How Productive Were Those Outs? Team Edition

Earlier this week, I threw some numbers together on the value of productive outs. I focused on Corbin Carroll, and rightly so: His electric skill set is a perfect entry point for explaining how hitters can add (or subtract) value relative to average even when making an out. Putting the ball in play? We love it. Avoiding double plays? We love that too. The Diamondbacks are a team full of speedsters, and Carroll’s productive outs gave their baserunners a chance to show off their wheels.

A quick refresher: I calculated the difference in run scoring expectation between the average out and a specific type of out (strikeout, air out, non-GIDP groundout, double play) for each base/out state. Then I had a computer program tag each out made in 2024 with that difference. For example, the average out made with a runner on second and no outs cost teams 0.35 runs of scoring expectation in 2024. Groundouts in that situation only cost 0.25 runs, a difference of 0.1 runs.

Thus, on every groundout that occurred with a runner on second and no out, I had the computer note ‘plus 0.1’ for the “productive out” value. A strikeout in that situation, on the other hand, lowered scoring expectancy by 0.43 runs, a difference from average of -.09 runs. So the computer noted ‘minus 0.09’ for every strikeout with a runner on second and no out. Do this for every combination of base/out state and out type, add it all up, and you can work out the total value of a player’s productive outs.

I say player, but that’s not the only way to slice the data. With a little sleight of hand (read: a line or two of code), I re-categorized the data by team instead of player. I did this for one main reason: I wanted to see how important this seems to be. When you look at a list of individual players, it’s hard to understand how that relates to the bigger picture. Maybe Carroll is at the top of the list because he had all the best opportunities to make “good” outs. Maybe the guy who batted after Carroll racked up the double plays and undid all the positive value. One guy might not be representative of the whole, in other words.

Carroll’s outs alone were worth 8.5 more runs to the Diamondbacks than you’d expect if you treated all outs the same. How about all the outs that Arizona recorded, from everyone? We’re talking speedsters with groundball tendencies like Carroll, boom-or-bust sluggers like Eugenio Suárez, mashers like Christian Walker and part-time backup catchers like Tucker Barnhart. Do you have a guess?

The answer is 12.3 runs. Less than you expected? It’s certainly less than I thought it’d be. There are so many great baserunners on that team! But, well, that’s not exactly what this statistic is about. Baserunning gets accounted for based on what actually happens – extra advances and avoiding outs on the basepaths or so on. Here, we’re assigning credit only to the hitter based on the average result of a fly out or what have you, so the great baserunning never touches this statistic, which I’m calling OAR (out advancement runs) in my spreadsheets to avoid having to come up with some new word soup name for it every time I mention it.

That’s not to say that the two aren’t correlated. There’s a 0.26 r-squared between team-level baserunning and team-level OAR. This isn’t about counting the same play twice, but it is about complementary skills. Carroll is a great example – his productive out exploits don’t add anything to his baserunning score by definition, but blazing speed helps him excel at both. It’s not a 1:1 situation. Fast players who strike out frequently (Luis Robert Jr., for example) often end up with positive baserunning value but negative OAR. Juan Soto is no one’s idea of fast, but he puts the ball in the air and doesn’t strike out, so he has negative baserunning value and positive productive out value. In general, though, great baserunners tend to make more productive outs.

The Diamondbacks weren’t the best team in the league by this measure, by the way. They were third, though the top four all finished between 11.1 and 13.2 runs of value. First place? That’d be the Detroit Tigers. They were the platonic ideal here. They put the ball in play quite frequently when they had RBI opportunities, avoided double plays, and while their strikeout rate was high overall, it was meaningfully lower with runners on and fewer than two outs. I’m not sure if that’s a skill, but it certainly could be; players absolutely change their approach based on the situation, and whether the Tigers were doing it on purpose or not, they got far more contact-happy when the situation called for it.

For what it’s worth, the Tigers were also an excellent baserunning team, at least if you exclude steals. They were second in non-steal baserunning value and 24th in value added via the stolen base. On the other hand, the Orioles finished second in team-wide OAR despite middling baserunning. They simply never hit into double plays – their 71 double plays was the lowest in the majors by a mile (15 double plays).

The other side of the coin is also interesting. Yankees fans, you can admit it: You think your team finished last. Aaron Judge was last on the individual leaderboard, after all, and he was hardly the only Yankee to have a double play problem; as a whole, the team hit into 138 of them last year. But the Yankees aren’t quite as woeful as you’d expect; they finished 27th in the majors with -9.5 out advancement runs. That’s bad, but not that bad. Judge himself finished with -8.8.

That’s for a few reasons. For one, the Yankees were quite good at avoiding strikeouts in high-leverage situations. With runners on base and fewer than two outs, they only struck out 19.1% of the time. That’s almost even with Arizona’s mark, and it helps explain all the double plays: For better or worse, they put the ball in play. Those double plays are misleading, too. The Yankees had more opportunities to hit into a double play than any other team in baseball because they had runners on base so frequently. Their rate of hitting into double plays was halfway between average and league-worst; in other words, there’s a rate statistic versus counting statistic mismatch going on here that makes them look terrible instead of merely bad.

No, the worst team in baseball when it comes to making productive outs was the Colorado Rockies. And they were the worst by a ton:

Ew. You know how I mentioned that the Yankees’ double plays didn’t look so bad on a rate basis? Colorado’s looked awful. The Rockies hit into 126 double plays, fourth most in baseball. And they did it in the fifth-fewest opportunities. It’s honestly hard to fathom; despite playing in an offensive paradise, they struggled mightily to put runners on base. When they did manage, they hit into double plays at a league-leading rate. Oh yeah, they also struck out at the fifth-highest rate in the league when strikeouts are at their worst (runners on, fewer than two outs). Do you know how hard it is to hit that many double plays when you’re also striking out all the time? It’s incredibly hard.

The Rockies weren’t completely hapless baserunners. They were one of the worst teams in the league at stealing bases, but that’s a different thing; when it comes to advancing and avoiding outs, they were roughly average. It simply didn’t translate to the plate, where their outs were as unproductive as you can imagine – and potentially more than you can imagine. Twenty runs is a ton!

The next question writes itself: How are productive outs, as measured by OAR, correlated with actual run scoring? To test that, I started with weighted runs created. I didn’t adjust for stadium, because runs scored don’t adjust for stadium either. Then I added baserunning runs, as measured by Statcast’s measures of baserunning and base stealing. That gave me an “expected” runs scored, measured by context-neutral offense and baserunning prowess.

From there, I compared expected runs scored to actual runs scored. Then I regressed the residuals against my measure of runs from productive outs. Sadly, the correlation wasn’t huge; we’re talking about an r-squared of about 0.075. The best way to think about that if you’re not knee-deep in a statistics textbook is that roughly 7.5% of the variation in “unexplained” run scoring (as in, the difference between actual runs scored and the sum of wRC and baserunning) can be explained by productive out value. In other words, very little of teams’ over- or under-performance on offense is explained by how productive their outs are.

That might be a disappointing result, but as I look into it further, it makes sense to me. The Rockies actually scored more runs than you’d expect from their component numbers, despite their terrible OAR. That’s because they hit better with runners in scoring position than without. They hit even better with runners in scoring position and two outs. Productive outs are a lot less meaningful than reaching base.

That raises a great question about my accounting for productive outs: Are they predictable? There’s plenty of research showing that hitting better or worse with runners in scoring position isn’t sticky; your split in one year doesn’t have much to say about your split in the next year. If the residual value of outs is similarly noisy, it’s more of a curiosity than a building block statistic, good for telling a story, but bad for predicting how the future will go.

My answer? Check back later. The way I built my query in this iteration would require a good amount of tinkering to generalize to past years. For those who don’t care about the specifics, you can skip the rest of this paragraph. For those who do, I wrote one script that calculated the values for each pair of base/out state and out type, then manually entered those into a second script that went through game logs and assigned a value to every out. To run past years, I’d either have to do a ton of manual work or rewrite the script in a way that calculates weights for several years, then applies those to a multi-year dataset. It’s doable, just not in my current setup, and I’m pretty slow at working in SQL still.

But while that might be unsatisfying, I don’t think that the overall conclusion is anything short of fascinating. It’s like this: Better baserunning teams also tend to get more value out of their outs than the league as a whole. They also demonstrate a weak tendency to score more runs than you’d expect based on their hitting and baserunning. The effect isn’t enormous, but it’s clearly real. Putting the ball in play and beating out double plays really does count on the scoreboard – just not all that much.