Forget (Some of) What You Know About Runners on Third

I’ll spare you the description of how I came up with the idea for this article. There was a lot of Alex Rodriguez’s announcing involved, and this is a family website, so my opinions on that will remain undiscussed. The point is, though, that it made me wonder about something I used to take for granted but have increasingly questioned: how do pitchers change their game plan with a runner on third base?

Depending on who you talk to, it might matter a lot or a little. Maybe pitchers won’t be willing to bounce one. Maybe they’ll pitch to a strikeout (assuming fewer than two outs), trying to keep the run from scoring. Maybe pitchers will completely ignore the runner on third and pitch normally. I’m legitimately uncertain. Not I think it’s 50% likely to be one and 50% likely to be the other — I have absolutely no idea how to weigh the relevant probabilities.

First things first; what about those bounced pitches? This is a classic announcer trope, but it’s a trope for a reason; throwing a pitch in the dirt really is more dangerous with a runner on third. Through the magic of run expectancy tables, we can see how much a one-base advancement costs the fielding team, based on whether there’s a runner on first, second, or third (I ignored other base/out states for brevity’s sake):

Change in Expected Runs After WP/PB
Runner On 0 Out 1 Out 2 Out
1 0.21 0.15 0.10
2 0.22 0.24 0.05
3 0.18 0.35 0.72

With no one out, everything is more or less the same; that runner on third was pretty likely to score anyway, in fact. As the outs pile up, allowing the runner from third to score hurts more and more — quite logical.

With that out of the way, here’s a factoid: In 2020 and 2021 combined, pitchers have bounced 6.7% of their pitches with runners on base. I’m not looking at passed balls or wild pitches specifically — just any pitch that hit the dirt, whether the batter swung or not. That gives us a good baseline for how often we should expect a bounced pitch before considering context. Next, let’s consider context:

Bounced Pitch% by Base/Out State
Runner On 0 Out 1 Out 2 Out
1 5.4% 6.5% 8.4%
2 6.0% 7.3% 8.7%
3 6.6% 7.9% 7.5%

Pitchers ease up ever so slightly on pitches in the dirt when there’s a runner on third and two outs, but only ever so slightly; they’re still bouncing pitches far more than league average. Why is that league average lower than all of these situations? Strikeouts are more important with runners on base, and pitchers hunting strikeouts are more likely to bounce pitches.

So far, so good. Pitchers are aggressive about throwing risky but rewarding bounced pitches, particularly with runners on. They pull back a hair when a mistake would be most costly, but generally speaking, pitchers don’t seem to care about balls in the dirt. We know — know for a fact — that they’re capable of being more cautious. The data in other situations bears that out. The desire to avoid contact still wins out.

Okay, so we know how pitchers behave in aggregate. What about when the pressure is on, though? I looked at pitches thrown with a runner on third when the game was in the seventh or later, and that run would either tie the game or put the pitching team behind. As we’re getting into small sample sizes here, I bulked up the data to include 2015-21.

With no one out, pitchers hit the ground 6.8% of the time. With one out, they risked a wild pitch more often; 7.8% of the time, to be precise. With two outs, they eased off — only 6.8% of the time again. But it’s just not that big of a difference. With runners on base in close and late situations, pitchers bounced 6.5% of their pitches in aggregate. It certainly appears that they’re mostly trying to avoid contact, conceding the chances of a wild pitch or walk in exchange for a better chance at a strikeout.

Okay, so pitchers mostly go for strikeouts with runners on base, never mind the consequences. Do they succeed? Eh, kind of. Here’s that same grid, only with strikeout rates (2020-21):

Strikeout Rate by Base/Out State
Runner On 0 Out 1 Out 2 Out
1 29.8% 21.5% 23.6%
2 20.8% 22.9% 24.6%
3 23.1% 21.9% 23.0%

Wait, what the heck? Where are all these strikeouts that are supposed to accompany the breaking balls in the dirt? The pitchers have been robbed! Someone start an investigation!

Actually, the pitchers haven’t been robbed. It takes two to tango, and hitters are doing their best to avoid leaving runs on the table. The overall first-pitch swing rate hovers around 29%, but batters swing far more often with runners on, trying to avoid deep counts and the possibility of a strikeout:

First-Pitch Swing% by Base/Out State
Runner On 0 Out 1 Out 2 Out
1 29.8% 30.8% 31.6%
2 33.4% 31.5% 29.1%
3 39.8% 38.9% 32.3%

So, what happens when a runner reaches third? Pitchers get aggressive — particularly if there are fewer than two outs and a strikeout is the only clean way out of a jam. Hitters get aggressive right back; they don’t sit and wait to get into a deep count, but start trying to put the ball in play immediately.

There are levels upon levels. Pitchers know that hitters come up to the plate swinging. For all first pitches in aggregate, pitchers throw sinkers or four-seamers 55% of the time. With a runner on third and less than two out, that number falls to 42.2%. But hitters know that pitchers do that — again, it’s hardly rocket science. So their increased swing rate is at worse pitches in general, and they’re still fine with that; in this era of increasingly rare contact, putting something into play, anything at all, is an increasingly difficult goal.

So, what did I learn about runners on third in my announcer-induced deep dive? In one sense, not so much. Pitchers don’t alter their behavior all that much, and if they do, they’re more willing to bounce pitches rather than less. Hitters don’t strike out more. Things look relatively static, strangely enough.

In another sense, there’s a ton to learn. Both pitchers and hitters are strongly in tune with the changing incentives as runners move around the basepaths. They alter their approaches to maximize those incentives. And the two sides cancel out, when it comes to strikeouts at least. Both sides change their behavior significantly, and the push and pull evens out nearly perfectly — give or take some balls in the dirt — so that despite wildly different incentives, strikeout rates with a runner on third don’t budge much in either direction. What a wonderful coincidence.

Ben is a writer at FanGraphs. He can be found on Twitter @_Ben_Clemens.

Newest Most Voted
Inline Feedbacks
View all comments
1 year ago

Interesting! I have to watch A-Rod on mute. Some of the stuff he spouts is just mind-numbingly bad.

1 year ago
Reply to  kylerkelton

I don’t think anything can top his “you always want even leads, versus odd leads” take, but there are many worthy candidates for second place.

1 year ago
Reply to  HowBoutDemOs

Nonsensical analysis is everywhere – not just ARod.

1 year ago
Reply to  RonnieDobbs

That’s true, but the national broadcasts are noticeably worse to me than the local ones. I think *most* teams broadcasts have shown improvement over the past few seasons.

1 year ago
Reply to  kylerkelton

With the merciful retirement of Hawk Harrelson, are there any really bad local broadcasts left? I’d rather listen to the other team’s crew than a national one basically every time

1 year ago
Reply to  kylerkelton

I have to mute most crews. More accurately, I have to mute every traveling crew. ESPN is terrible in general.

Jason Bmember
1 year ago
Reply to  RonnieDobbs

I miss 50% of the old Morgan and Miller crew…

1 year ago
Reply to  Jason B

Fortunately Jon Miller does the Giants radio broadcast. We’re blessed

1 year ago
Reply to  kylerkelton

If no statcast I watch it in Spanish. Better home run calls for sure. It is weird because he had some decent post game analysis on fox back in the day, but his announcing is just godawful. There isn’t even a close second worst. He is far and away the worst I’ve ever heard.

1 year ago
Reply to  carter

Worse than Joe Buck?