The "Magic" Mile

One of the things I've yet to figure out about running is knowing how fast I could possibly get and what speed I should target in an upcoming race. I always desire to target a personal record (PR), but often times, the question comes down to by how much. What I really need is a why to translate my training speed to race speed.

A co-worker of mine has pointed me to a website that attempts to answer that question. A professional running trainer, Jeff Galloway has devised a formula for predicting 5K, 10K, and half marathon performance based on the speed at which one can run "a magic mile". The rules and explanation for the magic mile can be found here. The short explanation is that Jeff devised this formula from working with a large number of runners over many years.

Intrigued by the concept and aided by the fact that Jeff lists the formula for translating the magic mile to a 5K, 10K, and half marathon time, I decided to reverse the formula to attempt to predict my magic mile time based on my time for the Shamrock'n Half Marathon, the Super Bowl Sunday Run IV (10K), and my 5K PR from October of last year.

Starting off with the half marathon, my 1:53:04 time translates to a 8:37.9/mile pace during the run. Reversing the formula for the magic mile results in a one mile time of 7:11.6. That seems pretty fast for me (of course, so did the half marathon time).

My Super Bowl Sunday Run 10K time was 51:58 for 8:22.9/mile. Reversing the magic mile formula gives a one mile time of 7:17.3. Pretty close to the other time, and that shows I improved between the 10K in February to the half marathon this month.

As for my 5K PR from October of last year (this course was no where near flat), I finished it in 25:33 for 8:14.5/mile. That translates to a 7:41.5 magic mile. Hard to believe that I would have taken 30 seconds off in roughly half a year.

The theoretical is nice, but I thought I'd put it to the test. Rather than my regular 5 mile run tonight, I decided to do the one mile warm up, followed by the magical mile, then the five minutes recovery walk, before finishing with three more miles of running.

I had to keep telling myself to slow down on the warm up mile and averaged a 10:38/mile pace. I was pretty amped up when I finally got through the warm up and started way too fast (faster than 6:30/mile) before trying to hold a ~7:00/mile pace. It quickly became apparent that was too fast as I kept slowing down until I was at 7:30 and then finally 7:40. As I approached the end of the mile, I was able to pick it up a little bit (~7:30). I finished the mile in 7:36.6.

I was a bit disappointed with the time after calculating the theoreticals. Based on that 7:36.6 time, the magic mile formula predicts that I could run a 25:18 5K (15 seconds better than my PR), a 54:16 10K (2:16 slower than my Super Bowl Sunday time), and a 1:59:38 half marathon (6:34 slower than my Shamrock'n time). Granted, the half marathon time is around what I had expected to do, but my actual time was much better.

So my first attempt didn't really follow the formula (and maybe I don't), but I think it may take an attempt or two to figure out how a fast one mile should feel, and from there maybe I'll get a time in range with his numbers. If not, maybe I'll find a co-efficient to apply that matches my output.

Thankfully I really don't need an answer yet. While I am running the Sac Town Ten Miler next weekend, I already have a plan for the run: keep my miles between 8:25 and 8:35. I'd like to keep my opening miles around 8:25, and if I feel good drop down to 8:20. But my overall goal is every mile under 8:35.