Did Game Seven Delay the Bullpen Revolution?
For much of the postseason – with the exception of Buck Showalter’s decision to strand Zach Britton in the visiting bullpen at Rogers Centre in Toronto – it seemed the game might be on the cusp of a new revolution, a bullpen revolution.
For many, the major takeaway from October was how some managers were employing their top relief arms. Kenley Jansen recorded at least five outs in four of his seven postseason appearances, pitching three innings in Game Six of the NLCS. Aroldis Chapman entered seven playoff games before the ninth inning, and nowhere was the trend more dramatic or effective than in Cleveland.
Trying to piece together a pitching plan with an injury-depleted rotation, injuries in part allowing him to operate unconventionally, Cleveland Indians manager Terry Francona turned Andrew Miller into perhaps the most valuable player of the postseason.
Miller made 10 appearances and each began before the ninth inning. He entered most often in the seventh inning (four occasions), but entered as early as the fifth three times. The lefty also entered in the sixth twice. He appeared as late as the eighth. Miller recorded at least four outs in every appearance and went at least two innings six times.
Everything was going so well for the revolution until Game Seven…
And later, this…
Miller pitched 19.1 postseason innings. He allowed 12 hits, three runs, walked three and struck out 30. But all three of the runs he allowed occurred in the World Series, including two costly ones in Game Seven, when he was pitching for the fourth time in the series.
Chapman allowed six total runs in the postseason in 13 appearances but three of those were allowed in Game Six and Game Seven when he recorded at least four outs for the second and third time in four days.
Despite the success of the relief-ace model for much of the postseason the industry, will recency bias have an impact on decision makers? Will the events of Game Seven delay a change in the way relievers are deployed? Will managers continue to save their best arms for the ninth inning and save situations?
Game Seven is probably not going to help the case for a move away from traditional bullpen usage and alignment.
A number of managers were asked about bullpen tactics at the Winter Meetings in December in Washington, D.C., and there seemed to be little hint of any significant carry-over effect in 2017.
Said Colorado Rockies manager Bud Black at the Winter Meetings: “Managing a bullpen during the regular season is different.”
Black added that, because Cleveland had the luxury of a deep and talented bullpen, Francona had multiple trusted arms to handle multiple high-leverage situations.
“If you have the versatility and you have guys who do perform and you can sort of mix-and-match, and if, again, like the Cleveland model, per se, with Miller and Cody Allen they had two good guys, and you throw in another really good guy like Bryan Shaw, it’s really nice.
“I think with what you saw Joe [Maddon] do late in the year with Chapman, he felt as though he might not have other weapons.”
Tigers manager Brad Ausmus told reporters at the meetings the structure of the postseason simply warps the way games are played.
“The playoffs are going to change the way a bullpen is used, even the way starters are used,” Ausmus said. “Starters are often taken out earlier, as a result, if you have a stronger bullpen. Just doesn’t happen in a regular season because you get to August and the bullpen would be shot…
“You know, Tito did a great job with his bullpen, but if you can find me another Andrew Miller, let me know, we’ll stick him in our bullpen and see if it works. It’s not that simple. I don’t know that you could throw that many innings on a regular basis.”
Even Francona in speaking with reporters during the postseason expressed his doubts about the strategy he was employing carrying over to the regular season, though for a different reason: economics.
“I don’t think you’re going to see as much [change] as people think, just because of the way our — like the arbitration system. The way people are rewarded for saves. Again, I’d love to see that changed because I think if that was changed you would see how pitchers are used differently and I think we’d have a better game.
So long as saves are rewarded in the arbitration process, many pitchers will want to accrue them, and managers have interest in harmonious clubhouses and in avoiding the creation of disgruntled players.
So arbitration is an issue that might need to be addressed before any significant structural changes are made to bullpens. (While arbitration is not kind to a non-traditional reliever role, perhaps such a pitcher’s open-mindedness and versatility would be more valuable than saves to clubs on the open market.)
Regardless, after listening in December to managers, it doesn’t seem like we are headed for a sea change in April. It seems the non-traditional bullpen usage will mostly remain a postseason tool. There’s perhaps too much skepticism on trying to extrapolate the strategy over six months. There’s too much comfort in leaning on the starter-setup-closer model, which can also reduce second-guessing in postgame press conferences. Perhaps too many decision-makers saw Chapman and Miller’s diminished velocity, location and results in Game Seven. Perhaps that’s the lasting image that will guide decision-making.
But if there isn’t league-wide change, that means there’s a competitive advantage for a team with decision-makers and relievers who are willing to buy in.
Few ,if any, are proposing that a pitcher like Miller or Chapman would pitch two innings every other game like in the postseason, where the stakes and off days are greater in number. Even without drastically increasing workloads there are still opportunities for clubs to spike win probability throughout the season by employing their best arms in difficult situations, or in facing difficult stretches of lineup.
In an era where every team is trying to reduce inefficiency, much inefficiency remains in the bullpen. But if Game Seven is the lasting image, if Chapman and Miller are thought to have been worn down due to workload and lack of routine, then those wanting to see change might have to wait.
A Cleveland native, FanGraphs writer Travis Sawchik is the author of the New York Times bestselling book, Big Data Baseball. He also contributes to The Athletic Cleveland, and has written for the Pittsburgh Tribune-Review, among other outlets. Follow him on Twitter @Travis_Sawchik.
Fans sometimes forget that relievers often warm up without getting in a game, and that impacts their fatigue factor. They simply can’t be used as often in the regular season as they are in the post-season.
Game 6 is what delayed the bullpen revolution. Maddon unnecessarily used Chapman because he ignored (or never understood) the most important reason why leverage is a thing: the costs of fatigue.
I’m still waiting for a team to really use AAA as a taxi squad and have RPs go up and down. Make more suspect starters into RPs, and have them throw 1 time thru the order, no matter how they do, every other day or two, and then go down to throw less in AAA, then come back….
Teams with terrible SP should really think about some kind of “starter by committee” thing, where guys only go 6-9 outs (other than their 1-2 good starters), but which team is brave enough to really try something different?
Buck Showalter is notorious for that, hence the Orioles success, which has alot to do with bullpen management. Yea, he fucked up putting Jimenez in the Wildcard game when he should’ve used Britton, but overall everything you just said is what the Orioles have been doing.
There is a limit on this, in that your AAA team also plays every day, and you have to manage those players, too.
You would run out of player options rather quickly and then they’d have to make it through waivers each time you dropped them back down.
Option-years don’t care about the frequency of transactions.
“which team is brave enough to really try something different?”
— The 2012 Rockies used a 4-man rotation with a 90-pitch limit. It was a disaster; their starters posted the 2nd-worst ERA in Rockies history, and the record-high bullpen innings overtaxed their better RPs. That’s not a prediction about your plan, but I’m sure it was discouraging for any MLB exec trying to break out of the box.
The AAA shuttle idea can help with bullpen workload on the margins, but there are three rules that limit movement so that AAA can’t really be used as a “taxi squad” for relievers.
Each player only has 3 option years. In addition, any player with at least 5 years of service time has to consent to a minor league assignment, even if said player has option years left. So the pool of RP’s who can be put on the minor league shuttle is limited.
Another key limitation is that any player sent to the minors has to remain there for at least 10 days before being recalled.
http://www.fangraphs.com/library/principles/contract-details/player-options/
Players don’t want to change, because they prefer defined roles and they all hope they will rise to become a closer and command $20 million.
Managers don’t want to change, because the players don’t want to change, and because having a “book” to follow makes their job easier and insulates them from second-guessing.
GMs might want to change, but the potential incremental gain is not worth the cost of alienating your players, managers, and the 90% of fans that aren’t analytically minded but do love singing their closer’s walk-on song in the 9th inning.
I think MLB could fix this by re-working the arbitration system. I know they don’t want to introduce any advanced metrics, but if you made saves and holds closer in value, players and teams wouldn’t be under as much pressure to allocate innings the way they do now.
Your first two points are fallacies. It would only take one front office to make changes to impact the whole system.
If that single front office started rewarding younger pitchers for performance but leaving out saves, and the player saw that they were getting paid on other merits other than saves, they would be completely happy with that. That front office would just have to avoid paying “closer” money to a designated closer, and instead allocate that money to the relievers who pitched the best no matter where they were deployed.
Managers would then change, because relievers are getting paid based on actual performance instead of a stat that actually has no real meaning. Managers would then manage to win the most total games to keep their jobs.
Then outside relief pitchers would see that whichever team does this, values their true value much higher and wish to sign with that team more than others. Thus creating the necessary demand for other teams to step up.
While this is the oversimplified way to look at it, it is 100% true that once players are getting paid for different things, they will focus on that rather than what they used to get paid for.
One guy that I am going to follow in 2017 and 2018 is Dellin Betances. He has pitched the 6th most games (217) and the most innings (247) out of all RPs these past 3 years. Over this same span he has faced the most batters, and he’s 66 Ks ahead of 2nd place (Andrew Miller). This is a pretty heavy workload for 3 years running.
Now, Betances has been terrific as evidenced by his 8.5 fWAR over these past 3 years, which also leads the majors. However, if his effectiveness takes a big hit, or if he has a major injury, I could see teams using his workload as a cautionary tale. That would only slow down the bullpen revolution more.
The bullpen revolution will likely remain mostly a playoff thing for now; however, if the league ever drops back to a 154-game schedule and/or increases the active roster to 26 players, that will allow for it to creep into the regular season.
Jonny Venters
What you could do is look for similar workloads in the past and see how those players did going forward, that save you the hassle of waiting 1-2 years for results 🙂
2010-2012 Tyler Clippard, Sean Marshall and Matt Belisle were all top 10 in fWAR and batters faced among RPs.
2010-2012
TBF/yr fWAR/500TBF NAME
338 2.1 Tyler Clippard
338 2.3 Matt Belisle
290 3.6 Sean Marshall
2013-2014
TBF/yr fWAR/500TBF NAME
276 1.7 Tyler Clippard
292 1.3 Matt Belisle
59 0.4 Sean Marshall
Clippards batters faced and fWAR rate dropped by about 20%. Belisle’s batters faced dropped by about 15%, but his fWAR rate dropped like 45%. Marshall completely collapsed.
That is a pretty decent range. Some of the drop is probably age related; 2010-2012 they averaged 28.3yo and 2013-2014 30.8yo, so some amount of age related decline is probably to be expected. Some of the drop is presumably that by plucking the very elite in a three year range that it just happened to be their peaks and significant decline is just to be expected.
In any case, do this for a bunch of years and you can probably build a reasonable model to predict what will happen with Betances…
What might affect things even more is how miller (and maybe chapman) do this season, whether there are long-term effects of being heavily used in the postseason. It’s hard to miss that the royals, who kind of pioneered this strategy, saw their 2014 closer go down in 2015 and their 2015 closer go down in 2016.
I don’t think they will go away from using relievers more than 3 Outs but the collapse of chapman and miller probably was a reminder that you can’t ride your horse unlimited. especially chapman looked gassed and threw a couple mph slower.
going with your best guy is good but you probably have to Limit the workload of your bullpen ace and occasionally go a risk with the lesser guys to Bridge an inning.
it worked out in the end but especially the use of chapman in game 6 with what was I think a 4 run lead or so was a total head scratcher.
of course I understand that the cubs were facing Elimination and every lead can be blown but you have to pick your spots were you expose your lesser guys.
even in a PS series you can’t just manage for today 100%.
So I think the take home message is yes to an increased use of your top arms and yes to a quicker hook but there are Limits to what an arm can take and francona and maddon at least pushed that Limit if not exceeded that.
so I don’t think there is any room to increase the workload from what chapman and especially miller did, what we have seen is probably the abolute upper end and maybe a Little beyond.