What Is the Physics-First Algorithm?
The Physics-First algorithm is the core analytical engine behind StrideOdds. Unlike traditional handicapping models that rank horses by speed figures or public consensus, the Physics-First approach starts from the physical reality of the race — the biomechanics of the horse, the physics of the surface, and the environmental conditions — and works outward to a true win probability.
The result is a fair-odds line: a calculated probability for each runner that exists independently of what the public thinks. When the Physics-First line diverges from the tote board, that divergence is where the edge lives.
The 10 Dynamic Contingency Variables
The algorithm processes 10 primary variables that interact dynamically — meaning a change in one shifts the weight of others:
1. Live tote movement — where sharp money is flowing in the final 20 minutes
2. Stride velocity — the horse's recent speed of stride relative to optimal
3. Stride length signature — individual biomechanical pattern vs. career baseline
4. Sectional timing — fractional splits adjusted for pace scenario
5. Track moisture — real-time surface reading beyond the official designation
6. Wind velocity and direction — effect on final-turn performance especially on turf
7. Track bias rating — live post-position and running-style bias for the day's card
8. Connection strength — trainer-jockey-horse combination win rate in specific conditions
9. Class differential — horse's rated ability vs. competition level today
10. Pace shape projection — energy expenditure model based on expected early fractions
Real-Time Matrix Modifiers
Beyond the 10 variables, the algorithm applies Matrix Modifiers — dynamic adjustments triggered by late-breaking conditions:
- Gate change — a horse moved to a significantly different post position within 24 hours
- Scratch impact — the removal of a horse changes the pace scenario for all remaining runners
- Late weather shift — a rain system arriving on race day changes surface reads
- Equipment change — blinkers on or off, or a change in shoeing, affects stride pattern
Each modifier recalculates the full probability distribution, not just the affected horse, because racing is a zero-sum game: every probability adjustment to one runner changes the expected win probability of all others.
Signal Latency: Under 150ms
One of the most critical performance requirements of the Physics-First algorithm is latency. Tote odds move in real time, and the window to act on a mispricing narrows as post time approaches. StrideOdds processes the full algorithm cycle — from tote data ingestion to on-screen Confidence Score — in under 150 milliseconds. This means users see a mispriced odds flag before the public can correct the price.
How It Differs From Speed Figure Models
Traditional handicapping tools like Beyer Speed Figures or Brisnet ratings provide a backward-looking assessment: how fast did this horse run last time? The Physics-First algorithm is forward-looking: given all current conditions, what is this horse's true probability of winning this specific race on this specific surface under these specific conditions today?
The difference is the difference between a weather forecast and a history of past temperatures.