Aleks points us to this beautiful dynamic graph by Noah Veltman showing the heights and weights of NFL players over time. The color is pretty but I think I’d prefer something simpler, just one dot per player (with some jittering to handle the discrete reporting of heights and weights). In any case, it’s a great graph. Click on the link to see it in action.

P.S. Even better, once we move to a dynamic scatterplot, would be to use different colors for different positions, and to allow the reader of the graph to highlight different positions. On the linked page, Veltman writes, “the blob separating into multiple groups in the 1990s . . . likely reflects increased specialization of body type by position.” But we should be able to see this directly, no need for speculation, right?

I’d prefer a set of densities (could be separate for height/weight, or a 3D plot). Dynamic requires one to remember where you’ve been. Or, I’m wondering if something could be done along the lines of Tufte’s Napoleonic retreat from Moscow.

I’m confused: What are the colors mapping to? What are those tiny percentage numbers? THe percent of players in each bin?

Anyone know a source to get the raw data file?

I tried to add up the squares to make sure they equalled 1 but I can’t tell the difference between these colors when they’re scattered around and placed next to similar colors which throw off how you see. I can do it if I work at it, but why would I want to work at that?

“If all athletes have gotten taller and heavier at similar rates, then it would reflect a general trend”

but a line-backer isn’t like an athlete in any other sport. You can see a separation in the distribution, where a heavier set group is away from the rest. Alternatively, this could be a realization of the effectiveness of these individuals and bringing their potential to a logical conclusion –more weight. I don’t necessarily see this as a reflection of society as well –quite a biased sample, no?

Still, a very well made graph! Thanks!

It’s kind of cool that you can see the specialization of the offensive line position over time (the separation of a distinct group over 6 foot and 280ish lbs.)

I don’t really like animated graphics because I don’t have time to verbalize for myself key comparisons. This seems like a good opportunity for the use of Tuftean “small multiples.”