An F1 tyre gets a little slower every lap. Everyone knows that. What you want to know is how much slower, and you want to know before the lap times make it obvious. Once a driver is visibly struggling, the pit wall is already late. So RaceHooks publishes four tyre estimates for eligible drivers in supported sessions, aimed at the two questions a strategist asks: how much longer can this set last, and what does staying out cost?
The four numbers we publish
They arrive on supported analytics feeds when the model has sufficient inputs. Availability depends on the session, race state and plan:
| Output | Range | What it tells you |
|---|---|---|
| Tyre health | 0.0–1.0 | How much usable performance is left in the set right now |
| Degradation rate | seconds / lap | How much lap time the tyre loses with each extra lap |
| Predicted cliff lap | lap number | The lap where we expect degradation to turn sharply non-linear |
| Cliff risk | NONE / GRAINING / THERMAL / BLISTERING_RISK / BLISTERING | A quick label for the kind of trouble the tyre is in |
These are inferred tyre states, not direct tyre measurements. Health is published on a 0–1 scale; healthPct and percentage displays convert it to 0–100.
Health is the number people look at first. The rate and the cliff are what drive the call. Two cars can both sit at 0.45 health while one loses 0.05s a lap and the other loses 0.30s. Same health, very different race.
Raw lap times hide the tyre
The obvious way to measure degradation is to watch lap times climb. That doesn't work on its own, because three other things move lap time at the same time:
- Fuel burn. The car gets lighter and faster every lap, which pulls against tyre wear and partly hides it.
- Track evolution. More rubber goes down over a session and the surface gets faster.
- Traffic, dirty air and lift-and-coast. Noise that has nothing to do with the tyre.
A raw lap-time delta mixes all of that together. Our tyre-state model pulls the tyre's share out of the lap time and updates the estimate when each lap completes. The degradation rate you read is the tyre's, not the fuel load's or the track's.
You can get lap times from any timing screen. We publish the part of the lap time that belongs to the tyre, worked out live, because that's the number a strategy call depends on.
What the cliff is
Tyres don't wear in a straight line. For most of a stint they lose performance gently and predictably. Then the surface or the structure gives up and lap time drops away fast. That's the cliff, and a few laps past it can cost more than the whole stint before.
So the predicted cliff lap is its own output. A steady rate can still be two laps short of a cliff, and if you only watch the rate you won't see it coming. We model the cliff directly and publish the lap we expect it on. Your question changes from "does this feel like it's about to go?" to "do we pit before lap N?"
Reading cliff risk live
Cliff risk is the field you read at a glance. It looks at the shape of the last few fuel-corrected laps, plus track temperature and compound, and names what the tyre is doing:
- NONE: nothing wrong; the stint looks normal.
- GRAINING: degradation has stalled after an earlier climb. The rubber may come back.
- THERMAL: lap time is rising every lap, and rising faster each time.
- BLISTERING_RISK: hot track, soft tyre, one sharp lap of loss. Keep an eye on it.
- BLISTERING: a lap a second or more slower than the one before.
In a live session nobody has time to trace a lap-time curve in their head every lap. A single label tells you what kind of trouble the tyre is in as soon as it starts. If you'd rather have a number than a label, our tyre-cliff model publishes a calibrated cliff probability alongside it.
Tyre health tells you what's left. The degradation rate and predicted cliff lap tell you what staying out costs, and cliff risk tells you when that cost starts climbing.