So, just how good was Damon’s steal?

As I’m sure you all have heard, Johnny Damon took advantage of the shift to advance an extra base on a steal of second last night. It was 2 outs in the ninth, and after a walk to Texeira, next batter A-Rod hit a walkoff double. As you can expect, Damon’s steal was given a lot of credit for impacting the outcome of game. However, while it was a very heads up and exciting play, it simply didn’t make that much of a difference. As Jeff Sullivan of Lookout Landing notes:

# Advancing to third base with two outs provides a boost, but a small one, as runners will generally score from second on any hit. The win probability added of the first steal was +4%, as Damon got himself into scoring position. The win probability added of the second steal was +1.6%. The steal of third did not dramatically improve the Yankees’ odds of winning.

However, that fact didn’t stop the steal from becoming mainstream media fodder.

Yes, Damon’s daring steal of second and third, after a memorable nine-pitch at-bat against Brad Lidge in the ninth inning, was as unique as it was brilliant. The bottom line was the Phillies went to sleep on the play, forgetting that third base was wide open because of the shift they had on Mark Teixeira, but it was Damon who saw the potential for how it could change the game dramatically.

Mainly he was thinking of how much pressure it would put on Lidge.

“I felt like being on third base, it possibly takes away his slider – that tough slider in the dirt,” Damon said. “Alex got two fastballs so it did work out for us.”

So while advancing on the steal didn’t actually improve the Yankees odds of winning that much, Damon argues that the added pressure it put on Lidge was significant in itself. The main justification behind that is that being on third instead of second forces pitchers to thrown fewer offspeed pitches than they normally would, for fear of throwing one in the dirt.

Is that claim verifiable? Well, let’s check it out. Using Pitch f/x data from 2007-2009, we can see that with a runner on 3rd base, pitchers threw breaking balls 29% of the time. When there was no runner on third, pitchers threw breaking balls 25% of the time. Amazingly, we see that the opposite is true! Pitchers actually throw more offspeed pitches when their is a runner on third base.

The reason for this is likely that pitchers are trying to strike out more hitters when they have a runner on third, as most contact will score a run, and the improvement in strikeout odds overrides the risk of throwing a wild pitch. For Lidge, the same pattern holds true. Over the past 3 years, with a runner on third base, he’s thrown his slider 54% of the time compared to 49% of the time in all other situations.

Now, it’s still possible that Damon on third improved A-Rod’s odds of getting a hit in that at bat; however, that would already be included in the Win Expectancy figures. So while it was a very heads up play, let’s not pretend that it had much of an effect on the outcome of the game. Damon’s single, and A-Rod’s double were much, much more important.


12 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Wooden U. Lykteneau
14 years ago

I’m sorry, is there a time when a pitcher *doesn’t* want to get a strikeout with a runner on third?

Dan Novick
14 years ago

Interesting. Damon could have done the same thing and said, “I wanted to get to third base so that I could score on a slider that gets away, since Lidge is more likely to throw a slider with a man on third.” Chances of him actually saying that? Not very likely.

RedRobot8
14 years ago

Don’t you need to look at pitch selection with a runner on 3rd and 2 out?  That was what Damon/Lidge was in, and then you wouldn’t see that strikeout effect.

Nick Steiner
14 years ago

My database actually doesn’t have the number of outs before each pitch, just after, which is kind of obnoxious.  I assume you’ve added that data to your parser, do you think you could take a look at it for me?

RedRobot8
14 years ago

Looking at my database with 2 outs, my numbers don’t seem to agree with yours, Nick.

For all pitchers:
  Man on 2nd (3rd empty): 31.7% slider/curve
  Man on 3rd (2nd empty): 30.6% slider/curve

For Lidge specifically, the effect is more pronounced:
  Man on 2nd (3rd empty): 55.0% slider
  Man on 3rd (2nd empty): 48.3% slider
Interestingly, Lidge doesn’t throw more fastballs with a man on 3rd, instead going to his changeup more.

Even so, I don’t really think that Lidge throwing his slider 48% of the time instead of 55% of the time is that much of an advantage to the hitter; it is not like Lidge totally abandons his slider.  It would be fun to narrow down further to 1-run/tie games, but I don’t have that easily available.

Nick Steiner
14 years ago

I just looked at the slider/curveball usage when a runner was on third, and when a runner was not on third.  I didn’t care about the other bases. 

Thanks for getting the data with two outs, I’ll have to find a way to query that information.

RedRobot8
14 years ago

Now, it’s still possible that Damon on third improved A-Rod’s odds of getting a hit in that at bat; however, that would already be included in the Win Expectancy figures.

That would be true unless there was some particular characteristic of Lidge (i.e., over-reliance on a pitch he struggles to control) that made him more susceptible to that particular base-out state than the normal MLB pitcher, correct?

Dave Studeman
14 years ago

I guess I’m confused by the sentence, which goes from the personal (Damon/A-Rod) to the general (Win Expectancy).

As a general rule, I think applying generic Win Expectancy tables and things like that to specific situations is great as a guideline, but I’m always hesitant to say that all the specific contingencies and personal attributes of a situation are built into the generic tables. I wasn’t sure which Nick was referring to.

Nick Steiner
14 years ago

Dave – you’re right that WE tables wouldn’t account for specific situations like this.  However, if the average player improves with a runner on third base, which he does I believe, than it’s included in the marginal gain in WE.  So, at least proportionally, it should reflect ARod’s improvement with a runner on third, unless of course there is something very specific like RR mentioned. 

Anyway, I think we can all agree that while the play was very heads up, Damon should be considered the second coming of the Mesiah.  And Keith Olbermann should probably get his head checked out.

Doug
14 years ago

Not sure if your data shows the innings and score. I would expect to see a change in pitch selection to less offspeed stuff if it’s the ninth and the game is tied. As the potential penalty for a WP or PB changes (ie. no chance to make up for it next inning), so should the behavior… at least in my mind.

Also, no idea if you’d have enough of those situations to constitute a meaningful sample size.

Dave Studeman
14 years ago

Now, it’s still possible that Damon on third improved A-Rod’s odds of getting a hit in that at bat; however, that would already be included in the Win Expectancy figures.

How is that?

Ted
14 years ago

Another question might be *where* Lidge throws his slider in that situation. He has for years tried to bury the pitch, in other words to fail low rather than up in the zone.

Even if he throws the slider more often with a man on third, there would seem to be the possibility that he leaves it up in the zone more often, for fear of a wild pitch, and thereby making it a more hittable pitch.

Just a thought…