Playoff Hindsight and Short Rest
The playoffs often cultivate an ‘anything goes’ environment, where aces like Zack Greinke and Cliff Lee become legitimate bullpen options and starters get lifted before the fifth inning ends for a more favorable matchup. In a sense, playoff rules trump what fans grow accustomed to during the regular season as every single game is of tremendous importance. While many managers alter their mindsets to treat each playoff game as a must-win, the 2011 postseason has shown that the more holistic view of the series can lead to better decision-making processes.
Last week, Dave Cameron discussed Jim Leyland’s refusal to use Justin Verlander in Game Four of the Championship Series. Leyland’s quotes — mainly that using Rick Porcello over Verlander was a no-brainer, or that not using his ace was the best thing for the team — drew significant ire but ultimately made sense. Verlander pitching on short rest would reduce his effectiveness, and the Tigers still needed to win at least two games after that outing. Pitching him on short rest would have been an example of managing for the present as opposed to the future.
Though that type of approach is traditionally more optimal, especially in a five-game division series, it makes little sense in the latter playoff rounds.
Leyland wasn’t alone in that line of thinking, as both Ron Roenicke and Tony LaRussa made decisions based on playoff series, as a whole, instead of a collection of immensely important individual games.
Shaun Marcum pitched very well for the Brewers this season, and without his acquisition it’s unlikely that Zack Greinke waives his no-trade clause and joins the team. Despite that, Marcum will be viewed as a major reason the Brewers bounced out. In three starts, he produced a 16.20 ERA over 9 2/3 innings. His strikeouts and walks were identical at five apiece, and though he kept the ball on the ground more often than not, he wasn’t missing bats. Perhaps it was standard fatigue — he is still two years removed from major surgery — or random ineffectiveness, but the common narrative was that he looked cooked and eminently more hittable than at the regular season’s midway point.
Even though Marcum’s struggles to end the season carried over into the playoffs, manager Ron Roenicke unequivocally stated that the former Blue Jays starter was his Game Six starter. Yovani Gallardo was deemed unavailable, even in the bullpen, because managing completely for that sixth game would hinder the Brewers’ ability to win a potential deciding game. Gallardo didn’t figure to perform well on short rest in Roenicke’s eyes, and the perceived difference between him on short rest and this version of Marcum was minimal. Maybe they gained a smidgen of win probability by going with Gallardo on short rest, but they lose a good deal of win probability for the potential deciding game by not using Gallardo and having to use someone else on short rest.
Hindsight is 20/20, and it’s easy to buy into the idea that Marcum shouldn’t have pitched given his recent struggles. But that’s really just hindsight. There was never any question he would pitch, and he wasn’t clearly ailing to the point of predictably poor pitching. On the other side, hindsight proved Leyland correct, as Porcello pitched decently even though the Tigers lost that game, and they won behind Verlander in Game Five. The results were divergent, but the thought processes were identically built off of the proper foundation.
LaRussa took a slightly different approach to managing a series on the whole. He actually chose to start Chris Carpenter on short rest in the division series to ensure that the Cardinals ace would make two starts against the Phillies. Carpenter was never considered for the opening game, but LaRussa could have kept him out until Game Three, when he would have thrown on short rest. However, that would have prevented Carpenter from making more than one start in the series. LaRussa may not have believed the series would go all five games, but starting Carpenter in the second game, on short rest, was a risk management move in case the series did get to a deciding game. Lo and behold, Carpenter, on full rest, pitched a shutout against the Phillies in that deciding game.
LaRussa’s thought process was different here, despite accomplishing the same goal of managing for the entire series, as he made the most of the situation dealt instead of creating the situation. Three managers, all of whom adopted a holistic view of their playoff series, two of whom experienced success in that style. Though Marcum struggled, it didn’t invalidate Roenicke’s decision. While managing for a future that might not exist by hurting the present odds of succeeding is typically ill-advised, it can be correctly utilized in the playoffs when it comes to starters throwing on short rest. Kudos to Roenicke and Leyland, especially, for making the logical moves instead of those born out of panic.
Eric is an accountant and statistical analyst from Philadelphia. He also covers the Phillies at Phillies Nation and can be found here on Twitter.
The Brewers and Tigers chose to maintain rotation order and the Cardinals chose to skip ahead and start their ace on short rest. You are essentially complimenting the managers here on doing two different things, but shaping them in such a way (both used a “holistic” approach) that they appear similar, when in fact they are the exact opposite strategy.
If anything, this shows that both strategies can be successful or (in the Brewers’ case) unsuccessful, not that abandoning a short rest strategy is always correct.
“LaRussa’s thought process was different here, despite [[accomplishing the same goal of managing for the entire series]], as he made the most of the situation dealt instead of creating the situation.”
“Kudos to Roenicke and Leyland, especially, for making the logical moves instead of those born out of panic.”
I hope those quotes illustrate his point for you, which you evidently missed. If not, I’ll summarize:
The author (Seidman) is trying to say that both managers made Starting Pitcher decisions with a view towards winning the series, not towards winning that night’s game. He is making a comparison with other managerial decisions that are tailored towards winning a particular game at all costs, without considering the series as a whole. For example, a hypothetical decision to pitch Verlander on short rest wins you the game he starts, but not the start he was supposed to take, so you lose the series anyway.
So, LaRussa and Roenicke are making a different decision, but have either stated in an interview or logic has made clear that their decision was geared towards winning the series as a whole, not a particular game of the series.
Thanks for the explanation, but I understood the point just fine. My issue is that the point is convoluted.
My contention being, if Leyland had decided to start Justin Verlander in game 4, I can assure you that the decision would have also been made holistically, as it would have put Justin Verlander in position to start game 7 as well. In fact, this is essentially what TLR did in Game 2 by starting Carpenter on short rest; he made that decision so that he would be available to start Game 5. So no, it is my opinion that you are incorrect about Leyland’s decision-making process (as is Seidman), if had chosen to start Verlander on short rest he would be taking the same “holistic” approach that TLR took.
Basically, they’re all taking a holistic approach.
There was a time when pitchers routinely had three days between starts.
There was a time when pitchers were not nearly as durable as they are today.
jmonkey – where is the evidence of that? i’ve done quite a lot of studying of the pitch count myth, and there is no statistical evidence that i have found that today’s players are more durable than in the past. at best, you can argue that modern surgery can resurrect a career, but that’s hardly the same thing as arguing that players are more durable now than they were when 3 days’ rest was the norm. if you have some empirical data, i’d love to see it.
They are conditioned to pitch on more rest nowadays. There is a good article (can’t find link) that shows pitchers perform slightly worse on less and less rest.
Joele-
I don’t have empirical data, because I haven’t studied the question. My conclusion that pitchers are more durable was an impression drawn from this piece by Jonah Keri: http://jonahkeri.com/2010/09/13/pitching-injuries-and-rangers/
He doesn’t delve into historical statistics, but if you scroll to about halfway down (although, I have to tell you, it is an incredible article, I would read the whole thing anyway), you’ll see a subchapter that begins thusly:
“Ryan is hardly the first person and the Rangers are hardly the first team that’s tried to squeeze more out of its starters. Plenty of managers have talked about the need to toughen up pitchers, push them harder and make them throw more pitches and more innings in the past. The results have, for the most part, been frightening.
In the past 15 years alone, pitchers like the Royals’ Jose Rosado, the Mets’ highly-touted “Generation K” of Paul Wilson, Bill Pulsipher and Jason Isringhausen, and the supposedly indestructible Cubs duo of Kerry Wood and Mark Prior have all seen their careers severely curtailed, or simply ended, by the strain of heavy workloads at a young age. Before Generation K, Wood and Prior, phenoms like Mark Fidrych also flamed out of baseball after the strain of overuse became too much to bear. These burnout cases have become so common, we remember those who heaped the high pitch counts on their young charges – Tony Muser, Dallas Green, and statheads’ Patron Saint of Pitcher Abuse, Dusty Baker – almost as vividly as the pitchers themselves. ”
The rest of the subchapter delves into why this might be, and what other people think about it. To me, this suggests strongly that 3-days between starts is insufficient.
Also, I think there is do research that identifies the optimal amount of rest, and its 5 days between starts, with a slight drop-off in performance when using 4 days of rest instead. Don’t really know where I read that though.
There’s really no way of knowing for sure no matter what data you collect. All we know is that pitchers are treated differently nowadays, and that practiced repetition must have an effect… Different bodies react in different ways… I would argue that you can’t generalize durability and conditioning much beyond an individual. There are certainly trends that take place, but you’d have to go back into a pitcher’s usage as a little leaguer… high-school workload… NCAA workloads as compared to HS draftees that are integrated into a workload more similar to what they would experience at the major league level… you have coaching variables at every level… it’s endless.
I would assume that it may or may not be a myth because it’s impossible to articulate patterns that are accurately translated between individuals. The larger the empirical sample, the more mush you get. The smaller the empirical sample, the more it speaks to maybe that individual’s unique circumstances.
“Also, I think there is do research that identifies the optimal amount of rest, and its 5 days between starts, with a slight drop-off in performance when using 4 days of rest instead.”
We certainly have a more robust collection and understanding of data since moving to the 5 man rotation, so to compare what we have and know now to anything off into the past is probably misleading or incomplete at the very least. I think it would be reasonable to see a generally slight drop off using anything but 5 days rest between starts when a pitcher has been trained to perform that way through an extended period of time. But, it doesn’t mean that every pitcher will react the same way. In the case of the playoffs, it’s a very individual and I think almost spontaneous understanding of what that pitcher may or may not be capable of, since there aren’t enough similar past occurrences with any pitcher to base a reasonable decision.
I don’t fault RR for not pitching Gallardo, but Narveson was clearly the correct choice yesterday. Soft-tossing lefties are exactly the type of pitcher the Cards always struggle against and there was considerable evidence that Marcum was cooked.
I am guessing you are being sarcastic but I can’t be sure. I know this is hindsight and all but the Cards did not have a problem solving Narveson yesterday either.
Is it possible the cards roughed up Naverson because was pitching for the 3rd time in 5 days, when he’s used to only pitching once every 5 days. RR made 2 major blunders. The first was pitching Naverson later in game 5 when the outcome was already decided. The second was starting Marcum in game 6 when Marcum was clearly toast.
And it’s not entirely clear the cards would have roughed up Naverson. 2 of those 5 ER they score on him came in when Hawkins let a couple of inherited runners score.
Nobody brings up these points for some reason.
Hindsight is indeed 20/20, but Runnin’ Ron Roenicke’s foresight is the real problem here. He was questioned all week if he was serious about starting Marcum in game 6. He seemed to be the only one who believed it was the correct decision. I’m not sure what Marcum’s opinion was.
Marcum was clearly done for the year. He threw 200 regular season innings. His previous high was 195. Instead of his fastball/changeup being at 88/79 like it was mid-season, It dipped to 87/81. We saw the results in atrocious September and Post -season starts.
Jack Moore outlined it well here:
http://www.fangraphs.com/blogs/index.php/is-shaun-marcum-burnt-out/
WPA for starting Shaun Marcum:
In Game 6: -.315
In Game 7: -.500
I wasn’t convinced by Cameron’s earlier article that starting your ace(s) on short rest to gain him (them) an extra start or two in the playoffs is a bad idea, and I’m not convinced by you.
It is also not short-term thinking, but exactly the opposite.
+1, the team that did it is in the World Series. I’m starting to think this fangraphs isn’t the be all and end all of baseball blogging, my suspicions were first raised when seeing that “bleacher report” widget on the side. I have no evidence that pitchers are tougher or just as tough these days, but i do know i have read of guys throwing 200 pitches in games back in the golden age, of throwing 150 pitches TWICE in one day in a Doubleheader!
Ron is not looking good here esp. with Mr. T. Plush, those antics are cute when you win, when you lose and put your foot in your mouth you wonder if the manager is doing anything to control this asshat mentality.
Also the whole “closer” deal i still don’t buy, while i embrace the new stats, i guess i’m still old school about certain things in baseball.
I don’t think it’s hindsight when you look at how horrible Marcum has been since September. This was a predictable and preventable start. He should not have started that game, and Roenike was playing to not hurt anyone’s feelings by starting him. It’s foolish to worry about a game 7 that you might not get to. Gallardo was their best chance to stay alive in the series. Then they’d have Wolf for game 7, a veteran pitcher that could gut it out. If Gallardo could have thrown 6 innings and given up 3 or 4 runs, the Brewers most likely win the game. It was a really foolish decision.
One other note: Not starting Gallardo was akin to when managers stupidly don’t bring in their closers in the extra innings of a game on the road. You often wind up saving your ace for a chance that never combs. And you lose.
Unfortunately I think that’s much of the state of the game. Today, it’s perceived that players (pitchers especially) are driven by a manufactured label of a specific role they play and the routine that follows, and it’s assumed that when they are used outside of that scope, bad things are more likely to happen. Maybe that’s the reality of the situation… you end up relying on that routine when it’s all that you’re used to. It’s kind of a shame, that so many are afraid to break a pattern.
Excellent post, I will make sure I look here frequently from now on.
I want tell you, I think you will find there’s problem with your RSS, it’s not at all displaying correctly in my Feed reader. It merely started occurring earlier this week, did you change anything on the blog?