Activity intelligence for competitive cycling
Certified data. A new model.
A new way to see.
Competitive racing produces the closest thing to lab-controlled data that exists
at population scale: maximal efforts, consistent protocols, independent
device recordings, hundreds of thousands of athletes. The data has always
been there. No one built the tools to read it.
RaceProof compiles, cross-references, and certifies that data into a single
authoritative record. Then it analyzes what certified data makes visible:
seven bioenergetic thresholds where your physiology actually
shifts, not the single number the sport has been stuck on for twenty years.
What we built
Every layer required the one below it.
When Banister published the impulse-response model in 1975 and Coggan translated it
into TSS, CTL, ATL, and TSB, they needed the same thing: certified data, a specific
schema, and a new visualization to make it legible. That framework defined endurance
training science for a generation. It has been largely stagnant since.
RaceProof is the next iteration. Three layers, each enabling the next.
01
Certified Data
Cross-source compilation and verification. Independent devices recording the same effort, resolved into a single authoritative record. When the data passes, it earns the badge.
02
Power Threshold Array
A new analytical model that replaces single-threshold FTP with seven bioenergetic transitions detected across the full power-duration curve. Each corresponds to a distinct energy system. Together they form a fingerprint that develops along fourteen independent dimensions.
03
Effort Signature + Loom
Every ride has a shape. RaceProof captures it as an Effort Signature: your effort rendered as color against PTA thresholds. Sprint power burns gold at the top. Threshold effort glows deep plum at the bottom. Line up a season of signatures and they become Loom, a biographical view of athletic development over time.
How each layer works →
Research
The first population-scale evidence that racing develops the full curve.
Not at a single FTP threshold, but across every duration from 5 seconds
to 20 minutes, for every rider in a validated four-year corpus. The data
required certification. The analysis required PTA. The findings required Loom to
present them.
+21.2W
Annual development gap between year-round and seasonal racers, full power-duration curve
2 / wk
Racing frequency above which whole-curve fitness development accelerates
145K
Classified riders in the validated four-year corpus
Read The Racing Effect →
Publication
Torque
RaceProof's research arm. Population-scale analysis of competitive cycling
fitness, rider behavior, and platform engagement. Four years of event data,
hundreds of thousands of riders.
Read Torque →