A Brief Tangent on Arm Strength

Ceddanne Rafaela has a weak arm. He also has a strong arm.
This is an analysis I’ve wanted to do for a while. It’s not that important or complicated, and most of it is fairly obvious. But it gets at something that comes up from time to time in the various places baseball is discussed online. The conversation tends to start like this: Team A should sign Player X and move him to a new position. Inevitably, one of the first questions asked about such a plan is whether Player X has the arm strength to play that new position.
The number that gets cited to “yay” or “nay” such a follow-up is arm strength, in miles per hour. But ask any baseball fan to sit with this for a moment, and they’ll raise a concern. Arm strength, to some degree, is a function of position. A third baseman has a longer throw to make than a second baseman. A right fielder has a longer throw to make than a left fielder. This means players with better arms tend to play those positions, as we can see in this plot:

It also means the official arm strength figures reported by Baseball Savant aren’t quite transferable between positions. We can’t say whether Player X has the arm strength for a given position if he’s never played there before.
That brings us to Rafaela. Most fans think of him as the next-best center fielder in the majors after Pete Crow-Armstrong. But Rafaela also happened to play 24 games at second base last year. His published arm strength at second base in 2025 was 78.5 mph. In center field, however, his published arm strength was 93.4 mph. Since 2021, that’s the largest gap for a fielder between two positions (excluding first base):

Looking at the plot, if we assumed Rafaela’s published arm strength at second base would be his arm strength in center field, he would have one of the three weakest arms in center last year. Instead, he wound up having one of the strongest. That’s because this is the type of throw that gets an out at second base:
While this is generally what it takes to get an out from center field:
It’s worth noting the methodology behind the arm strength figures you see on Baseball Savant. Basically, it’s the average of a player’s strongest throws. From the Baseball Savant arm strength leaderboard:
Since the demands of each position grouping are different, the averages and qualifiers are different as well.
– 1B: average of top 1% of throws (minimum 100 throws to qualify);
– 2B/SS/3B: average of top 5% of throws (minimum 75 throws to qualify);
– OF: average of top 10% of throws (minimum 50 throws to qualify)
So, how much arm strength might we expect a player to gain or lose with a position change? I found all of the players with the minimum number of throws at multiple positions in a season since 2021. I then took their arm strength at each position and averaged the differences. It’s about what you’d expect. Players throw harder when they’re in right field than they do at first base:
| Pos | Gain/Loss (mph) |
|---|---|
| RF | +1.5 |
| 3B | +1.1 |
| SS | +0.8 |
| CF | +0.3 |
| LF | +0.3 |
| 2B | -2.0 |
| 1B | -4.8 |
While performing this analysis, I found that Tom Tango had done a more comprehensive piece on these positional equivalencies, looking specifically at players who moved between any two positions. Using his method gives us this nice grid:

For instance, a fielder moving from second base to center field adds about 6.5 mph on average to their arm strength figure. Rafaela gained nearly double that, going from a roughly average throwing second baseman to one of the hardest throwing outfielders in the league. But how much is all that arm strength worth?
In terms of value, arm strength is the lesser of the fielding skills compared to range. There are simply fewer opportunities to make use of a strong arm. Baseball Savant only publishes arm value components for outfielders, where arm and range are easier to separate than they are for infielders. Arm value is derived through a series of estimates and rates, comparing how often a baserunner should attempt to advance on a fielder with how often they actually do (and how often they do so safely). This is done using a distance-time model, which considers the speed of the runner, the arm of the fielder, and the distance each are from their target.
In other words, arm value isn’t just about pure strength. Preventing runners from advancing requires a variety of other skills, like getting to the ball, fielding it cleanly, and making an accurate throw. And even a great throw is dependent on another fielder receiving the ball and applying the tag. Plus, a runner’s decision to advance is partially based on the game state and other nuances specific to a play. There are many reasons a strong arm doesn’t always equal a valuable arm, and vice versa.
This also means that each outfield position has a different baseline arm value:
| Value | LF | CF | RF |
|---|---|---|---|
| Advance | -2.26 | -1.62 | -2.52 |
| Hold | +1.14 | +1.29 | +1.52 |
| Out | +0.87 | +0.52 | +0.91 |
| Total | -0.25 | +0.19 | -0.09 |
Arm strength seems to be most impactful for right fielders. That’s interesting, though it’s also fairly intuitive. Right fielders are often asked to make very long throws, and can get dinged for allowing runners to advance. Those with strong arms are both more likely to throw runners out and to convince runners to stay in place.
I also wonder if there’s a reputation advantage. Balls hit to left field, for instance, allow the runner to read and react to the play in front of them, while balls hit to right field force the runner to refer to the scouting report. I’m sure many runners feel uneasy when they can’t see Addison Barger:

Center field arm value is the least responsive to arm strength. There’s just not a lot even the strongest arm can do from 400 feet away. A center fielder with a valuable arm is a center fielder who takes a good route to the ball and cuts it off clean, rather than one who racks up assists.
Rafaela happens to have all these skills. He has both tremendous range and a strong arm. He finished 2025 with two runs above average from his arm, seventh among center fielders and in the top 15% among outfielders. That’s not a lot, but hey, it isn’t bad for a soft-tossing second baseman.
Ryan Blake is a contributor for FanGraphs and Lookout Landing.
Very interesting. But none of this explains why the Red Sox sometimes wear surplus UCLA jerseys.
The color scheme of the Boston Marathon.
The real question is why do UCLA hats have B on them instead of UCLA. Most teams put the letter for the school they represent on their hats, not the mascot. Their baseball hats say “B”, they named their mascot the Bruins… who is copying whom?
The Cardinals’ broadcast recently mentioned that Nolan Gorman’s offseason prep for moving from second to third base was building up his arm strength. Seems to align with this article’s premise. (Gorman has been pretty good over there this year, but statcast doesn’t have enough sample size on his arm yet this year, last year he was about 5mph stronger at 3rd)
I don’t understand how there’s enough data to create the heatmap grid. How can you get a reliable number for, say, SS vs LF?
Because utility players exist? The most common place for a utility infielder to dabble in the outfield would be left field, for your specific example.
I guess. If a player’s data would qualify, what’s the minimum number of innings you’d use at each position? And I’m assuming it would be in the same season? I would think that the data is noisy from the get-go, and then finding players at multiple positions causes it to be sliced a bit thin. Or maybe not
I believe in my data there are four LF/SS combo players who met the throw minimums set by Savant. So yeah, tiny sample. Most other combos had a much larger sample. Tango did offer a proposed best-fit version of the grid in the comments of his post, which looks pretty similar. But this is definitely more exploratory than anything.
I’d guess there’s also chaining involved–if you have a decent sample of 1B/3B players, of 3B/LF players, and of LF/CF players, you could add those numbers together for an estimate of the 1B/CF shift even if the only actual 1B/CF is Cody Bellinger.
Great article. I’ve been really interested lately in the positional adjustment and how one determines the relative value of players at different positions. It’s a very slippery concept. This analysis does a great job summarizing this one aspect, albeit a small one. Thanks.
Excellent piece – love learning how arm strength is valued.
The actual mechanics of throws are often quite different between positions, which complicates calling these differences purely “arm strength”. The fastest throws from the outfield typically involve players with plenty of time to set up for a big throw (by crow hopping as they catch a fly ball, meeting a grounder with momentum towards the infield, etc) but infielders usually don’t have the time for it. It makes plenty of sense to me that a player might have better mechanics relative to the rest of the league at one type of throw than the other, or be stronger in the muscles emphasized by one type of throw.
I’d be curious to see if guys who primarily play outfield tend to be worse throwers than expected when switched to infield and vice versa. If ranking in arm strength relative to the position goes down when moving to a non-primary position you might call that a mechanics/years of practice effect, if it stays steady maybe it really is about pure strength.
And when there’s a tradeoff between air speed and setup time, it makes perfect sense that the throws that have to cover more distance are the ones that the fielder takes more setup time to throw faster. That’s why you don’t see right fielders scooping the ball and throwing on the run or shortstops taking a huge runup to put more zip on the ball.
Unless the shortstop is Oneil Cruz, whose fastest-ever infield throw was one where he took four steps and barely got the runner on what should have been a routine groundout. In related news, Cruz is now an outfielder.