Rolling Mileage Windows

A few days back, I completed the final 2 miles of a long run much slower than the initial 16. Scientists call this a bonk. It was hot, and yes, a there was a hill during mile 17, but these elements do not fully explain what happened.

The first metric I looked at to understand my bonk was weekly mileage. At the start of training, I neurotically plan to avoid a sudden hike in distance, which I've learned is a sure fire path to injury. I am also a devoted adherent to the 80/20 rule for endurance training, which recommends that one never run more than 20% of mileage at a strenuous pace. My plan also makes sure that I never come close to the 20% ceiling. So, I know it's not my total mileage or the total hard miles I'm doing.

However, my planning consists of 4-week cycles, with three hard weeks followed by an easy week. I know I'm not breaking any weekly limits that might result in overtraining or injury. Yet, what if I change the timescale for how I assess mileage?

As a definite nonprofessional athlete I must weave my runs into daily life. As a result, sometimes my mileage and hard efforts cluster. Conversely, I watched a Kofuzi video explaining his 9-day cycles where he separates all hard efforts with two day recovery periods. Must be nice.

I decided to look at 3, 5, and 14-day distance totals in addition to the 7-day totals I use for long term planning. My 7-day total was 60 miles, typical for the past few weeks. However, on the 3, 5, and 14-day scales I hit 36, 52 and 136 miles respectively, each a peak not seen for at least the past two and a half months.

Another approach is to convert these timescale distances into 7-day equivalents. Running 36 miles over three days seems fine, but the 7-day equivalent is 84 miles.

That is about 10 more miles than I've ever run in a week and I'm currently aiming for 50-60 miles per week. A 40% increase over my aim would have raised concerns during my planning had I been able to see it coming.

The rolling 5-day total for the week was 52 miles, a bit steep, but again it masks the real load due to clustering. 52 miles in 5 days is proportional to running 73 miles in a week. Here again, I'm floating 20% above what I intend to run.

The same overtraining story emerges when looking at my 14-day rolling total of 136 miles, averaging 75 miles per week, which is significantly high for me.

This week has shown me that tracking weekly distance and the percentage of hard effort alone isn't enough to avoid overtraining. Looking at 3, 5, and 14-day rolling totals along with their 7-day equivalents, illuminates how real-world clustering of miles can trigger mini periods of overtraining within just a few days.