WR hand size predicts drop rates (1 Viewer)

Do WRs with bigger hands drop fewer passes? Yes they do, but it's a modest difference.

I found combine hand size data going back to the 2009 draft class, and got data on NFL drops from Pro Football Focus. There were 176 WRs that I found who had at least 1 NFL target and a combine hand size measurement - I split them into 3 groups based on hand size (small, medium, and large). Then I calculated the total drop rate (drops / catches+drops) for each of the 3 groups. The results:

9.125" or less: 9.93% drop rate (534/5380)

9.25" - 9.625": 9.09% drop rate (511/5620)

9.75" or more: 8.43% drop rate (352/4176)

The gap between the small hands group and the large hands group is statistically significant (p = .012 by a chi square test). It is a 1.5 percentage point difference in drop rate. This translates into one extra catch for every 67 catchable passes (roughly 60 catches & 7 drops for a small handed WR vs. 61 catches & 6 drops for a large handed WR).

Another way to look at the data is to treat each player as one data point, and do a linear regression predicting NFL drop rate from hand size. Doing this with the 75 WRs who have 100+ NFL targets gives a slope of -1.4, meaning that each additional inch of hand size translates into a 1.4 percentage point reduction in drop rate. This is very close to the estimate that we got with the other method (it's actually slightly larger than the other estimate, since the "large hands" group had hands that were about 1.2 inches bigger than the "small hands" group, on average).

The data set is here, broken down into all 21 hand size groups (8.25" hands, 8.375" hands, etc.) and including targets, receptions, drops, yards, and TDs for each group.

Let me go on... like I blister in the sun.
Let me go on... big hands, I know you're the one.

Nice find ZWK. I am a bit surprised by this.

That said a one catch difference every 67 catchable targets (which begs the question of what is catchable?) isn't going to impact a seasonal projection enough to matter. On a weekly basis even less so.

Two things.

1. The hand size data needs to be cross tabulated by round drafted.

The reason is if more guys with small hands rate being drafted in the first round based on other attributes over lower drafted guys with large hands those other attributes would trump hand size as the primary reason for drops.

Just a suggestion but if it is possible this hand size data should also be broken out by round drafted.

2. Instead of having drops as the end goal, why not have overall quality and longevity of career cross tabulated by hand size?

The reason is I think the overall quality and longevity of a WRs career should be the end goal is because achieving the most positive end goal

= lengthy quality career

Would be the primary objective in drafting a WR over reducing a negative such as drops. Drops would be a sub-set of a lengthy quality career but should not trump the primary positive goal.

Can you share your data set of each receiver as a data point? Thanks

Also I think it would be useful to add at least another explanatory variable to the regression, such as round drafted (h/t BS)

I've been curious about this exact thing and was trying to make a point in the Perriman thread, so up front, thanks for crunching the data. I do agree with the others that hand size has to be factored into round drafted because the early round picks generally have the requisite skill set to be a fantasy contributor. 3rd round picks and later, if they produce at all, tend to be situational players or guys that have a big year or two and then disappear very quickly. I get it that there are the Antonio Brown's, Wes Welker, Colston types, but they are the exception.

Thanks for starting the discussion!