Imaginary Racing API – Real World Example

All this talk about releasing data could use an example, no? I’ve got one!

In our many “what-if” discussions about racing, Jessica Chapel and I refer to the “imaginary racing API” (as seen here!). The discussion usually goes something like this: “Hey, wouldn’t it be great if we could make something to do XYZ?” “That sounds great! Now if we just had the imaginary racing API.”

Our dormant site, Omnisurface Stars (done in collaboration with Kevin Martin and Valerie Grash), is the perfect example of something that could benefit from a not-so-imaginary racing API, like the ones that the city of New York make available for free so people can build cool things like this, this, this AND this (just to name a random few).

Let’s say we could programmatically access the following data for free:

Horse name
– date
– track
– SURFACE (important for this example)
– distance
– race condition
– finish position

And that’s it. We could set up a site like Omnisurface Stars to automatically populate an up to date list of horses that have won on all three surfaces and, depending on how the data came back, we could also derive whether horses who have won on all three surfaces have won in graded races and visually display those differently. And, if any horse that had won all three surfaces at the level of Grade 1, we could even display those differently, all without having to manually update the site.

As it stands now, if we wanted to keep this site up to date, we’d have to  manually look through charts and past performances on a daily basis for EVERY TRACK! And really, who the hell has that kind of time?

I’ve always wanted to do a post where I came up with a dataset for the Imaginary Racing API and then asked clever people with good ideas who have a grasp of the development or dataviz worlds (like Jessica or @o_crunk) to brainstorm on what sorts of things could be created if the IRAPI existed. Perhaps it’s time to dust off that idea (you know, after the BC… )