Friday, January 10, 2014

My Running Log

As an engineer I love numbers and tracking things. This naturally bled into running. I remember back in 2005 when someone showed me a GPS running watch. I got one the next day. It's not that I need to run a certain distance or pace - I just like to know what I did. 

As my education and job taught me how to use Excel I eventually put my running log and stats on a spreadsheet. On day I started plotting my races and paces. I was curious as to how my pace varied with distance. I was surprised to find not just a relationship, but a very strong relationship –a logarithmic one. In fact, when I plotted my personal bests I got an R squared value of 0.99. Logarithms show up in many neat places with our bodies from how we perceive light levels or sounds and even the amount of time it takes us to make choices (based on the number of choices) so I guess this should make sense.

My personal bests
I decided to see if this held for world records. Low-and-behold, it did – with an R squared of 0.99 again.

A small selection of world records
As you could guess these folks' coefficients are much smaller. (And perhaps the mile record is a bit soft - but probably because it is not raced as often as the 1500)

When I went back and looked at my races I could then see which ones were above average efforts and which ones I had slacked on. Well . . . to a point. This trend worked as long as the races were roughly comparable.  In other words, as long as the races were all road races with modest elevation gains. So a race with some intense elevation gain or on trails was always going to be a little slow..

So now my nerdy running goal is to figure out how to quantify these other variables so that I can tell is that trail marathon was a great effort or if I have a little more room to target a PR. Right now I think the most important (an reasonable easy to quantify) variables to look at would be:

-Age
-Elevation gain
-Elevation loss
-Terrain (not quite sure how to quantify other than giving a variable of 1 for road, 2 for dirt, 3 for technical trails)
-Altitude
-Wet bulb temp

So eventually I will have a part two once I have enough data. As we say at work, happy nerding.

2 comments:

  1. You might pop altitude in there as well. Especially where most of the trail races are.

    ReplyDelete