Watts Target — L’Étape du Tour 2026

At the very front (overall finish): 1. Luca Cavallo 5:20:20 · 2. Thibaut Clément 5:24:31 · 3. Antoine Berger 5:32:25 · 4. Nicolas Dricot 5:33:54 · 5. Thomas Seigneuret 5:35:59

Why Sarenne reads low. Its timed split is 20.8 km at ~3% — the front rode it at ~31 km/h (Cavallo, 39:52). That’s a power-on-the-flat effort, not a steep col, so average climbing W/kg is naturally lower than the real cols. It’s excluded from the grimpeur classification and left uncalibrated here — treat its number as indicative only.

See how these Strava times stack up against the Tour de France →

Method & formulas

The window = two ranking bases. “On the climb” takes the riders placed in the 0.9N–N band on that climb’s split (e.g. 900–1000 on the Galibier). “At the finish” takes the riders whose general rank is 0.9N–N and reads what they climbed at. On-climb is the harder end; the gap between them is how much the strong early climbers fade before the line — widest on Croix de Fer (the first climb).

For each end I take the band’s median climb time, then:

v      = distance / time                         (m/s)
θ      = atan(elevation / distance)             (slope)
VAM    = elevation ÷ (time / 3600)              (m/h)

P      = [ m·g·sinθ·v                            (gravity)
         + m·g·Crr·cosθ·v                        (rolling)
         + ½·ρ·CdA·v³ ]                          (air)
         × drivetrain × k_climb

W/kg   = P / m

Constants: g = 9.81, Crr = 0.005, CdA = 0.32 m², ρ = 1 kg/m³ (altitude), drivetrain = ×1.02. k_climb = per-climb calibration fitted to 3 power-meter files (61/71/85 kg): Croix de Fer ×1.19, Télégraphe ×1.11, Galibier ×1.13. It absorbs variable-power vs. steady-power, drafting/wind, and parameter slack; it’s larger on Croix de Fer because its mid-climb descents make net elevation understate the effort. The same model runs on the time you enter above.