uphill notes · running fingerprint
Strava × Claude · a mirror, not a coach

What the data says
about how you run.

Not advice — a portrait. Pulled from 233 moving data points across five recent runs (easy road, tempo, two trails and the Trail d'Alsace race), this is the shape of Martin's running: how the terrain bends his pace, and the rhythm his legs default to.

The signature curve

Speed as a function of gradient. Every runner has a different curve — this one is the fingerprint. Hover to read it.
Uphill costs pace, downhill returns it — but never all of it.

Pacing personality

Last third vs. first third of each run. Positive = finished faster (negative split). It reveals a split personality.
finished fasterstarted faster
Natural cadence154 spm

His legs default to 154 steps/min — the exact tempo the Run-to-the-Beat matcher chased. Range 146–162.

Climbing tax−28%

Uphill pace (6.8 km/h) vs. flat cruise (9.5). Every metre of climb is paid in speed.

Descending flow10.5 km/h

Fastest on gentle downhills — willing to spend the quads for free speed.

Overall pacing+1.4%

Across all runs, almost perfectly even — a disciplined pacer who fades on tempo but negative-splits trails.

Data: Strava streams via the Claude MCP connector (read-only) — activities 18541461784 (TDC race), 18629712572, 18447334682, 18474249566, 18604022514 · velocity & gradient smoothed by Strava · 233 moving samples · This is descriptive, not prescriptive: a mirror of past runs, not a training plan.
Part of the uphill notes Sunday Long Run · uphillnotes.com ↗