Every gallery card carries up to three small tags — Status, Speed and Timing — read off the market's fitted S-curve. Status is absolute (80% share and the year 2035). Speed and Timing are relative — a market is compared with the other markets of its vehicle type, so trucks are measured against trucks, not against cars.
Only base variants define a scale. For cars that is the main new-car curve (Whole) of every market, one per country. Car subsets — private and company keepers, rental and non-rental, used imports, and other car slices — do not shape the scale; they are placed on Whole's scale, like a dot on Whole's boxplot.
Vans, trucks (HDV), buses, two- and three-wheelers and so on are their own reference groups. A subset of one of them is compared against that parent — Canada's pickup trucks, for example, are placed on the Vans scale.
A reference group needs at least 4 markets. Smaller groups — and their subsets — get no speed or timing tag.
Right under the search box, country and chart types, the gallery offers the tags themselves as chips, in four groups — Fit (Reliable, Volatile / unreliable, No transition), Status (Done, On track, Behind), Speed (Fast, Avg, Slow) and Start (Early, Mid, Late). Start is the Timing tag; Mid is the middle half, which carries no pill on the cards.
Nothing picked shows everything. Within a group any picked tag matches; across groups all must match — Fast and Early means fast and early, Fast and Avg means either.
Every chart has exactly one Fit value. No-transition charts appear only under No transition, never as Reliable, and Volatile / unreliable or No-transition charts carry no Status, Speed or Start — they drop out as soon as one of those is picked. So does a market whose group is too small for a scale, once a Speed or Start value is picked; it is never counted as Mid.
Each chip shows how many charts it would give with the other groups' picks applied; chips that would give none are dimmed. Picked tags also appear as removable chips above the results.
Status — where the fitted curve crosses 80% BEV share of new registrations. Done (already crossed), On track (by the end of 2035), Behind (after 2035). The same for every vehicle type.
Speed — the slope of the fitted curve at its steepest point. Fast = top quarter of the reference group, Slow = bottom quarter, Avg speed = the middle half.
Timing — the year the fitted curve crosses 10% BEV share. Markets still below 10% count with their projected year. Early starter = earliest quarter of the reference group, Late starter = latest quarter; the middle half carries no timing tag.
{"Curves that show no transition get a single No transition tag, and fits that can't give a reliable year yet a single pill": 'Volatile curve when only the re-fit swing is too large (the curve keeps moving with every new month, nothing is wrong with it), Unreliable fit when it has collapsed (a near-vertical step the data has not reached, or no valid S-shape) or its data are stale. Neither counts towards any scale.'}
One row per market and variant from params.csv (the S-curve parameters, the latest data month, the observed BEV share and the re-fit stability).
| # | When | Then |
|---|---|---|
| 1 | fitted 80% crossing before today | Done |
| 2 | fitted 80% crossing by end of 2035 | On track |
| 3 | fitted 80% crossing after 2035 | Behind |
| 4 | speed at or above the group's 75th percentile | Fast |
| 5 | speed at or below the group's 25th percentile | Slow |
| 6 | speed in between | Avg speed |
| 7 | 10% year at or below the group's 25th percentile | Early starter |
| 8 | 10% year at or above the group's 75th percentile | Late starter |
| 9 | 10% year in between | no timing tag |
Live from params.csv with the gallery's own tag code. Each dot is one market's reference curve (hover for its name); the box is the middle half, the braces mark the tag bands with their cut-off.