Torque by RaceProof

Model depth

The accuracy of every activity recommendation a wearable device produces is constrained by the physiological model underneath it. If the model sees one dimension of a user's fitness, the recommendations will be right for some users and wrong for the rest. The platforms that solve this will build durable user trust. The ones that don't will compete on hardware while their software feels interchangeable.


The bottleneck

One threshold, applied to everyone

Every major wearable platform uses a physiological model built around a single threshold measurement. This measurement was not chosen for its physiological depth. It was chosen because it is reproducible in a laboratory. The protocol was designed for researchers, adopted by the industry, and applied identically to every athlete.

The result: two athletes with identical threshold values but fundamentally different physiological profiles receive the same training zones, the same recovery guidance, the same post-workout interpretation. The model cannot distinguish between them. The device treats them as the same person.

The cost of generic
When a device tells a user to rest and they feel strong, they ignore it. When it says push and they are exhausted, they resent it. Each wrong recommendation erodes trust in the device. Enough wrong calls and the user stops looking, then stops wearing it, then cancels. The model is the bottleneck.

The correction

An industry moving toward more dimensions

The industry has already acknowledged the problem. Over the past decade, every major platform has independently moved away from a single threshold and toward more physiological dimensions. The direction is not in dispute. The question is how far it goes.

2017
Wahoo / Sufferfest introduces 4DP (Four-Dimensional Power). Measures neuromuscular power (5s), anaerobic capacity (1min), maximal aerobic power (5min), and FTP (20min). Four distinct power dimensions instead of one.
2020
Garmin acquires Firstbeat Analytics. Makes VO2max estimation, lactate threshold detection, EPOC-based training load, HRV analysis, and training effect exclusive to Garmin. Cuts competitor access to twenty years of Finnish exercise physiology research.
2021
COROS launches EvoLab. Adds VO2max, lactate threshold, training load, race prediction, and training focus classification. Built in-house after Garmin's Firstbeat acquisition closed the licensing door.
2025
Wahoo releases Dimensional Training Load. Extends 4DP by breaking training stress across all four power dimensions with independent decay rates. The most explicit acknowledgment that single-dimensional load tracking is insufficient.
2026
RaceProof releases Power Threshold Array in beta (Q1). Seven detected thresholds plus the positional dimension (PT Seat™). Commercial licensing targeted for Q4 2026.
Each move independently validates the same conclusion. Boutique tools like INSCYD and WKO5 have explored deeper models for years in professional coaching contexts. The wearable industry is now catching up, bringing multi-dimensional physiology to consumer devices.

The model

Power Threshold Array

RaceProof's Power Threshold Array detects seven distinct physiological thresholds across the full spectrum of human power output, from neuromuscular through oxidative. Each threshold corresponds to a bioenergetic transition: the point where one energy system yields dominance to another.

Each threshold is detected, not calculated, from sensor data. PTA is a software layer that operates on data wearable devices already capture. It requires no new hardware, no new sensors, and no change to the user experience.

PTA is currently validated on cycling power data. The same detection architecture will extend to heart rate, opening the model to every wearable device on the market regardless of sport or sensor type.

PT-NPT-SPT-GPT-XPT-OPT-VPT-T
ThresholdEnergy system
PT-NNeuromuscularPeak phosphocreatine sprint capacity
PT-SSprintPhosphocreatine depletion onset
PT-GGlycolyticAnaerobic glycolytic activation
PT-XMaximalGlycolytic ceiling, maximal anaerobic output
PT-OOxidativeOxidative phosphorylation dominance
PT-VVO2maxMaximal aerobic capacity, fully expressed
PT-TThresholdFunctional threshold, lactate equilibrium
Each threshold is detected individually for each person. Each is tracked over time as an independent development signal. The first six are invisible to every model that begins and ends at a single threshold.

The dimension

PT Seat™

Every existing platform measures power at a fixed duration chosen for testing convenience. Wahoo's 4DP, the most dimensionally advanced competitor, measures at 5 seconds, 1 minute, 5 minutes, and 20 minutes. Those durations are the same for every athlete.

PTA detects where each threshold actually occurs for each individual. Two athletes can produce identical power, but if their thresholds occur at different positions, their physiology is meaningfully different in a way that power alone cannot distinguish. That positional dimension is what RaceProof calls the Seat™.

Scalar analysis
Garmin, Apple, COROS, Polar, Wahoo, Whoop. A single power value at a fixed duration. One number per threshold. Can go up or down. That is the only development signal it produces.
Vector analysis
RaceProof. A coordinate pair: power and position. Each threshold is a point in two-dimensional space. Power can shift, position can shift, or both can move together. Each pattern is a distinct physiological signal.
What the Seat reveals
When a threshold's PT Seat™ shifts outward without losing power, the energy system is extending. When power increases without the PT Seat™ moving, the system is strengthening but not extending. When both move together, the development is unambiguous. Each movement pattern is a distinct physiological signal that scalar analysis cannot produce regardless of how many thresholds it measures.

PT Seat™ is a truly novel contribution that adds a multiplier effect to depth of analysis. Where every existing model produces a scalar value at a predetermined duration, PT Seat™ resolves a two-dimensional coordinate. No other platform performs vector analysis on physiological thresholds.
The landscape

What each platform measures

The matrix below maps the physiological model depth of every major wearable platform. Each column represents a dimension of the model. PT Seat™, the positional dimension, is exclusive to RaceProof.

PHYSIOLOGICAL MODEL DEPTH

The implication

Fewer wrong answers, at every level

The physiological model touches everything the device says to the user. Activity analysis, fitness forecasting, training recommendations, readiness assessment. When the model sees one dimension, every output built on top of it inherits that limitation. The post-workout summary is generic. The suggested workout is generic. The readiness score is generic. The user learns to ignore all of it.

This is not a theoretical problem. The accuracy of readiness and recovery features on major platforms has been publicly questioned by researchers and experienced athletes. When recommendations don't match lived experience often enough, trust erodes. Once trust is lost, the feature is ignored regardless of how prominently it appears on the screen. The underlying model is the limiting factor.

Fourteen physiological signals instead of one or two means the model can distinguish between athletes who look identical under a single-threshold lens but are physiologically different people. Activity analysis becomes specific to each person's energy system engagement. Fitness forecasting reflects development across the full spectrum, not a single number trending up or down. Training recommendations respond to what the individual actually needs, not what the population average suggests. The device stops guessing. It starts knowing.

Software, not hardware
PTA is a software layer. It operates on sensor data wearable devices already capture. No new hardware. No new sensors. No change to the user experience. PTA is currently validated on cycling power data. The same detection architecture will extend to heart rate, opening the model to every wearable device on the market regardless of sport or sensor type.

Related
Loom & Effort Signature — how RaceProof renders athletic development as color and light.
The Racing Effect — population-level evidence that competitive racing develops the full physiological spectrum.
RaceProof Product — the full platform including Loom, Trust Score, and Proof Network.
Torque by RaceProof · PTA Methodology · RaceProof LLC