Was Castro’s Mild Dash Worth It?

Photo courtesy of David Banks-USA TODAY Sports

My wife and I were driving to Chicago last Friday to help our younger daughter move from one Lakeview apartment to another. It really wasn’t the ideal time to move, with the Cubs at home, but a confluence of events made these days the only ones that would work. It would be fine—after all, the game had a 1:05 start, and we weren’t planning on getting to Chicago until around 6:00 pm, so the game would be over by then. What could go wrong?

I believe I was on I-355 when Hector Rondon gave up a two-out (and two-strike) double to Josh Harrison, who would come around to score the tying run. As a result, the game ended almost literally the moment we pulled up to our daughter’s place. As we were stuck on the Kennedy coming in, I heard the play in the bottom of the tenth when Starlin Castro attempted to score on what Pat Hughes called a “short fly ball” to right. Not being able to see the play, I had to rely on Pat’s call, and he made it sound like Castro was out by a mile. The video suggests this was the case.

Later in the evening, as we were walking around Wrigley, I saw this tweet from Len Kasper:

I'm fine w/10th-inn send. OK to get thrown out on 1-out SF attempt. Pete Palmer says its statistically acceptable & not enough teams try

— Len Kasper (@LenKasper) May 16, 2015

Len followed that with:

Palmer said 90% success rate on SF's is way too high. 1-out try almost always makes sense bc of 2-out run probability (which is LOW).

— Len Kasper (@LenKasper) May 16, 2015

I knew something was nagging at me when I read it, but I couldn’t put my finger on it. Plus, my computer and my data were at home, so there wasn’t anything I could do about it.

Saturday morning, as we got smart and hired some nice young men to move the heavy stuff, I realized what had been bugging me. Recall last year’s World Series when Alex Gordon hit a single that was misplayed and allowed him to go all the way to third. At that time, Tom Tango, one of the foremost sabermetricians in baseball today (and a huge hockey fan also) tweeted out suggesting that Gordon might have been justified in going for it, since the odds suggested the Royals only had a 27-percent chance of scoring with a man on third and two outs (second chart). In other words, consider the situation—in the case of the Royals, they had Gordon on third with one out to bring him in, and the historical odds weren’t in their favor.

I sent this tweet to Len, which Tango added to:

@ScottLindholm @LenKasper Pretty close. 27% on average (see middle set of charts) http://t.co/m0O9glphID

— Tangotiger (@tangotiger) May 16, 2015

This is the crux of the matter. Matt Szczur flied out, leaving the Cubs with that same 27-percent chance of scoring a run with that last out. It was already an extra-inning game, both the bench and the bullpens were being taxed, and Joe Maddon saw an opportunity to end the game. Four things had to go right for the Pirates to complete the double play—Gregory Polanco had to field the ball cleanly, make an accurate throw to home, and Francisco Cervelli would have to field the throw cleanly and apply the tag. The entire play from the moment Polanco caught the ball took around 3.3 seconds, and it takes at least four seconds to safely score on a fly ball. Joe Maddon took a chance, and it didn’t work out.

Through Saturday’s games, there have been 426 fly balls hit with runners on third and less than two outs, of which 362 were not hits, and as such potential sacrifice fly candidates. This table summarizes how those balls turned out:

It’s pretty clear that if a fly ball makes it to the outfield, chances are very good the runner can score from third. Daren Willman’s Baseball Savant site shows Szczur’s fly ball only traveled 227 feet, which is a relatively easy throw for a right fielder with even an average arm.

The key to understanding Maddon’s thought process is realizing that sending Castro was probably the best chance the Cubs had to score in that inning. If it was the top of the first, perhaps the thinking would have changed, but the chances of Castro scoring only have to be higher than 27 percent in order to justify the chance.

One way to describe the Cubs this year is aggressive, particularly on the basepaths. They’re third in all of baseball in stolen bases, and are also among the league “leaders” in being thrown out on the basepaths. There will be times when such aggression won’t pay off, but it is a function of the increased number of baserunners the Cubs are getting this year.

Using my superior 20/20 hindsight, would I have held Castro? Keep in mind the next batter up was the pitcher slot, so sometimes it pays to try to take advantage of the what is, instead of waiting for what if. This is what Maddon did, and it didn’t work. It didn’t make it the wrong move, merely one that didn’t work. It’s not like Maddon did something crazy like attempting a 47-yard field goal on second down in overtime and watching it miss—that would be insane. Who would do anything like that?