Bochy? his bullpen management was as big as any reason why the giants won the world series… other than of course brooks conrad

    Comment by Jack — May 15, 2012 @ 5:46 pm

  4. Papelbon closed a one run game on sunday afternoon against the Padres. Minor correction.

    Comment by Chomp — May 15, 2012 @ 6:02 pm

  5. I remember Cody Ross as the post season anti-hero

    Comment by Josh G — May 15, 2012 @ 6:03 pm

  6. Change ‘Monday’ to ‘Tuesday’ and you should be set.

    Comment by Paulie — May 15, 2012 @ 6:08 pm

  7. Just because few managers make correct decisions does not mean we should allow it as a matter of course. One day the strategy of not using closers on the road in a tie game will be looked on as archaic and wrong by baseball mainstream, too.

    The real indictment of Papelbon’s misusage is he was used on May 12 in the 9th inning while the Phillies were losing. Why? He had only been used once in the previous 11 days while these tie games on the road were coming. So when a true save situation comes up on May 13 and he warms in anticipation of the same situation on May 14 he becomes unavailable on May 15. If he had pitched only May 13 and May 14 would he be available May 15? I don’t know, but I would think the answer is yes – he clearly can pitch 3 days in a row.

    Comment by Josh G — May 15, 2012 @ 6:17 pm

  8. Once Paps warmed up on Monday, he was already not going to pitch today even if Charlie pulled him back. I think a full warm-up counts as an appearance. Also Bastardo would have been the guy in the 9th today except he also pitched in 2 straight. Hopefully Diekman continues his recent minor league dominance and becomes a reliable late-inning arm. Dude is filthy.

    Comment by nik — May 15, 2012 @ 8:00 pm

  9. He has also saved two 1-run leads this year. Opening Day and this past Sunday. Fact check, my friend.

    Comment by Kevin Wilson — May 15, 2012 @ 8:09 pm

  10. Living in Pittsburgh and being a Phillies’ fan, I got to see firsthand Charlie Manuel misuse the bullpen in two games opening weekend and once more when we went to see the Phillies in DC May 4th.

    April 7
    April 8
    May 4

    On April 7th, he allowed Michael Stutes to pitch to the 2-3-4 hitters in a tie game in the bottom of the ninth and also allowed Joe Blanton to stay in the game with one out and a man on third in the bottom of the tenth, where a strikeout would be the best outcome. Blanton had already allowed Rod Barajas to nearly hit a home run, missing literally by inches. LI range of 1.69 to 5.49 with no Papelbon. He also tried to use Laynce Nix to bunt Hunter Pence from first to second with no out in a tie game in the top of the ninth, which ended not well.

    On April 8th, there were a number of times Papelbon should have been brought in. First and second with both one and two outs in the bottom of the eighth in a one run game, facing PH-1-2 in the bottom of the ninth in a tie game, man on third with one out in the bottom of the ninth in a tie game. LI 2.20 to 5.34 with no Papelbon.

    On May 4th, while there were several opportunities to bring in Papelbon, most egregious was after Michael Schwimmer had walked Bryce Harper with two outs in the bottom of the 11th, with Schwimmer having already pitched 2.2 innings, with Jayson Werth and Chad Tracy (4-5 hitters) coming up. Schwimmer walked the bases loaded and gave up a single to Wilson Ramos to end the game, all with their clear-cut best reliever sitting in the bullpen. LI from 1.26 to 5.37 (1 out in the 8th until the walk-off) with no Papelbon.

    Charlie Manuel is clearly not interested in using Papelbon in anything but save situations, situations where he has warmed up for the save but the lead grew, or the top of the 9th in a tie game. I have been to five Phillies’ games this year and three of them have seen the Phillies lose in walk-off fashion with Papelbon nowhere to be seen. It’s not that the Phillies haven’t had tough situations in which to use Papelbon, they just choose not to use him. With this mentality, Charlie has already thrown away as many as three games and with the way the offense has been struggling, there will likely be a few more cases of this throughout the course of the season.

    Comment by elkabong — May 15, 2012 @ 9:56 pm

  11. Where is the guarantee Phillies would have won those games. Papelbon could have given up a winning run himself, or maybe he would simply delay the loss by an inning.

    Comment by nik — May 15, 2012 @ 10:08 pm

  12. Agreed that Papelbon could have given up the winning run in any of those, but would you rather have Michael Schwimmer or Jonathan Papelbon on 40 pitches pitching with the bases loaded and two outs? David Herndon or Papelbon pitching to Andrew McCutchen with a man on and two outs in the bottom of the 10th? While bringing in Papelbon does not guarantee victory, it gives you a significantly better opportunity. If you keep betting bad hands in poker, you will end up losing more than you win.

    Comment by elkabong — May 15, 2012 @ 10:18 pm

  13. Correction, Michael Schwimmer on 40 pitches or Jonathan Papelbon, not Michael Schwimmer or Jonathan Papelbon on 40 pitches.

    Comment by elkabong — May 15, 2012 @ 10:19 pm

  14. Yeah, the days are off on this. It was Monday he finished a 4-run lead, his third game in a row, and Tuesday he wasn’t available.

    Also, the article misses the couple of 8th inning leads the Phillies blew when Papelbon was held out. Apparently it’s not done to either have a greater than 3-out save or put out an 8th inning fire and let someone else get the three outs in the 9th (if you’re worried about the former). The 15-13 loss to Atlanta was the most egregious example.

    Comment by Richard — May 15, 2012 @ 11:20 pm

  15. I’m surprised to hear the Phillies haven’t had good opportunities to employ Papelbon at the end of games. Has it just been pure chance? My expectation would be that a team with great starting pitching and a badly depleted offense would play a lot of close games.

    Comment by Jon L. — May 16, 2012 @ 12:37 am

  16. That depends upon if you consider protecting an 8th inning lead to be a good opportunity to employ Papelbon at the end of games. The Phils have already had at least three situations of this sort and they decided not to use him.

    Comment by hk — May 16, 2012 @ 6:42 am

  17. Manuel said he’s not opposed to running Papelbon in the 8th but not this early on in the season. Protecting the investment.

    Comment by Nik — May 16, 2012 @ 7:26 am

  18. Bochy? Romo’s IP is a pretty good indicator of his lack of acumen

    Comment by gobstopper — May 16, 2012 @ 7:44 am

  19. I agree. We tend to focus on mistakes that managers make but when we compare any manager to his peers, who tend to make the same mistakes he made, there’s not much of a difference.

    I really don’t think that managerial decisions, when compared with other managers, have a tremendous impact on the number of wins a team has at the end of the season. They tend to all hurt their teams relatively equally.

    Comment by chuckb — May 16, 2012 @ 9:20 am

  20. Jack, shouldn’t we mention the context of how that high leverage is obtained? Bastardo, as the team’s top LH reliever, is going to come in during more mid-inning fires, which generally have more leverage than starting a clean inning, even with a one run lead. When people point to closers not having the top LI spot on a team, it’s largely because of this point.

    Comment by DD — May 16, 2012 @ 9:57 am

  21. But that’s exactly the point of this article. Teams should be using their “closer” (who is often their best reliever) in these situations. While Bastardo should be near or above Papelbon for highest LI on the team simply because he will be called upon to face left handed hitters in tough situations, Manuel should be using Papelbon in situations like May 4th (bases loaded, two outs, bottom 11th, right handed batter) or April 8th (man on third, one out, bottom 9th, right handed batter). He shouldn’t be saving Papelbon for when his team is leading by two in the next inning, because a lesser pitcher (Qualls, Contreras, et. al) should be able to handle that lower leverage situation. Charlie is letting an arbitrarily defined individual statistic take precedence over the single most important statistic in baseball. They aren’t going to let the Phillies into the playoffs if Papelbon has 50 saves but the team finishes fourth in the NL East.

    Comment by elkabong — May 16, 2012 @ 11:17 am

  22. This article should have been called “Exhuming the Dead Horse in Philly”, and this comment “Let me have a go at that!”

    Everyone knows managers don’t always use their best reliever in the highest leverage situations… now that it’s the year 2006, can we move on?

    Comment by The Real Neal — May 16, 2012 @ 11:41 am

  23. “He shouldn’t be saving Papelbon for when his team is leading by two in the next inning, because a lesser pitcher (Qualls, Contreras, et. al) should be able to handle that lower leverage situation. ”

    (completely ignoring the fact that the impetus for this article was Qualls blowing a 2-run lead).

    Comment by The Real Neal — May 16, 2012 @ 11:52 am

  24. “Manuel should be using Papelbon in situations like May 4th (bases loaded, two outs, bottom 11th, right handed batter) or April 8th (man on third, one out, bottom 9th, right handed batter).”

    Was Papelbon warmed up and ready to go both days? If not, how long does it take him to get loose? If it takes 5 minutes to get loose, should they have predicted 5 minutes ahead of time that there’d be a bases loaded 2 out situation with a right handed batter?

    Lookit, it’s easy to look back on a game and pick the critical at bats. It’s much more difficult to see these coming ahead of time, and have your top reliever loose and ready to go just in case. You don’t want Papelbon getting loose every inning and throwing 75 warmup pitches in the bullpen over the course of the last few innings, in the event that you might need him to put out a fire.

    If you want to argue that he should be used in the 9th of a tie game, that’s fine. To expect him to get called into a game in the middle of an inning on short notice – that’s hindsight at its worst.

    Comment by vivalajeter — May 16, 2012 @ 12:15 pm

  25. I’m not comfortable with the suggestion (implied by many here) that every single manager in baseball is clueless. Is there any evidence that teams would actually win more games using the FanGraphs managing-by-simple-algorithm approach? If not, then the FanGraphs approach is a hypothesis to be tested, not a tool to question the intelligence of professional baseball managers. You might predict that teams would do better with your method, but you don’t know it. We shouldn’t be so bold.

    Comment by Jason H — May 16, 2012 @ 12:15 pm

  26. @Real Neal – I’m not saying it’s going to work 100% of the time. As I stated before, it’s all about putting yourself in the best position to win. Ultimately, Papelbon can’t pitch every day, so the Phillies definitely need guys who can handle low leverage situations. However, maybe they should have thought about bringing in Qualls in the even lower leverage situation the day before, so Papelbon would have been available on Tuesday.

    @vivalajeter – I agree that you can’t constantly be getting Papelbon up and down all the time. But how many times have you seen a pitching coach go out to the mound to kill time to get a reliever warm? Each of the PIT-PHI games described started the inning with a leadoff double and a sacrifice bunt. You’re telling me that in that situation, where 99% of the time there’s going to be a sacrifice bunt, Papelbon can’t get warm? Schwimmer had over 40 pitches in the WAS-PHI game and proceeded to walk three batters, they couldn’t get Papelbon (or anyone, really) up over that time? Teams bring in their closers in the middle of an inning all the time. Papelbon entered in the middle of an inning 5 times out of 63 last year, 9 out of 65 the year before. It’s not standard, but sometimes it has to be done. There are ways to make it happen.

    Comment by elkabong — May 16, 2012 @ 1:03 pm

  27. Well maybe I missed it, but I don’t see where the author stated the Phillies would have won more games with ‘better’ use of their bullpen. the closest I see is ‘Could the Phillies find more spots to use Papelbon? Probably.’…

    Comment by Eric R — May 16, 2012 @ 1:17 pm

  28. Qualls has just been awful the last couple weeks. He thought he had found a “hitch” in his mechanics, solved it, and yesterday happened.

    On the plus side Jake Diekman is looking like a name we’ll be hearing a whole lot more of. 95-96 with movement and location, good sweeping slider he can locate, and all from a side-arming lefty. Philly’s next Bastardo.

    Comment by Muggi — May 16, 2012 @ 1:19 pm

  29. It’s fair to say the manager badly bungled the Monday game, because he had a 4 run lead and Qualls had only pitched 1/3rd of an inning when he lifted Qualls for a pinch hitter.

    You’re telling me that in that situation, where 99% of the time there’s going to be a sacrifice bunt, Papelbon can’t get warm?

    I’ll tell you that your estimated percentage there is not reasonable .Even if there was a hitter who could sacrifice successfully 99% of the time, managers aren’t going to choose to bunt there 100% of the time (well, Dale Sveum may chose to sacrificie bunt to lead off an inning). The real odds of a sac bunt there are probably more like 50%.

    But the larger point is, that these situations due happen quite quickly, and certain relievers can get up and sit down and get up and sit down repeatedly, but all can’t. Papelbon has never been used this way, so to imply that he can with no effect on his health and effectivenss, is speaking for a position of ignornace. Maybe he can, more likely, though he cannot.

    For the most part, when you warm up a reliever, you need to use him, or at least count it as a use. Baseball managers understand this, because they have pitched and talk with pitchers and coaches on a daily basis. People perusing box scores, where “warmed up” isn’t counted as a stat, often times don’t.

    Comment by The Real Neal — May 16, 2012 @ 1:29 pm

  30. I agree, Jason. There seem to be 2 possibilities:

    1) All managers over the last couple of decades are morons when it comes to bullpen management, and Fangraphs Commentors are much better suited for the task.

    2) Bullpen management is a lot more complicated than looking at a game log after the fact and deciding which reliever should have been used in which spot.

    Comment by vivalajeter — May 16, 2012 @ 1:50 pm

  31. So only half the managers in baseball would sacrifice bunt with nobody out and a man on second where one run automatically wins the game? Even if that were the case, which I think is a conservative estimate, I guarantee you Clint Hurdle bunts there 100 times out of 100. This is a guy who bunts the leadoff man from first to second with nobody out in a walk-off scenario. I’m agreeing with you that Papelbon shouldn’t constantly be up and down all the time. However, when you know you want him in the game, there are ways to stretch out his warmup time. Mound visit by the catcher, mound visit by the pitching coach, throws over to “check on the runner”, etc. I’m also contending that he should be up at the beginning of some of these innings anyway, which would give him plenty of warmup time.

    Comment by elkabong — May 16, 2012 @ 2:00 pm

  32. Most often is that managers go ‘by the book’ so when it doesn’t work, they don’t get roasted…?

    Comment by Eric R — May 16, 2012 @ 2:24 pm

  33. Eric,

    Read the comment thread to see what I am talking about. I was thinking of the commenters, not the author.

    Comment by Jason H — May 16, 2012 @ 2:27 pm

  34. I personally attended three of the four extra inning losses mentioned in the piece and questioned at the time why Papelbon was not being used. I was most definitely not alone in questioning Charlie Manuel’s decisions in the stands. I can assure you that this mismanagement is not only being questioned in retrospect, but during the flow of play. But thanks for assuming they only way FanGraphs commentors get their baseball information is via game logs. It’s not like the game logs are used to support points being made. I much prefer the “all managers do it, so it must be right” argument, it works much better than real evidence.

    Comment by elkabong — May 16, 2012 @ 2:30 pm

  35. I like to think that if a manager’s concern was not getting roasted when their moves don’t work out they’d try to minimize making moves that don’t work out rather than trying to infer what other people think they ought to have done. …at least in the former strategy they are actively trying to win! In your suggestion, managers put the opinions of others before their own when two are in conflict.

    Comment by Jason H — May 16, 2012 @ 2:31 pm

  36. The ideal usage for winning games short term is probably not the ideal usage to avoid injuries. Managers know this and choose to use the longer term strategy with their closer.

    Comment by kick me in the GO NATS — May 16, 2012 @ 2:32 pm

  37. elkabong,

    There is no evidence showing that managers, by and large, use their bullpens less effectively than they otherwise could have and that they lose games because of it. There are reasonable hypotheses that they do.

    “I much prefer the “all managers do it, so it must be right” argument, it works much better than real evidence.”

    Managers lose their jobs by losing and keep their jobs by winning. The system actually selects for managers that employ successful strategies whether they understand why they do what they do or not. This is a process akin to evolution. Evolution removes variation from populations by driving successful variants to fixation and unsuccessful ones to extinction. If all managers employ a strategy and there is no variation, you might guess that it is actually a pretty successful strategy. It is not an unreasonable position to take.

    Comment by Jason H — May 16, 2012 @ 2:40 pm

  38. I’m guessing most managers do this. It doesn’t take a statistician to realize that coming in with no outs, runners on 2nd and 3rd, middle of the lineup coming up, with a close game, that you probably need your best reliever to come in as opposed to clean inning with a lead with (I’m assuming) only having to face the meat of the lineup 1/3 of the time.

    I know people argue the psychological strain of not knowing when you come in or whatever or how coming in with just 3 outs to get for a victory is stressful. I just don’t buy it.

    Freddi Gonzalez drives me crazy with this. In the Cards series he brought in Livan Hernandez with the top of the order coming up bottom 10. He ended up getting out of a bases loaded jam twice. Once they got the lead, then he brought in Kimbrel. Most of the time, Livan doesn’t strand 6 guys in two innings and the game is over with Kimbrel still in the pen.

    He does shit like that with Venters too. If it’s the 6th inning with runners on 1st and 2nd he won’t bring in Venters. Even though Venters has a super high gb rate, he’ll bring in Medlen (which isn’t bad) or Durbin (which is stupid). He has Venters slotted in the 8th inning or at best, 2 outs in the 7th to get out of a jam.

    Doesn’t make sense to me but it’s baseball culture I guess and most people love the save stat.

    Comment by Antonio Bananas — May 16, 2012 @ 3:00 pm

  39. Jason, while I do like your comparison to evolution, I would argue that the reason managers don’t vary from the set strategy is because if they gamble and lose, they will be held accountable, whereas if they stick to the accepted norm and lose, the players are held accountable. Ultimately, I would posit that no other strategy has been tested with any sort of significance, therefore there has been no potential for variation. Additionally, the “natural selection” is not done strictly by wins and losses, but perception by a particular group, be it management, ownership, and/or the fans.

    I think this is a difference of opinion between you and I that won’t be resolved in this conversation. I do respect your take on it, however, and appreciate your support of your point.

    Comment by elkabong — May 16, 2012 @ 3:09 pm

  40. Elkabong,

    In my recollection, the history of bullpen management goes something like this:

    1) Teams used their best relievers as “firemen” and brought them into the game in tough spots.
    2) Tony La Russa introduced the use of a structured bullpen with predefined roles into baseball (mutation).
    3) La Russa’s strategy swept threw baseball like a wildfire to the point where in just a few seasons every team in baseball had adopted it (selective sweep).
    4) Every now and again teams will try alternate strategies like “closer by committee”, etc. but it always fails and the team quickly switches back to La Russa’s strategy or the manager and/or general manager gets fired (purifying selection).

    Classic evolution!

    Its not that other strategies aren’t tried, it is just that they haven’t yet been successful.

    …of course, the ultimate fate of every genetic lineage is extinction, so there may be a better strategy out there yet!


    Comment by Jason H — May 16, 2012 @ 3:21 pm

  41. So only half the managers in baseball would sacrifice bunt with nobody out and a man on second where one run automatically wins the game?

    Only half the time will the manager elect to bunt and have the bunt succesfully excecuted, depending on all sorts of circumstances. I’m not really onboard with the premise that he shouldn’t have been brought in to start the inning, anyway.

    Comment by The Real Neal — May 16, 2012 @ 4:13 pm

  42. What happens if you get Venters up in the 6th, then before he’s ready to come in, the other pitcher gets out of the jam, and then you need him in the 7th, 8th or 9th? So you get him back up, and he works out of a jam in the 8th, and then you can’t use him the next day.

    It’s not that cut and dry. There are some bad decisions made by managers, but really in this article, there’s only one or maybe two.

    Comment by The Real Neal — May 16, 2012 @ 4:15 pm

  43. What happens if the pitcher doesn’t get out of the jam and gets clobbered for a 3 run jack? If you were down by 1, suddenly you’re down by 4. I think preventing whatever runs you can at the time is the best way to go.

    I would think that it’d make more sense to bring in a long reliever with a lead in the 8th and let them pitch the 8th and the 9th. Use your better relievers for the intense situations when the starter is gassed.

    You’re right, it’s not always that simple, especially not in the NL. I’ve noticed Freddi Gonzalez likes to use Chad Durbin with more than 1 out with the pitcher spot due in the next inning. Basically using Durbin so he doesn’t have to “waste” a reliever on just 1 or 2 outs. As long as Durbin keeps getting out of innings with deep flyballs (like his last 2 outings) and not home runs, that’s fine I guess.

    Comment by Antonio Bananas — May 16, 2012 @ 4:27 pm

  44. Generally speaking, if all managers do it, it probably isn’t a mistake.

    Comment by Joebrady — May 16, 2012 @ 4:30 pm

  45. Clearly not all situations are cut and dried and there is the risk of having one of your top releivers (i.e. Venters) throw many pitches in the bullpen without even being used. However, there are also situations like the Phillies vs. Braves game on May 2 in which the Phillies led 12-8 after tacking on 3 runs in the top of the 8th. In a game against a division rival, the manager has to be prepared to go to his best reliever if the game gets close. Jose Contreras started the bottom of the 8th and was replaced during the inning by Michael Schwimer with one out, the bases loaded and the Braves having cut the deficit to 12-9. If there was enough time for Schwimer to get ready, there should have been enough time for Papelbon to get ready. Schwimer gave up a walk, a single and a sac fly to squander the lead before retiring the final batter of the inning. We don’t know that Papelbon would have done better, but my thought was at the time and continues to be, I would rather my $12.5M per year pitcher come into this situation than a guy who was in AAA the week prior.

    Comment by hk — May 16, 2012 @ 4:44 pm

  46. Most of the article is based on two assumptions that assumed to be true, but which cannot be prove. Mostly the assumption that anyone can close. If anyone can pitch in the 9th like they can in any other inning, then that would be true. However, it’s alway been my experience that many pitchers simpy do a real bad job in a save situation, that might otherwise be a good pitcher. I don’t think I have ever seen a manager who thought these innings were interchangeable.

    IRT to putting in Paps without a save on the line, I’ve seen plenty of managers do that. You don’t sit your closer because an arbitrary number like 3 becomes a 4-run lead. Are you really better off sitting your closer, after warming him up, only to have to warm him up again if your setup guy puts two guys on?

    I’d like to save IPs as much as anyone, but once a guy is ready to go, unless you’ve truly blown out the game, you’re probably best off going with him.

    Comment by Joebrady — May 16, 2012 @ 4:45 pm

  47. The whole warm up time is something that I don’t think is studied enough. If a guy has too little/too much etc. I was at the Braves/Cards series finale last Sunday. Martinez must have been warming up on and off for 30 minutes in the Braves pen. He came out sharp and then ended up getting relieved after allowing an enormous 3-run bomb to Allen Craig. I am convinced it was because he would warm up, sit down, warm up, sit down, play catch to stay loose, sit down, warm up, etc for 30 minutes.

    Does anyone know if managers have maybe notes about approximately how long each reliever needs? I’d guess that relievers with more experience need less time because they know their routine better. Or maybe guys with different pitch types require different times.

    Comment by Antonio Bananas — May 17, 2012 @ 3:42 am

Leave a comment

Line and paragraph breaks automatic, e-mail address never displayed, HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Close this window.

3.704 Powered by WordPress