Data & sources

Where the numbers come from

All grip and longevity scores come from a single spreadsheet, Bike Tyres Comparison.xlsx, compiled by Gerhardt Kirchner from public sources — independent tyre tests, owner long-term reports, community threads and video reviews.

They are judgements, not measurements. No braking distance was instrumented, no wear rate was measured under controlled conditions, no lap time was taken. They are one experienced rider's synthesis of published opinion, expressed to a half-point. That is a legitimate way to build a comparison and a poor way to imply precision. Treat a 0.2 difference as noise.

The biggest problem with this data

The source workbook's notes carry citation markers numbered 1 to 53. There is no source list anywhere in the workbook, and there are zero hyperlinks in any of its four sheets — this was checked cell by cell.

So every score is attributable to the compiler, but no individual claim can be traced to a URL you could go and read. Recovering that source list is the single highest-value outstanding task for this dataset, and until it happens the honest description of the provenance is "one rider's compilation" and nothing stronger.

The citation markers are left visible in the quoted notes on each tyre page rather than being scrubbed, because hiding them would hide the problem.

Scope — the caveat that matters most

A tyre that is confidence-inspiring on a 220 kg middleweight with long-travel suspension may behave quite differently on a 260 kg flagship or a 150 kg dual-sport. Nothing in this dataset accounts for that. If your bike is much heavier or lighter than the reference, treat every ranking as weaker evidence rather than as an answer.

Why "tarmac at 180 km/h" is excluded

The source workbook scores fifteen conditions. One of them, tarmac grip at 180 km/h, is excluded from every ranking on this site.

In that row, several tyres score 0.5 or 1.0. Those are not grip assessments. Those tyres are speed-rated below 180 km/h — a Q symbol means a 160 km/h maximum, an R symbol means 170 km/h. The row is recording a rating breach as though it were a grip score.

Presenting that as grip on a public page would be unsafe: it invites a reader to conclude that a Q-rated knobby merely "handles poorly" at 180 km/h, when in fact the tyre is being operated beyond its rated limit. So this site lifts speed rating out into its own section, as a hard constraint with an explanation, and drops the category from scoring.

The workbook's own weighting already set this category to zero, so no ranking changes as a result. The change is in how it is presented.

Is the data good enough? Use case by use case

These verdicts come from the data audit written before this site was built. They are deliberately blunt.

Tyres with limited or conflicted evidence

Flagged by the compiler in the source workbook, reproduced here.

What is missing entirely

• Price. There is not one tyre price in the source workbook. Price ranking is therefore not offered at all, rather than being faked.
• Comfort, noise and rolling resistance — which matter disproportionately to touring riders.
• Puncture resistance, carcass stiffness, tube versus tubeless, bead retention — which is why the off-road-biased use case is rated thin.
• The date the scores were compiled. Not recorded anywhere in the workbook. It has not been invented; the workbook's last-edited date is shown instead and labelled as such.
• A second reference bike, which would test how far these scores travel.
• User-review sentiment. If it is ever added it will be labelled opinion and kept visually separate from specifications. It is not here today.

Research added on top of the workbook

Every specification on this site is attributed to a named source with a retrieval date, and manufacturer sources are marked as such and preferred over retail listings. Where a manufacturer does not publish a figure, the figure is shown as "no data" rather than being filled from a weaker source without saying so.

How this data stays current

• Versioned JSON in the repository. The data lives as data/tyres.json, data/bikes.json and data/meta.json, each carrying a schema version, a build date and per-datum source attribution. Diffable in git, which a spreadsheet never was.
• A committed build script. build/xlsx-to-json.py regenerates the data from the source workbook, so the transformation is reproducible and reviewable rather than a one-off act.
• A monthly review prompt is drafted in the repository at build/MONTHLY-REFRESH-TASK.md. No automatic schedule has been created — that is the owner's decision to make.
• Corrections by email. Every tyre page carries a correction link that pre-fills the tyre name and data version. No form, no backend, no data collection.

Report a correction