Terry Francona: Small-ball using, run environment chameleon

Before the season there was much interest in the potential of the Cleveland Indians, some going as far as to suggest they were World Series favourites. A miserable April (7-14, -19 run differential) caused their playoff odds to drop by the largest amount in the game, but since then they have played well (64 - 58, +21) and have themselves within striking distance of the second wild card spot in the American League. Through Tuesday's games they are 5.0 games back with 4.5 percent odds of making it according to FanGraphs.

Ok, maybe they are not within striking distance so much as within small miracle distance, but they are not yet eliminated. That chance, slight as it may be, gives the perception that Cleveland's games this month, and the managerial decisions within them, are of greater importance. We know this is not the case, as a win now is worth the same as a win in April, but the perception of there being an inequality remains. Thus, Terry Francona needs to be at his best right now. Push the right buttons and pull the right levers, or his team will be going home. Here I look at how Terry Francona has changed as a manager, specifically looking at his newfound propensity for using small ball strategies.

I am focusing on Francona because during Tuesday night's game against the Royals Francona made a questionable decision. Trailing 2-0 in the ninth, the first two Cleveland batters reached base against Royals closer Greg Holland. Holland, typically an elite reliever who throws his fastball in the mid-90s, was throwing only 87-89 mph and appeared to be struggling. Holland's velocity chart over the season, which has been flying around Twitter the past few days and was written about by Mike Petriello, suggests that he may be hurt. In any case, Francona's boys had just added 22.6 percent of win expectancy (WE) and seemed to have a chance to knock off the Royals with Holland not at his best. Things were still in the Royals' favour (WE: 69.6%) but not to the extent they were seven pitches prior.

Then Francona had Mike Aviles (68 wRC+) attempt a sacrifice bunt. Aviles popped up to Holland, who misplayed it, but because of the popup the runners had to freeze. Holland was still able to get the lead runner; it was an odd play. Cleveland essentially gave the Royals (and their struggling closer) an out, dropping the WE down 12.2 percent. Two batters later the game ended in a Cleveland loss.

Mike Aviles not getting the bunt down is a problem, but even before that, Francona's decision to bunt is questionable. According the 2015 run expectancy table at Baseball Prospectus, with runners on first and second and nobody out the run expectancy (RE) is 1.44. With runners on second and third and one out (the likely situation following a sacrifice bunt) the RE is 1.29. Even if the move worked as intended, Francona's decision could cost his team runs. On Twitter, David Schoenfield, editor of the ESPN Sweet Spot Blog, wondered why Francona has become so sacrifice bunt-happy since moving to Cleveland.

Yep, been meaning to write on that. Never bunted in Boston, but has led the AL in sacrifices past two years. https://t.co/BztPFklIeB

— David Schoenfield (@dschoenfield) September 16, 2015

The answer is that Francona appears to be modulating his sacrifice-ness to the run environment and the offensive capabilities of his team.

Consider the following table of Francona's 11 seasons managing in the American League (AL):

Note: I have included data only for the AL because that is a more relevant comparison group for Francona, and pitchers hitting in the National League affects the prevalence of sacrifices.

In high run environments (and when his team is better than league average in run scoring) Francona does not call for many sacrifice bunts. He routinely ranked at the bottom of the AL when managing teams in Boston that were offensive monsters. From 2004 to 2011 the Red Sox were pretty consistently on par offensively with the 2015 Toronto Blue Jays (5.5 R/G), an offense by whom we are all currently mesmerized. As the league's run environment declined to the much lower level it has been during the last three seasons, and his team's relation to it went in the wrong direction, Francona has been more willing to sacrifice. This relationship is clearly demonstrated in the figure below, which plots the number of sacrifice bunt attempts per 500 plate appearances as a function of the difference between team and league R/G:

The Cleveland offenses of the last three seasons were much less capable of scoring than were Francona's Boston squads. Francona has responded by adjusting his managing.

The adjustment is astute. While sabermetric research has shown that small ball tactics are generally ineffective for run production, which has led to a decline in the prevalence of things like sacrifice bunts, they make more sense in low run environments because the value of an out is different. This is made clearer by returning to run expectancy tables. Above I showed that within the 2015 major league baseball run environment (4.25 R/G) a successful sacrifice bunt with runners on first and second and nobody out reduces RE by 0.15 runs^*. If we consider the same situation in 2004 (a high run environment) the reduction in RE would be 0.18 runs (1.54 to 1.36). In 2009 (a run environment in between 2004 and 2015) it would be 0.16 runs (1.49 to 1.33). So while the decision is always costly in terms of RE, though the difference is small, it is less so now because of the lower run environment.

From the data it appears as though Francona is aware of this relationship, which could be driving the increased frequency with which his Cleveland teams attempt sacrifice bunts. Considering that the Cleveland offense is a fair amount below league average this season, the decision on Tuesday night is perhaps even more reasonable.

I do not mean this as a defense of Francona's decision to attempt a sacrifice on Tuesday night. There were a number of other factors at play, like Holland showing signs that he was not at his best. Perhaps, given this information, swinging away was the much better play, even for the weak Cleveland offense. Regardless, as noted by other baseball writing analysts, Francona's willingness to implement such strategies over the last few seasons stands in stark contrast with how he managed previously. This change appears to be driven by the context of the league and his team's offense. Good for him for knowing that the relative efficacy of strategies changes with the run environment.

--

* Note: the RE values given in the article are not AL-specific. They include data from both leagues, so the cited differences could be slightly larger if only the AL was considered, as it typically has a higher run environment. Also, the numbers are inning-neutral, meaning they are averaged across all innings. Things change for situations in late and close games, as was the situation in the Indians-Royals game that led me to this discussion.

. . .

Chris Teeter is a featured writer and editor at Beyond the Box Score. He is also a contributor at BP Boston. You can follow him on Twitter at @c_mcgeets.