Jan
22

The fireman and organizational incentives

By

When Rafael Soriano joined the Yankees, the cause célèbre of Yankee bloggers quickly became the use of Soriano as a fireman in the bullpen. EJ Fagan from The Yankee U and our own Ben Kabak have both discussed it recently in detail. In a nutshell, the Yankees now have two closers in Rivera and Soriano. Given that Rivera will pitch the ninth and that the eighth inning isn’t always the highest leverage moment that the bullpen will face, they argue persuasively that the Yankees should use Soriano to put out fires, whether those fires arise in the sixth, seventh or eighth innings. This concept is logical and well-founded, yet I think there’s a good reason to believe that the Yankees, or most other organizations for that matter, won’t employ it in 2011.

It’s no secret that the New York media is unforgiving. While Brian Cashman seems to avoid a lot of the nastiness, plenty of reporters assume a sarcastic and critical approach towards Joe Girardi. Their Twitter accounts during games are rife with jokes about Girardi’s matchup binder, and they seem to enjoy playing “gotcha” with Girardi’s information about player injuries and explanations about decisions. Very simply, an unorthodox idea like using Soriano as the fireman would likely be met with criticism in print, in the airwaves and on the internet. One can imagine the reaction if Soriano blew a lead in the sixth inning and the Yankees lost the game, or if Soriano saved a lead in the sixth but saw Joba Chamberlain surrender the lead later in the game.  It’s easy to picture the back page of the New York Post with a gigantic headline like, “Bonehead: Why is Joe Girardi using his $35 million dollar man in the sixth inning?”

Of course there is a very good answer to this question, one built on data, logic and research. But it’s a complicated answer and it doesn’t lend itself to a sound bite. It’s easy to say that Soriano is our “eight-inning guy, period”. It’s way more difficult to explain that the manager is going to try to maximize win probability by utilizing the best relievers in the highest leverage spots. It would also require Girardi to explain why Rivera isn’t used in the highest leverage spots, and only in the ninth inning, a question which would require him to admit that this idea is a bit of a hybrid between the traditional use of a closer and the more sabermetric-inclined concept of leverage and probability. In a media environment not known for kindness, friendliness to new ideas or nuance, one can imagine how badly this would play out. Who’s ready for a summer of arguing with the beat writers!

Of all the reasons not to do something, though, worrying about how the New York media would perceive you has to rank near the bottom. This reason would also be moot, and Girardi wouldn’t be the focal point of the criticism, if it was clearly communicated that he had the full confidence of the organization to execute this plan. As such, whether Rafael Soriano is used as a fireman or strictly in the eighth inning is a question of organizational incentives, a cost-benefit calculation that all relevant actors in the organization have to perform. Traditional bullpen management works well enough. Put another way: traditional bullpen management is orthodox, accepted by fans, media and other organizations alike. There may be a much better way to do it, but no one at the moment seems to be trying it. The potential gain is not losing a lead, something that most people assume as a given anyway. Think about it: in the best-case scenario the team doesn’t surrender a lead that it already has. The downside risk is a bit greater. For one, the team could actually lose the game in question, should the fireman give up the lead. The manager could lose the faith and confidence of the fans, or worse, his superiors. Ultimately it’s at least possible that that the manager could lose his job.

Unfortunately, we don’t have a very good sense of what organizational incentives are like for the New York Yankees. Our knowledge of the inner-workings of the Yankee organization is quite limited. Our first-hand information is self-selected by the GM and the Manager, or comes at times of unrest and dissension (the Cashman-ownership split on Soriano, for instance). Our second and third-hand information is even more problematic. It comes through the conduit of reporters and leaks. Often times one has to wonder whether the information made public is designed to serve some sort of  Machiavellian purpose. What part of the organization is leaking this information and why? Are they attempting to undermine another part of the organization? Are they lying in order to throw competitors off the scent? Is this simply good information? But this is about as deep as most fans can go. We simply will not know what Cashman and his cadre of advisors did this winter in secret. We won’t know if they looked at the possibility of a four-man rotation, batting Cano second or using Soriano as a fireman. We won’t hear about the ideas that they entertained, researched, debated and ultimately rejected.

As such we have to at least consider that the concept of Soriano as fireman has been explicitly research and rejected by Yankee management. Yet we also have to consider the flipside, that the organization is more by-the-book in certain areas, and that taking big risks in high-profile situations isn’t encouraged. This would mean that no matter how sound or logical the concept is, the Yankees will never be the first team to adopt it. Ned Yost of the Royals perfectly captured this sentiment yesterday when asked about the concept of a fireman. Kings of Kauffman has the quote:

Yost is still a baseball guy, and there’s a way things are done in baseball and a way to not do things.  Innovation isn’t a popular idea.  Using Joakim Soria in an early situation might make sense by the numbers “but you won’t catch me doing it.”

This is an unsatisfying feeling and it’s one familiar to anyone who has worked in a corporate environment and found that doing things as they’ve been done in the past and not looking like an idiot is more important than trying to invent new ways of doing things and possibly failing. It’s flat frustrating when the best idea loses out to the more familiar idea. It’s also bad organizational management, because it aligns the interests of the employees with keeping their jobs and not screwing up, rather than allowing a certain amount of room for failure and fostering innovation and productivity. But the Yankees don’t particularly need to reinvent the wheel. They don’t need to discover untapped markets of value like the Rays or the Athletics need to in order to succeed. In New York, where the lights are as bright and as hot as anywhere on the planet and where “what have you done for me lately?” is a way of life, there is little margin for failing and looking dumb.

This summer, Girardi is going to leave Soriano and Rivera chucking sunflower seeds against the plexiglass as a lesser reliever blows a lead. Sergio Mitre may pitch the 13th inning of a tie game on the road while Rivera waits for the team to get the lead before coming in. When this happens, it’s important to recognize that it’s not necessarily because Girardi is thickheaded or stubborn, too smart for his own good or intentionally trying to annoy the curmudgeonly beat writer crew, although the latter would be spectacular. Girardi may be a very public face of the Yankees, the one who projects authority and whose face is on television every night, but ultimately he’s another organizational actor subject to peer pressure criticism from his superiors. What Girardi doesn’t want to become is another Jeff Zucker, who risked job safety and ratings certainty on an unknown quantity with arguably higher upside and long-term success only to have to back out of it when it turned difficult. At that point, Zucker’s fate was written on the wall. It was only a matter of time before the house came down.

Categories : Death by Bullpen
  • http://twitter.com/JoeRo23 The Honorable Congressman Mondesi

    As annoying and awful as his columns are, Easterbrook has (correctly) noted this before in a football concept. I honestly haven’t read TMQB in years but I remember he’d rail against the idea that head coaches would never buck the conventional wisdom, no matter the mounting evidence for doing so, because to do so would put them in position to be judged by management/media. It’s unfortunate, but such is life. Over time things do change and new ideas are adopted, it just doesn’t happen as quickly as it ought to.

    • http://twitter.com/JoeRo23 The Honorable Congressman Mondesi

      “As annoying and awful as his columns are, Easterbrook has (correctly) noted this before in a football concept context.”

    • Sayid J.

      Most notably the lack of attempts to convert 4th and shorts, especially in that zone of “we aren’t in our own territory but I don’t want Nick Folk to kick a 55 yarder so instead we’ll just punt it into the endzone and game 20 years of field position”

      • V

        Thank goodness the Chiefs have a head coach who’s willing to play the numbers. Very refreshing.

        I really really wish someone would round all of the mediots up and toss them onto an island somewhere.

  • http://jukeofurl.wordpress.com Juke Early

    As with the earlier post this week regarding using Soriano ( and/or his calibre) to come in for leverage spots, I could NOT agree more. Being a fireman is what can really make this guy valuable. Of course, when & how many times this can be the right call – is key.

  • http://twitter.com/steveh_MandAura Steve H

    I don’t see why losing teams wouldn’t try to re-invent the wheel when it comes to bullpen use. It’s a no lose situation. I also don’t see why bad teams employ full time closers either, but that’s a whole different story.

    With regards to the 2011 Yankees, I’d rather them try to manage towards leverage a little more, but I don’t think there would be a huge benefit, in that they have several good relievers. David Robertson has shown an innate ability to put out fires (though I’m sure a lot of luck is involved), and Joba is still a damn good pitcher. In a tight spot in the 6th, would I rather see Soriano than Joba or D-Rob? Probably, but the difference really isn’t that great. I’m not sure if it’s a benefit or a detriment, but with a loaded pen like the Yankees have, the manager doesn’t need to think too much to do his job, he can pretty much just assign certain innings to guys with a good chance to succeed. Joe Torre might find a way to screw it up. BP management has been one of Girardi’s better traits since becoming manager, and in a way, having a great BP might diminish his overall value as a manager just a bit (if that makes sense).

  • smurfy

    I enjoy a game managed flexibly, and some managers with a great feel for it have worked wonders with their pinch hitters and pitching changes. (Zim had some great seasons.)

    This does seem a rare opportunity to test the dichotomy, save the best for last v. play the best at critical times, and I would love to see the Yanks experiment. Robertson and Joba are viable alternatives though, and I would hope it could happen without a set of rules.

  • http://twitter.com/steveh_MandAura Steve H

    One of the problems I have with the leverage argument is that you are making assumptions about the future. While you may encounter a high leverage situation in the 5th or 6th, you might, in the same game, encounter an even higher leverage situation later, and you may have already burned your best reliever. It’s like going for 2 in the NFL too early. I can’t stand when you see an NFL team go for 2 anytime before the 4th quarter on the assumption that the other team won’t score again. That being said, I do think there is a way to utilize situations with higher leverage as opposed to set rules, but it’s not black and white by any means.

    • V

      If only there were a way to utilize something called probability…

    • whozat

      Yeah, but Steve…the flip side of that logic is that you might NOT have that higher leverage situation, and then your best reliever sits in the pen waiting for, at best a lower-leverage situation than the one you just had your third-best guy throw in.

      You can only make a decisions based on the information at hand. It’ll never be perfect, but you kinda have to set a threshold of leverage and pull the trigger. It’s like selling stock.

      • http://twitter.com/steveh_MandAura Steve H

        I agree, it’s just not a black and white for every situation. If a team were to adopt this sort of bullpen management, I don’t think you could criticize them for not playing to the leverage situation in every single scenario. If you have a 2-1 lead you might, if you have an 11-10 lead, you might not. Either way it’s a 1 run game, but they are two different games.

      • YankeeJosh

        To that point, if you change the way you manage a game, that will also change the probabilities associated with when leverage situations arise. More simply, isn’t it possible that a reason later innings often have lower leverage situations is because the better relief pitchers are pitching then?

        For example, is it really hard to imagine a situation with runners on Second and Third and one out in the Sixth, Yankees up 1 run. Soriano comes in and closes the door. Now in the 8th, let’s say the Yankees have tacked on a run and lead by two. Joba gets in a mess, Feliciano comes in and fails to get a lefty and suddenly a run is in and the bases are loaded with no outs. Suddenly there’s a higher leverage situation in the 8th than in the 6th. However, if you used a lesser reliever in the 6th, even if he gave up the tying run, assuming the Yankees still take the lead, Soriano comes in and shuts down the 8th without every creating a high leverage situation.

        I agree with Steve H, it’s not black and white. High leverage situations would seem to more often be caused by lesser relief pitchers. It’d be interesting to see the numbers on that to get confirmation. The fact is that there are 27 outs to get in a game, and the bullpen has to get whatever number of them. Using your best pitchers in high leverage situations early leaves you vulnerable to higher leverage situations arising later in the game where they might not otherwise.

        Using pitchers based on leverage is a great concept in theory, but I’m not sold on it being practical in actuality. At minimimum, I think it needs to be looked at more. And with all that, I hate the idea of having one set “8th inning” guy a la Tom Gordon, Joba and now Soriano.

  • Jonathan

    Excellent post. Any relation to Steve Rhoades on Married with Children?

    I wonder if David Robertson would be a factor in keeping Soriano in the 8th exclusively. Not that I disagree with using your best relievers in the highest leverage situations, but the fact that we have a guy who strikes out more than a batter per 9 innings would make the gain smaller than most teams with only 1 or 2 good relievers. Joba can be thrown into this mix as well along with the Feliciano/Logan for tough lefties.

  • pete

    I’m generally a pro-fireman guy, but I don’t think it’s the best way to run this bullpen. To be honest, I’d be perfectly comfortable with any of Mo, Soriano, Joba, and Robertson in just about any situation, especially when you’ve got both Logan and Feliciano for tough lefties. Thus I really think the best way to manage this bullpen would be to essentially “ignore” situations (excluding exceptional circumstances, whatever they may be), prioritizing instead consistency; I personally believe, based on my admittedly limited experience pitching throughout high school, that routine and regularity are essential for maximizing a pitcher’s effectiveness.

    I really think that all four of those guys could put up sub-3.00 FIPs if they pitched on a regular and consistent schedule, and if the yankees had four sub-3.00 FIP bullpen guys, then it really wouldn’t matter which of them pitched when. I went into more detail about a way to organize such a thing here: http://riveraveblues.com/2011/.....nt-1561429

    and am curious as to what you guys think.

    • smurfy

      Yeah, Pete, that does make sense to me. It often seems a crapshoot, using several pitchers, one always seems to have a bad game. If he’s been called into a tight spot, we are then behind.

      Mo definitely prefers the regularity of his use: I emailed the fireman question to “ask Joe,” and he answered a couple weeks later that Mo’s health calls for adequate prep time, and no getting up, sitting down.

      I can readily see that everyone would prefer that. I would do it a little differently than you described: I’d add one of the lefties to the regular use list, but I wouldn’t quite schedule the guys, rather mix them, depending on the oppo lineup, and on who has been smoother lately.

      Sometimes I’d have one go two innings, and try, as you mentioned, to give them advance notice to get ready. So, we’d have a couple designated firemen, a righty and a lefty.

  • LarryM.,Fl.

    The Yankees have their meetings about many player moves that might be helpful to the organization during the year both in and out of season. Why should it become public knowledge? Do the Yankees have to be transparent to the public making the reporters job easier. I chuckle at the reporters who openly critique a managers move from the press box. Torre managed by feel and history of the players performances. Joe rarely employed the binder but Zimm’s head. Joe was guilty of killing relief pitchers careers and the butts of rookies planted on the bench.

    Girardi employs the binder. He loves stats as does Cashman. Giradi has an engineering degree, plenty of math skills and comprehension skills to you use when the facts are in front of you.

    Torre and Girardi have the same fault. They both rely on their beliefs in running a game. They rarely divert from the center of their beliefs seeking a left route or right route to a win. Much like our brilliant politicians in DC. So in the end you’ll see for the most part Soriano and Mo as the 8th and 9th inning guys but at times Soriano in the 9th when Mo is spent from prior days work. It should be great to watch as the Yanks play the Rays and Sox’s in 4 game series with those two at the end of the game. One final thought the wheel has been invented and its a wonderful invention just like a superior bullpen to bring home a win.

  • http://www.richardiurilli.com Richard Iurilli

    I agree with the premise of this article, that Soriano will likely be used in the traditional eighth inning role rather than in an innovative fireman’s role based on leverage.

    I do question the statement about tie games on the road, however. Without looking at the box scores, I seem to remember that Girardi was more willing to use Mo in a tie game on the road without a save situation after that debacle in Toronto where Chad Gaudin blew the game in the fourteenth while Mo sat in the bullpen. There were exceptions to this, even after the game in Toronto, but, if my memory serves me correct, they were because Mo was unavailable to pitch.

    Maybe I’m wrong and my memory is completely off (a likely possibility), but it would be interesting to study more in depth.

    Still, another great article from the new weekend crew. Good job.

    • http://www.richardiurilli.com Richard Iurilli

      The game in Toronto that I was referring to was this one (http://goo.gl/olYTK) on June 5. A quick look at Baseball-Reference shows that the next two games in which the Yankees were tied in the ninth inning or later on the road were this game (http://goo.gl/9lNC8) on June 23 in Arizona and this game (http://goo.gl/rml4c) in Los Angeles. In both games, Girardi went to Mo with a tie score in the bottom of the ninth. Interestingly, in both games, the Yankees scored in the tenth inning and Mo stayed in to pitch a second inning and earn the win.

      I know that two games is far from a large enough sample size to draw any conclusions, but I found it interesting that these were the first two times that the specified situation arose after the outcry in Toronto.

      • http://www.richardiurilli.com Richard Iurilli

        Okay, now I am officially perplexed.

        I looked through the rest of the box scores in 2010 and, to my surprise, Mo did not enter another tie game in the ninth inning on the road. There were six more examples of games in which the Yankees were tied going into the bottom of the ninth inning (8/10 @ TEX; 9/10 @ TEX; 9/13 @ TBR; 9/14 @ TBR; and twice in a doubleheader on 10/2 @ BOS).

        In two of those games (8/10 and 9/14), Robertson was the first man out of the bullpen in the ninth inning. In two more (9/10 and 10/2 part 1), that man was Hughes, in one (9/13), it was Wood, and in one (10/2 part 2), it was Nova. In four of those six games (8/10, 9/10, 9/14, and 10/2 part 1), Mo either started or entered in the middle of the tenth inning. In the second game of the doubleheader on 10/2, it made sense to not use Mo as he had already pitched that day and the Yankees would have wanted to rest him for the postseason. The only game in which a non-Robertson/Hughes/Wood reliever entered a tie game on the road after the game in Toronto was on 9/13, when CC was relieved by Logan in the ninth inning, who was followed by Gaudin and Mitre. I don’t remember a specific reason why none of Robertson, Hughes, Wood, and Mo pitched in that game, but I imagine that there is one.

        Conclusion: After the game in Toronto, Girardi went to a non-Robertson/Hughes/Wood reliever just twice in eight opportunities. One of these, the second game of the doubleheader on 10/2, is easily explained, but the other is noot. I think it is fair to say, however, that Girardi did, at the least, learn to use his setup relievers and then Mo before resorting to the back of the bullpen.

  • brian paul

    Great article Stephen! So far I am really enjoying your writing style. Great points too, about the contrast of new and old baseball philosophies, and how Ben’s proposed “hybrid” use of Rafi and Mo would not be well received in the media.

    The truth of the matter is, wins will ultimately affect how to media will crtique managerial decisions. So although I agree that the media may influence how Girardi uses Soriano, it would mean a lot to me as a fan to see my team being a bit more creative and strategizing in a statistical method.

    Again, great work!

  • Monteroisdinero

    I think Soriano will close more games than folks think due to Mo’s age/pitch count/minor injuries/inability to pitch 2 or 3 consecutive days etc. This will/should be a big plus for us.

    I could just see Billy Martin burning a binder symbolically during a press conference. He could have cared less. Not always successful of course, but cut of his own cloth. This does not contribute to job security however when things go wrong.

    • Billion$Bullpen

      Martin was rarely unemployed.

      • fire levine

        Doesn’t mean his job was ever really “secure”

  • NJYankeeFan

    I think Girardi has done a nice job in his time with the Yanks. Reporters don’t like him because he holds things back in a Belichek kind of way but most of them don’t know shit about baseball anyway. Sad thing is they can sway public opinion far too easily.

  • Billion$Bullpen

    NO valid reason why you would not use Soriano as a fireman. The way elite relievers are wasted in MLB is flat out silly. I have talked to a few former MLB relievers and closers and they all say the same thing “i need to know my role, i need to know when I am going to be put into a game” Seems like most of them just sit out there and eat candybars and get rubdowns until like the 6th when they leave the clubhouse and then hang out in the bullpen until they usual time they get in the game.

    But if a manager with balls changed that and let them expect to be used when he needs them and that could come at any time, maybe the elite ones would get it.

  • Kevin Ocala, Fl

    It’s generally too late if a pitcher is brought in the middle of an inning though of course it happens. But if this stategy is to followed to it’s logical end it would mean a pitcher would have his career burned up in a year.

    Therefore, since it’s the “third time around” that get’s many pitchers (especially the lower-rung guys) why not bring the “fireman” in to start the 5th or 6th inning when the heart of a line-up is due and the starter is showing signs of falling off the track. Naturally this stategy is score dependent, but IMHO, it makes more sense than having a man try to get loose quickly and boom, the game is broken open before he gets to the mound. Another thing, in most games the bottom half of the line-up is hitting in the 8th which is the real reason that their are “relievers” and there are true “closers”. Which would seem to further the idea of bringing the so-called “fireman” in to start an inning. I.E. the meat of the order will (or very likely)be coming up.

  • roadrider

    I’m confused – do the writers on this site have no knowledge of baseball history pre-Tony LaRussa or do they just have short memories?

    The “fireman” concept is neither new nor innovative. It was the dominant mode of relief pitcher usage prior to the advent of the modern closer role in the 80s and 90s.

    As much as I love Mariano Rivera and appreciate all that he’s accomplished for the Yankees I think that modern closers are overrated and not used to full advantage by their teams because of the distortion of the save rule.

    I remember the way Sparky Lyle and Goose Gossage were used and although they did often come in to finish games it was not uncommon for them to be called in before the 9th inning, sometimes in the middle of innings to put down a rally.

    I agree that the Yankees and Girardi probably think they’re playing it safe by conforming to the “LaRussa rule” but it’s a flawed assumption. Like Joe Torre before him, Girardi seems to think that that protecting a lead in the 7th inning is not as important as preserving the same or larger lead in the 9th. Since statistics show that modern closer usage has not appreciably increased the number of 9th-inning leads preserved (http://www.baseballprospectus......eid=12747- free article) it follows that reverting to the pre-LaRussa fireman pattern (and even expanding on it) would be a more optimum use of a valuable resource.

    Keeping a highly-paid specialist like a Mariano Rivera in reserve for a situation that might not even occur while refraining from using him in high-leverage situations in earlier innings so a meaningless statistic (the save) can be accumulated is not only a silly strategy it’s counterproductive. In fact, this “slave to the save” philosophy employed by Joe Torre was a large factor in the Yankees losing the 2003 World Series.

    How many times have we seen Rivera used in blowouts or obviously lost games because no “save” situations had recently emerged and he needed in-game work? Think how many times he might have been used to better advantage in the intervening games even though there were no 3-run 9th inning leads to preserve. The bottom line is to win tonight’s game (within reason – obviously rest and injury prevention are considerations) because tomorrow it might rain or the game might be a blow out.

    • MikeD

      Actually, people at this site do understand that. It’s been discussed many times. This article is even pointing to using Soriano in an “old style” fireman role. Most here would agree that the modern save (from Dennis Eckersley/Tony LaRussa forward) is stupid.

  • Jess

    I think we should use Soriano and Rivera in the 1st and 2nd innings. Those are the highest leverage innings. I hear there is a statistic that says over 70% of teams win if they score the first run. That’s too important a stat to ignore. This would be the best use of our bullpen. First three innings would be bullpen. Then the starters goes the last six. We should load up our lineup and pitching to try to get that first run.

    Great success!