Six years of my own running data, and the year the surface changed everything

This site used to carry eighteen study pages, thirteen numbered studies, two microstudies, three secondary studies, each proposing a hypothesis about adaptive envelopes or cadence protection or ecological narrowing. In September 2026 they were taken out of search and removed from navigation.

Three reasons, all of them mine. They shared between 11 and 21 percent of their text with each other. They were built partly on derived tables that contained sentinel values I had mistaken for real measurements. And they rested on running-dynamics metrics that turn out to exist for only fourteen months of a six-year archive.

This one page replaces all eighteen, and it is deliberately smaller than the thing it replaces.

Across six years, 652 sessions totaling 2,147 miles are logged under a running activity type. Apply a cadence threshold of 140 steps per minute. The conventional floor for running gait, and only 383 sessions and 1,561 miles qualify. The rest is walking.

That is not sloppiness on my part. Until the end of September 2021 my watch had no walking activity type in use at all. Walks had nowhere to go but the running label.

The low years are not injury. 2021 and 2022 I was mostly in a gym: 41 indoor cardio and 33 strength sessions in 2022 alone, against 47 runs. That is a training choice, not a foot.

In 2020 this was an outdoor activity, entirely. From 2023 it is a treadmill activity, almost entirely. The switch is abrupt, it lands in the same year the pace drops by five and a half minutes a mile, and I cannot point to an injury, a decision or a piece of equipment that caused it.

Running volume and running pace both changed when the surface stopped varying, and not before.

Long before I ran this analysis I had written on this site that level ground is close to unremarkable for me, and that uneven ground is where the missing hindfoot motion shows, because the joints that let a foot mold to the ground are the ones that were fused.

A published series of triple arthrodesis patients reports the same thing from a cohort rather than from one person: two thirds described significant limitation walking on uneven ground.1

This archive is a third line, and none of the three was collected to support the other two. My own account came first, the cohort literature I only read afterwards, and the watch was recording the whole time without being asked a question.

It is badly confounded and the confounds are obvious. A treadmill is easier than a road for reasons that have nothing to do with a fused hindfoot: the pace is imposed rather than chosen, the belt is cushioned, there is no weather, no camber, no traffic and no navigation. Any runner improves moving indoors. Nothing here isolates my foot as the reason.

It is one person, and there is no control. Both my feet were operated on in infancy, so I do not even have an untreated side to compare against.

Six years of training would improve pace anyway, surface or no surface.

And the causation may run backwards. I may have moved to a treadmill because I wanted to run more, rather than run more because I moved to a treadmill. I genuinely do not know which.

What survives all of that is narrow: in someone whose own account and whose cohort literature both point at uneven ground as the limiting factor, six years of activity data show running becoming possible at volume in the year the surface stopped varying.

The watch is on my wrist. Stance time, vertical oscillation, step length and running power are estimated from arm and trunk movement. They are not specific to either foot, and they cannot tell a fused side from an unfused one.

The one field that could have separated the two, stance time balance, was never recorded once across the entire archive. Those fields also only exist from 2025, so they cover fourteen months of a six-year record.

Analyzing them would mean describing an estimate of a body and calling it a measurement of a foot. That is most of what was wrong with the eighteen pages this one replaces.

My twenty-five year record is the clinical companion to this page. The operative history, an incidental x-ray and a scored outcome questionnaire. Running after an arthrodesis is the practical side. Uneven ground covers the surface problem properly. Research is where the rest of it lives.