Best of BP 2025: The Spider Problem

Let’s see if we can catch some flies.

This is the percentage of “air balls”—both fly balls and line drives—that made it to the outfield and were then caught by an outfielder. Something happened in the past few years and it’s been sapping the offense out of the game, turning what used to be singles and doubles into outs and at rates that are going to leave a mark. While the infield shift got all the press for how teams were turning pulled ground balls into outs, it was the outfield defense that was really tamping down on hits. It’s been enough that serious thinkers at MLB have played around with rule change ideas, mostly to make outfielders play closer to the infield, to encourage more balls to fall in behind them.

We have public positioning data for outfielders back to 2016, and from 2016 to 2024, center fielders moved back an average of seven feet (315 to 322 feet from home plate) while the corner outfielders pushed back three feet each. Some of that is simple calculus. There probably are balls that are landing in front of the outfielders that used to be catchable, but there are probably now balls that used to land behind the outfielders which are now outs. It’s a bigger problem to have the ball land behind you, because it’s deeper in the park and to retrieve it, you have to go further back, turn around, and make a longer throw. If you’re going to give something up, it might as well be the type of hit where you are charging the ball.

But the problem is more than just where the fielders are playing. A greater proportion of fly balls are being caught. Depending on where you want to benchmark it from, we’re talking about a two-percent increase (or, if you prefer, 20 points of batting average) over the course of a couple of years. What happened?

Why are so many flies ending up in the web of someone’s glove?

Warning! Gory Mathematical Details Ahead!

We can rule a couple of things out. One is that ballparks haven’t gotten appreciably smaller. Truist Park in Atlanta opened in 2017 and Globe Life Park in Arlington opened in 2020, so it’s not like there’s been a run on small new stadia the way that there was in the 1990s and 2000s. And while there have been a couple of teams who have fiddled with their walls, we can pretty much rule that out as a reason.

Maybe the ball is staying in the air longer. We know that over the last decade, the ball has “behaved differently.” Starting in 2015, it started flying further seemingly due to changes in the manufacturing of the ball. (MLB denies it was intentional.) But in 2018, the ball theoretically went back to normal, though the home run surge didn’t stop. Maybe hitters were putting the ball in the air at higher angles, which would mean a lot of lazy flies? If the ball hangs in the air for six or seven seconds, it’s a lot easier to get under it and catch it.

I used a very simple formula for hangtime using the launch angle and initial velocity (thanks 12th grade physics!). Granted, it doesn’t account for the Magnus force that keeps the ball aloft a little longer due to the way that the ball spins, but should do well to show the general trend. Hangtime dropped once MLB “fixed” the ball, and yet rates of catching the ball have gone up, and it’s pretty much gone sideways since.

Perhaps the fielders are faster now? There’s some evidence for that. We have sprint speed numbers from when outfielders run the bases, and here’s the chart for center fielders (the right and left field charts have the same basic shape.)

Sprinting around the bases isn’t the same thing as field speed, but it’s the best we have, since MLB doesn’t publicly publish field speed data. They do publish, for each fielder, how many more (or fewer) feet per fly ball each player covers compared to an average outfielder given the same amount of time, which is close to, but not quite, the same thing. That graph tells the same story. The sprint speed data tell us that the average center fielder has gotten faster in the past few years.

Then there’s this graph. MLB publishes data on catch probability. They don’t do it at the play-by-play level (*ahem*) but for the years 2016 to the present, you can get team-level data for each year and position and also individual player data. The individual player data may be confounded by the fact that some players play in different positions and it doesn’t say how many opportunities each player had, but it does give each of them an average “difficulty” level. The difficulty levels appear to be for only fly balls, as the percentages are far too high to include line drives. Catch probabilities are based on a ball’s hangtime and the distance that the fielder has to cover. Neither data set is perfectly set up to give a league-wide perspective, but here’s the graph from the individual-level data set:

The team-level data look the same (and if you really want to see the chart, it’s here), and the story is pretty obvious. Outfielders have a greater average catch probability. We’ve seen that hangtime hasn’t budged much. It means that outfielders are closer to the balls that they need to catch. The only way that can happen is that they are positioning themselves better before the pitch is made. It’s not clear whether that’s the result of smarter fielders or better index cards telling them where to stand, but either way, baseball has a Spider problem.

MLB also publishes its Outs Above Average (OAA) metric and calculates it – in yearly form – for each outfielder. It’s only calculated for balls where they are trying to catch balls that are not considered “gimmes” (i.e., they have a catch percentage below 90 percent.) More than that, they break down the individual components of the out. How much ground did a fielder cover in the first 1.5 seconds after the ball is hit (jump) and then in the 1.5 seconds after that (burst)? How close was their route to the mathematically ideal one? It doesn’t matter if you’re fast if you don’t know where to run. Problematically, MLB requires a minimum number of chances to appear on that leaderboard, but even with that limitation in mind, we can still see what the numbers tell us.

Because MLB seems to base catch probability on the combined corpus of the fly balls in its databases (everything since 2016), we see that the league-wide numbers don’t average to zero in each year. What we do see is this table, which is denominated in linear feet above average (i.e., Smith ran an average of 4 feet further than the average outfielder on a typical fly ball):

Year
Jump
Burst
Route

2016
-0.04
-0.09
-0.13

2017
-0.07
-0.03
-0.08

2018
+0.01
-0.05
-0.06

2019
-0.04
-0.02
-0.06

2020*
+0.01
+0.02
+0.06

2021
+0.07
+0.04
-0.06

2022
-0.00
-0.03
+0.02

2023
+0.00
-0.04
-0.06

2024
+0.24
+0.12
+0.25

Outfielders are getting better at getting a good read and a good jump on the ball and are covering more ground when they run to the ball and are taking better routes. Each of those has only grown an average of a couple of inches over the past decade, but put them together and now you’re talking about something close to an extra foot of motion. It doesn’t sound like much until you do a little geometry. Factoring in reaction time (when the brain has to recognize that the ball is coming toward me, I should move!), and acceleration, the average MLB runner has an effective velocity of around 20 mph. At peak velocity they run faster, but you have to include the time when they are literally standing still after the ball is hit. On a four second fly ball, that gives them an average of 80 feet of range in any direction. (Note: It’s harder to go backward than forward or sideways, but assume the spherical cow for a moment.)  Increasing the radius of the circle from 80 to 81 feet means increasing the area of the circle by about 500 square feet. Three outfielders, similarly upgraded, will cover 1,500 more square feet than previously. In an outfield that has an area of about 120,000 square feet, that’s an extra 1.25 percent of the outfield that you can cover.

That 1.25-percent coverage isn’t enough to explain the full increase in fly ball catching that we’ve seen over the past decade, so positioning is likely also involved. It does, given a very humble amount of ground distance, tell us though that if players—either by their own smarts or because their defensive coordinator told them where to stand—can be roughly one step, which is somewhere around 2-3 feet, closer to the ball than they had been in the past, it can have a huge impact on game outcomes.

The margins are pretty thin in baseball. A one-percent move in catch percentage is also known as 10 points of batting average, at least on those balls. Depending on when you want to benchmark it, we’ve seen batting average on fly balls drop by a couple of percentage points, and there are roughly 10 outfield “air balls” during the course of a game. Take into account that the swing in run value from changing a hit into an out is fairly large (roughly three-quarters of a run if it was a single being taken away, but obviously more for doubles and triples, which have also been affected) and you’ve got an effect that is suddenly very valuable. I’ve previously estimated that the increase in outfielder defensive efficiency is worth around two-thirds of a run per game, dwarfing just about any other single strategic change in the game in the past couple of decades.

Outfielders are getting too good at catching flies. They play deeper. They’re a little faster than they used to be. They are better able to anticipate where the ball is going. They get better jumps and run better routes. MLB could—in the sense that they theoretically have the power to do so—legislate where the outfielders stand. But they can’t tell the outfielders to be worse at their job, and the trend line suggests that they’re only going to get better at those jobs.

The Spider Problem is baseball’s silent problem. It might be its biggest one as well.