"Statlas also makes it easy to spot the littler plays"

October 24, 2013 4:04 PM

Statlas, a way to visualize baseball.

Tampa Bay over Cleveland in the AL Wildcard game.
Surgical Wainwright Shuts Down Pirates as St. Louis takes Game 1 over Pittsburgh
Still alive: Rays walk off in ninth to breathe life into season
Red Sox beat Cardinals, 8-1

and many more...


via fastcodesign

posted by the man of twists and turns (4 comments total)

Pretty design! The left-hand side is certainly an interesting semi-redesign of the conventional scorecard, and you could argue this design conveys some of the baserunner information in a more intuitive way with the vertical connecting lines and the nicely consistent use of the vertical axis as the time dimension.

But I find the strange vertical compound-curve S shapes that they're using far harder to read than the more conventional horizontal line graph that Fangraphs uses for win expectancy (e.g.). What do the varying areas and slopes of the S shapes mean? Or are they just decorative? It seems like a case of Tuftean chartjunk to me.

The FastCo thing says "It also tracks data on the different pitchers--what balls they threw against whom, and when--all with the same elegant and efficient design chops." But those pitch-type plots are really not analytically interesting at all, compared to the much richer PitchFX data they're drawn from. And similarly the hitting side doesn't actually capture many of the richer and more interesting contemporary generation of hitting stats — rather it just uses the old counting/outcome stats and adds the WPA column alongside them. I feel like once you get past the neat visuals, some of this stuff could use a rethink to be better as information design.
posted by RogerB at 4:29 PM on October 24, 2013

I actually like the weird S-shapes. I think they capture the fact that each at-bat is a discrete moment as far as win probability is concerned. The FanGraphs line chart sort of implies things are fairly continuous when they aren't.

As far as I can tell the slope of the ribbon thing is determined by how sharp of a change in probability that at-bat caused, if the change (length of the ribbon) is small enough. So the angle is sharper if it's a greater change, and the area you see is bigger. If it's long enough they just make it horizontal and rely on the length to convey the information. It makes intuitive sense if you pretend it's a physical ribbon.
posted by vogon_poet at 4:57 PM on October 24, 2013

Absolutely love the 'At the Plate' on these! The pitching stats are a little hard for me to parse, but all-around, these are pretty fascinating.

Thanks for posting!

/eponysterical
posted by graphnerd at 5:15 PM on October 24, 2013 [1 favorite]

It seems like a case of Tuftean chartjunk to me.

Tuftean Chartjunk sounds like a rad D&D character.
posted by Celsius1414 at 8:10 PM on October 24, 2013 [1 favorite]