(Red = race, blue = training run, green = treadmill run).

Drag to zoom in, right-click to reset.

Fitness measure

"Fitness" is an estimate of the speed (in kph) at which I could complete a half marathon on a flat course. It's calculated and plotted for every training run I do, provided I have heart-rate and elevation data (i.e. I had my Garmin watch and chest-strap on, or I was on a treadmill). It's based on the following measurements and assumptions:

• Speed. This is the starting number.
• Heart rate. I assume that (for a given fitness level) speed increases linearly with heart rate from a base of 80bpm, and that a HM would be run at 175bpm.
• Elevation change. I assume that for every 1% climb my speed falls by 0.75kph.

These assumptions are based on data from dozens of training runs on various courses (including a lot on treadmills) and at various paces. So for instance, a training run of 10km in 1 hour with a heart rate of 160, climbing (and then descending) a 200m hill, means:

• A speed of 10kph.
• A heart rate of 160bpm, which means scaling up the speed by (175-80)/(160-80)= 19/16.
• Total climb of 200m over 10km is 2%, which means adding onto the speed 0.75*2.

Overall that's 10*19/16+0.75*2=13.375kph on a flat HM course, or just under 1:35.

What's the point?

Theoretically this measure should be independent of what course I choose to run and how hard I push myself. It's therefore comparable across all my training runs. It doesn't affect my training - I don't need to run a particular way to get a "high" score and beat my previous fitness measure. I can therefore train whichever way I deem most effective.

The data is interesting in itself (e.g. to track my recovery from pneumonia), but will also be useful in setting expectations for races. The excitement of a race can screw with your heart rate, meaning a predicted race pace is more useful than running to a certain heart rate.

What are the flaws?

• After several months of decent training I'm limited much more by my body's ability to efficient use oxygen than my cardiovascular system's ability to deliver it. The fitness measure levels off but I continue to improve as I keep running.
• I find I run less efficiently in races than in training. I'm working on fixing that, but in the meantime this fitness measure predicts more than I am capable of delivering in a race.
• The measure takes no account of terrain: running on grass, mud, ice or snow is harder than asphalt or hard-packed dirt.

Does this cover all the running you've ever done?

No. I ran a 1:45 half marathon (my first) in mid-September 2012 after about 6 months of training from a base of nothing. After the HM I took a week off and then resumed light training. By late October I was back into serious training, and had bought a new chest-strap so I could start gathering heart-rate data on my runs. So this chart starts after ~7 months of training.

Why the big gaps and disastrous loss of fitness?

The 4-week gap in early 2013 was due to pneumonia. The longer gap over winter 2013-2014 was due to snow and lack of a gym membership or home treadmill.