Palmer CCW · 13 corners · 92 clean-pace laps over 4 track days · Eric Wood's corner-evaluation framework

By the book, Palmer has four roll-speed corners. You ride eleven of thirteen as drive corners.

The framework sorts every corner by two questions — is its minimum speed above the bike's threshold-of-traction speed, and does the exit feed a straight worth driving for — into a roll-speed, drive, or entrance line. This page runs that on Palmer's measured geometry, then puts your actual lines beside the prescription, corner by corner.

The circuit, sorted

roll speeddriveentrance

How the sorting works

V_min vs V_TOT. If the corner's minimum speed is above the speed at which the bike can spin the rear at full lean, horsepower can't hurt you there: take the roll-speed line — maximum radius, apex at the arc's centre, slowest point at the apex, light trail brake to it. Below V_TOT, the throttle becomes the limiting input and the corner is about the exit.

What follows the exit. A long straight rewards exit speed for many seconds: drive line — enter wide, apex late, gas-on-bike-up to the outside edge. Another turn right after gives the exit nothing to cash in: entrance line — brake deep, apex early, slowest point after the apex, set up the next turn.

The advantage runs to the next corner you brake for — read off the speed profile, not off V_TOT. Extra exit speed is banked up to the next braking point and no further, so what matters is where the median profile next dips. Out of T1 it does not dip for T2 (1 km/h), so T1's exit is carried through the T2 kink and pays to the 10 km/h roll-off before T3 — 15 s. T11's exit is carried through T12 and T13 onto the main straight, 29 s. T4→T5 is a 10 km/h dab: speed given away, so T4 is its own corner and its exit pays nothing. An earlier version of this page let a roll-speed corner extend the chain instead — T7 "paying through T8" for 13 s — which the profile contradicts: you shed 26 km/h for T8, so T7's exit pays 3.8 s, and by Wood's rule a braking entry with under 5 s of run after it is an entrance corner.

A connecting turn is a roll-speed corner whatever its V_min. Wood's chicane clause — a turn the rider neither brakes for nor drives out of — fires once at Palmer: T12, reached without braking out of T11 and left with 15 km/h of rise before T13. T2 and T3 are roll-speed corners on V_TOT anyway; T13 is driven onto the straight.

The thresholds are ours, not the framework's: "long straight" = at least 5 s to the next braking corner on the median speed profile, "brakes" = the profile dips at least 12 km/h before the next slowest point, "carried" (the exit still paying through the next corner) = a dip under 5 km/h, "connecting" = no braking in and under 20 km/h of rise out, "linked" = the next turn-in within 80 m. V_TOT per bike comes from the app's settings: the ZX-4RR at the framework's 400-class figure (80 km/h), the Tiger Sport 660 at 75 km/h — not the 600-class 120, because its measured wide-open-throttle curve sits in the same power-to-weight band as the ZX (the app now measures this per session; see the last section). The 2025 day was ridden on the Tiger and is scored against its strategies; 2026 on the ZX-4RR.

Corner by corner: the book vs your line

"Your line" is the median of all 92 clean-pace laps, read in the road frame: where your apex sits against the arc's centre, and where your slowest point sits against your apex. A drive-shaped line apexes past the centre with the slowest point before it; roll-shaped apexes at the centre with the slowest point there; entrance-shaped apexes early with the slowest point after. Lean-corrected to the contact patch and aligned for each day's GPS bias.

Each corner

Strip: your median line through the corner per day, inside limit at the bottom, outside at the top; the tinted band is the tarmac, warm ticks are kerbs, the dashed vertical is the arc's centre. 2025-07-28 dashed, 2026-08-02 bold, the other two days grey. Targets: the bar is the share of laps meeting the framework's target for the corner's strategy; the number is how much faster (or slower) those laps exited than the laps that didn't, relative to the day's median — the framework's prescription checked against your own data. "—" = fewer than 3 laps on one side.

Does the book pay? All corners pooled

Per target: the median across corners of "exit speed of laps meeting it minus laps not meeting it" (each relative to its day), and how many corners it helps or hurts by at least 1 km/h. Read this before taking any single corner's number seriously.

What this says

Grain of salt

V_TOT is the softest number in the whole page. The app now measures it two ways per session: an upper bound — the speed above which wide-open throttle gives less drive than you already apply at full lean (ZX-4RR ≤ 92–127 km/h across the 2026 sessions, tight only when you were driving hard at lean) — and a modelled value assuming a warm sport tyre at μ 1.0 (ZX-4RR ≈ 66–73 km/h, Tiger ≈ 72). Both bracket the 80 / 75 used here. The corners on the edge are T13 (V_min 78–81) and T8 / T10 (84–85): on the modelled value they are roll-speed corners with margin, on the bound they are not.
The arc centre is not a recommendation. It is the framework's reference point ("apex 5–20 ft later than the centre" for a drive line). On a 150° hairpin the centre is mid-arc by construction; what a coach would call the apex may legitimately sit elsewhere.
Exit speed relative to the day's median is a blunt response. GPS is good to ~1 km/h; most per-target deltas are 1–3 km/h on 10–40 laps. The pooled table is the honest read; single-corner numbers under ±1 km/h are noise.
Body position, bar pressure, vision — the framework's other half — are invisible to a GPS/IMU/OBD logger. What is measured here is the consequence: lean re-added after the crack, throttle closures, roll-off shape.