Note: The following is a guest post by Mike Roberts of Pluralsight.
My wife recently ran the San Diego 100, an endurance race that demands a lot of training, focus, and grit. Trail runners from all over the world spend up to 32 hours running 100 miles of trails without stopping. Yes, it’s pretty epic and I couldn’t be more proud of her accomplishment.
As runners, there’s a need to make sure one can at least control the controllable aspect of any race. This can include, but isn’t limited to, food, pace, and gear. While we made sure the food and gear were set, there was the pesky notion of pace. How, considering so many variables, does one estimate and control pace? Data, of course!
I decided to create a dashboard. This wasn’t a tool for her to use on the run; rather, it was for me (and other crew members) to use to ensure she stayed within range to finish in the allotted 32 hours.
We needed to know a few major things: aid stations (and distance between them), elapsed time, total miles run, and total miles remaining. Lastly, and perhaps the most important to all runners, we also looked at weather data. If it’s hot and humid, a runner will need to adjust both pace and food. Our analytical tool, then, would need to also have live weather data, preferably on an hour-by-hour basis.
Before starting, we extracted two major data sources. First, we converted the PDF of the aid station data to a CSV file and added a few fields (most notably a field to sort the aid stations since they do progress in order).
Second, we needed weather data and wind speed to augment our analysis as she was running into the night. (Runners who've run into 20-mph headwind know why this might be important.)