Imagine you are Field Marshal Terry Francona, lined up for battle with your 50 divisions behind you. You and your troops have fought well, having just defeated skilled armies from Boston and Toronto. But your nemesis now is Generalissimo Joe Maddon, who has 70 divisions to throw at you. Picture these two armies fighting over seven separate battlefields—first to seize four fields wins. What’s an underdog to do?
Suppose Maddon puts 10 divisions into position for each battle. Francona could mirror his opponent and evenly spread his forces, but he would be outnumbered all along his front. Or, he could do what outnumbered commanders have done for a long time: concentrate his forces selectively.
In Military Power: Explaining Victory and Defeat in Modern Battle, Stephen Biddle explains how this concept works in land warfare, where creating a local preponderance of force can offset an overall disadvantage (emphasis added):
By using the initiative to concentrate forces at a point while accepting risks elsewhere, attackers can amass a large local numerical advantage at that point while yielding a much smaller disadvantage elsewhere. For example, by accepting a modest 2:3 disadvantage over 550 kilometers of a 600-kilometer frontier, an invader who concentrates on a 50-kilometer breakthrough sector can obtain more than a 4:1 advantage at that key point, even if the defender’s forces are numerically equal to the invader’s overall.
In fact, even outnumbered invaders can create a large local advantage on a chosen frontage if that frontage is narrow: an invader at an 8:10 theaterwide disadvantage can still obtain almost a 4:1 local superiority by concentrating against a 25-kilometer front while defending at a 2:3 disadvantage over the remaining 575 kilometers of frontier.
To translate this into baseball terms, a manager of an overmatched club can win a series if he has the pieces to dominate his opponent in just enough games while conceding some others. Those best players won’t be available all the time, but just enough to carry the series. This is the story of a very good 93-win team facing baseball’s consensus best club in the final series of the season.
The Cubs and Indians were a mismatch on-paper. FiveThirtyEight’s prediction model estimated the Cubs as 63 percent favorites to win the series, while CBS’ simulations gave them 59 percent and Vegas set the line at -175 in their favor. Chicago, after all, beats Cleveland in almost all aspects of the game—starting rotation depth, defense, bench options, power hitting, and 100-year curses to banish.
But Cleveland had two extremely effective relievers, in Andrew Miller and Cody Allen, and one of the league’s best starting pitchers in Corey Kluber. Francona’s challenge was to combine his strengths into a winning formula even as his opponent had many potential advantages. If he could get three wins out of Kluber and one from anybody else, the series was his. The trade-off came in knowing that his best players would become unavailable even in close games the Indians fell behind.
The logic behind this selective usage is easy. Let’s say, in a simple example, that every inning Miller pitches increases the odds of Cleveland winning by 10 percent. Over the course of two games, Francona could use him in three basic ways: primarily to help when behind, primarily to help when ahead, or without any regard for how the team is playing. In addition, he could choose to give Miller a larger number of one-inning outings or a smaller number of two-inning outings.
A strategy that concentrates Francona’s best players in winning efforts might require a certain prior chance of winning to pull the trigger on using the big lefty. For example, using him for two innings in a game to get the team’s chances from 10 percent to 30 percent would probably not be as useful to Francona, under this formula, as getting the team from a 60 percent to an 80 percent chance of winning one game. Francona went to Miller largely along this line of thought in Cleveland’s wins—Games 1, 3, and 4.
In Game 1, Miller entered in the seventh up three runs and got six outs (+.146 Win Probability Added). In Game 3, he entered in the fifth in a scoreless game and got four outs (+.120 WPA). Allen was also called upon for four outs of work (+.263 WPA). In Game 4, Miller entered in the seventh up six and got six outs (+.014 WPA). Although the decision to use Miller in the seventh was reasonable given that the Indians were only up three runs heading into the top half of the inning, when he started warming, perhaps Francona could have saved an inning of work by having someone else pitch the eighth.
In contrast, Cleveland’s losses provoked a different approach. Games 2, 5, and 6 forced Tito to concede relatively early on that he would have a tough time winning and that he might be better off conserving his most precious resources. Francona knew he had long odds in Game 2 with Trevor Bauer facing Jake Arrieta. With the Tribe down 2-0 and Bauer laboring in the top of the fourth with two outs, Zach McAllister came in to stop the bleeding instead of Miller. Francona passed on Miller again in Game 5 when he took Bauer out early, even though the Indians were only down two runs again. Game 6, meanwhile, was such a blowout early on that Cleveland’s management decisions became pretty straightforward.
The fun part is that if the Indians don't score by the 3rd it's better strategy to let Tomlin get smashed to have everyone fresh tomorrow.
— James O'Hara (@nextyeardc) November 2, 2016
And then came Game 7. Oh, man. Everything was set for the Indians to use only their three best arms in a triumphant series conclusion at home. Francona had meticulously monitored Kluber's workload in the series, keeping him to 88 and 81 pitches in Games 1 and 4, respectively. Miller and Allen were each ready to do extra duty. But the magic wasn’t there for Kluber, whose stuff was lifeless and hittable. He failed to strike out a batter for the first time in his entire career.
Worse, the invincible Miller was human for the first time all postseason, too, surrendering four hits in a relief appearance for the first time in five years. The game plan fell apart. Most ironic of all, the effect of Francona’s management approach was that he prompted a drastic change in Maddon’s own strategy in the second half of the series. Maddon, with his club on the brink of losing everything, threw Aroldis Chapman in for an eight-out save in Game 5. Chapman had that same workload in Games 1 through 4, combined.
Maddon followed it up with a Game 6 that had his closer working in the seventh inning with a 7-2 lead. Chapman worked into the ninth inning of that game, by which point the Cubs had a seven-run lead and essentially a 100 percent win probability. In back-to-back elimination games, Maddon panicked at the first sign of trouble and called on his stopper. Chapman’s heavy burden in those games no doubt adds scrutiny to Maddon’s continued aggressive use in Game 7. It cost Chicago a lead and a smooth close to the historic series. So, instead, Mike Montgomery got the final out in the 10th vs. Mike Martinez with the tying run on base.
In the end, a game of maneuver became a grueling attrition battle. Still, Francona left a blueprint for underdog teams in future World Series. Even if your club is missing key players due to injury, even if your starting rotation isn’t very deep, and even if your opponent is stacked with exciting talent, there might still be a way to combine your assets in a best-of-seven clash. All it takes is a well-rested Andrew Miller and, like Chicago had for once in its franchise history, a bit of luck.