The Worst Baserunning Play of the Year
Earlier in the week, we talked about Torii Hunter moving to right field, and it turns out that he was totally on board with the decision. He made a selfless choice for the betterment of the team, and should be lauded for his self awareness.
Last night, though, awareness is exactly what Hunter did not have. If you haven’t seen it yet, watch the highlight. The shock of the announcers will key you in that this was bad, but it’s even worse than their reaction may suggest.
The situation – top the 9th inning, Angels trail 9-7, nobody out, and Hunter is on second base. Howie Kendrick is at the plate, representing the tying run.
At least, he would have been before Hunter was thrown out trying to steal third.
The Angels, previously down 9-1, had rallied back and given themselves a 17.5 percent chance of winning the game. When Hunter was thrown out, that dropped to 4.1 percent. It was a death blow to the rally, and to their chances of winning.
The play was all downside. There is almost no benefit from advancing to third base in that scenario. If Hunter had been successful, he would have pushed the Angels WPA up by just over half of one percent. Whether he was on second or third was, essentially, immaterial.
When you’re risking a 13.4 percent loss in win probability and the potential reward for success is .6 percent of win probability, the breakeven rate is off the charts. Hunter would have had to successfully steal third 22.5 times to create enough positive change in win expectancy to outweigh the loss of one unsuccessful attempt. Even when you factor in the possibility of an error that would have allowed him to score on the play, you’re looking at a break-even rate of nearly 95 percent.
His odds of success were obviously not that high. For a runner to have that kind of expectation of making it to third safely, the catcher would have to be Venus De Milo. It was an unbelievably bad play, and Hunter knew it:
“That was stupid,” Hunter said softly. “That was so stupid. Can’t take it back, killed the rally, terrible. They teach you that in Little League — don’t make the first out at third. [It] might have been the dumbest thing I’ve done in years.”
Yes, Torii, it was.
Dave is the Managing Editor of FanGraphs.
I can’t even see how there is any positive expectation in stealing third. When can it help?
My guess is that it’s from removing a force option at 3B if the batter reaches 1B without the runner advancing from 2B. And I think Dave’s numbers show why that’s rarely important.
If Hunter stays at second, and then an Angels batter walks, the Orioles would have had a force at any base but home. It makes the double play more likely with runners at 1st/2nd than runners at 1st/3rd.
Additionally, if Wieters had airmailed the throw into left, and Hunter had scored, that would have eliminated the chance of a double play entirely.
But, these are really small, marginal gains. They’re barley worth anything.
Dave — I think Danny was asking when stealing 3rd base can help (what situations would warrant an attempted steal of 3rd, statistically speaking, if any).
I think Dave answered that question in the post…
“Hunter would have had to successfully steal third 225 times to create enough positive change in win expectancy to outweigh the loss of one unsuccessful attempt. Even when you factor in the possibility of an error that would have allowed him to score on the play, you’re looking at a break-even rate of nearly 95 percent.”
I don’t think anyone has that kind of expectation for stealing third, unless they have a perfect read on the pitcher and are stealing signs or something (ie, the pitcher is about to bury a curve or splitter in the dirt or something).
Dave responded to that specific situation — down 9-7, man on 2nd, 0 outs, stealing 3rd.
Again, I believe Denny was asking when stealing 3rd is a worthwhile option, if ever, statistically speaking. I’m not stealthy enough to figure that out on my own, so I will lean on those with the appropriate knowledge and skills.
Stealing 3rd has quite a bit of value if a team is down by a run and only has one out. Gets you into a situation where a run can score on a fly ball, rather than relying on a base hit. Odds of winning increase from 20 percent to 29 percent if you go from runner on second to runner on third with 1 out in the 9th inning when trailing by a run.
Perfect Dave, that’s what I was looking for.
One more question — what is the decrease in said situation if he is caught stealing?
Just as an aside, since we’re talking about the value of stealing 3rd and the Orioles happen to be kind of related:
It also pays to steal 3rd when you’re Brian Roberts. Since 2005, he’s 60-65 when stealing 3rd, good for 92.3 percent. Last year he went 14-14 on 3rd base SB attempts, while going just 16-23 trying to steal 2nd. He’s talked many times about how he finds it much easier to steal third, and apparently that’s more than just talk.
Also, it’s worth pointing out that it’s a fairly useful skill for a guy who’s averaged 46 doubles over the past six seasons.
Using Tango’s WE numbers (that are drawn from Markov chains, not historical data), the gain in stealing 3rd in bot 9, down 1, 1 out is +13%. Getting thrown out is -26%, which translates to a 74% breakeven point