Adventure Racing Geographic Analysis

“essentially, all models are wrong, but some are useful” –  George E. P. Box

I admit it, I’m kind of a nerd for data and data analysis. So the face that I’m sitting on more than 2 years’ worth of data about all the adventure races in North American has made me realized that there’s a serious gold mine just there waiting to be extracted. Earlier this year, newsletter subscribers of mine got the chance to look at my analysis of a year’s worth of races to see the density of races by month. You can check that article here. A few folks asked me to take a look at the geographic distribution of races, which I thought was a pretty good idea, I just didn’t know how. I discovered Tableau, a data analysis software that is like Microsoft Excel on steroids. It has a handy Latitude/Longitude analysis capability, so it was a simple process of exporting all of ARHub’s extensive catalog of previous and upcoming adventure races and using Tableau to examine the results. Hopefully some entrepreneurial race directors can use this stuff to spot opportunities and grow the sport!

Disclaimer: I think I’m pretty good at capturing all the ARs in North America, but I’m sure I’ve missed a few and a few more got lost in all the data cleaning. Overall, I’d give myself a “B”, meaning that directionally, this map is accurate but not perfect. I had to cut out Alaska (sorry!) to fit the rest of the continent in successfully. Spot any errors? Let me know at cy@adventureracehub.com

ar-heat-map

Map #1 is a density map of all ARs from the start of ARHub (late 2014). So it’s dominated by the 2015 and 2016 race seasons, with a few from 2014 and 2017. The size of the blue dot indicates the quantity of races in the area. So when REV3 holds 4 races in one weekend at the same venue for 2 years in a row, you end up with a large blue dot representing 8 races. This biases the map towards races that occur in the same location, but still provides a lot of valuable insights.

  1. What the hell, Montana? Listen, I know that in terms of total population and density, you’re a small state. But with endurance meccas like Billings and Bozeman, there’s no excuse for not hosting a single AR. Readers, if you’ve got some friends in MT, email them right now and tell them to get an AR setup already. You can’t have a state with a motto like “Big Sky Country” and not host an awesome adventure race.
  2. Same goes for you, AZ. Or least northern AZ, like up in Flagstaff. Another endurance mecca with no representation in AR. Nobody wants to race amongst a bunch of giant cactus, but north AZ is a fantastic location. Shame.
  3. You know what? I’m throwing the whole West Coast under the bus. We’re supposed to be the “healthy” coast with all the yuppies and hipsters and what not. Look at the sparse offers of ARs compared to just the mid-Atlantic. Especially you, California. Somebody in NorCal needs to step it up.
  4. There are some serious “Adventure Racing Destinations” thanks to a big AR series having preferred race locations. REV3, FLX, and BendAR have really established awesome systems to create such success.
  5. Kinda expected more from Texas, given its size.

Now here’s a graph of the quantity of races that have occurred by state/provincestate-province-quantity

Florida and Pennsylvania are officially the co-champions of the “what state is the best for Adventure Racing”. FLX pumps out a crazy quantity of high quality ARs, so credit goes to Ron Eaglin and the FLX crew for their commitment to the sport. Pennsylvania benefits from great series like Roostock and Goals ARA among others. Clear evidence that multiple race series can thrive while drawing from the same population. Anyone else a little surprised to see Kentucky that far up the list, besides Virginia? The line you see represents the statistical average for the number of races that occur across all the states and provinces that have races, which is 7.2.

Here’s that same graph again, but this time color coded by year.

state-by-year

Here’s the same bar graph again, but this time filtering just to 2015 and 2016, as those are the only 2 years that I have (mostly) complete data for all ARs in North America. Shifts up the ranks a bit 🙂

count-of-races-by-year

Finally, here’s the map analysis again, this time with the dots turned into pie charts to reflect when a city hosts multiple races year over year.

race-location-by-year