Shaded stretches are where your draft ratio sat at or above the clean air threshold. The lower band is the ratio itself, blue below 1.0 (someone else doing the work) and red above (you doing more than a solo rider would need).
CritLab starts laps where your recording did, which is wherever you pressed start — the staging area, the back of the grid. The actual start/finish is a painted line on the road and nothing in a GPS trace identifies it, so if it matters, place it yourself.
Wedge length is time spent pointing that way; colour is the median draft ratio there. With a real crosswind one bearing should stand out.
One row per lap, one column per stretch of circuit. A column that stays red top to bottom is a place you were working harder than usual every single lap — that is a positioning habit, not bad luck.
Unlike the map, this grid is scaled to your own median for this race rather than to an absolute 1.0, because its job is comparing stretches to each other. If you sat in the bunch all evening, every sector is sheltered in absolute terms and only the relative differences carry information.
Exit cost is the work above CP spent getting back to entry speed after the apex. Bars are coloured by the draft ratio on the exit: a red bar means you were paying for that acceleration in clean air, with nobody to follow.
Skiba's differential model: the tank drains above CP and refills below it. Shaded bands are detected surges. Where the trace flattens near the floor is where you stopped having a sprint.
The battery is your W′ balance at this instant — the anaerobic reserve above critical power. Ten cells rather than a smooth bar, because the underlying W′ is an estimate and reading it to the nearest 100 J would be false precision. The dashed notch marks the lowest point of the whole race.
Tick races in the sidebar to include them. Every race is re-analysed with the settings on the right, so the comparison is like for like.