About Tyre Reviews
How We Score Our Tyre Tests
From our winter 2026 tests onwards, Tyre Reviews scores its own tests with a new method in which each published weight is the share of the overall result that discipline actually decides. This article explains what was wrong with the old method, how the new one works, the data behind it and its limitations.
On this page
From our winter 2026 tests onwards, Tyre Reviews scores its own tyre tests with a new method. Under it, the weight we publish for each discipline is the share of the overall result that discipline actually decides. This article explains what was wrong with the old method, the options we considered, how the new method works, the data behind it and its limitations.
The way we test is unchanged and is described in How We Test Tyres. This article is only about how the measurements are combined into one overall score.
In brief
- Each result is still scored against the best tyre in the test.
- Each discipline then counts for its published weight in every test, rather than for however widely its results happen to spread.
- The weights come from a fixed profile for each kind of tyre, and the conversion rates are published and never changed once tests have been scored with them.
- Tyres within 0.4 points of the best tyre in their group share its place.
- Tests published before the change keep the scores and finishing order they were published with.
What the overall score is for
Every Tyre Reviews test measures each tyre in a series of disciplines: braking and handling in the dry and the wet, aquaplaning, snow and ice where relevant, noise, comfort and running costs. The overall score combines those results into one number and a finishing order, using a published weight for each category and for each discipline within it. For road tyres we weight wet performance at least as highly as dry, and braking above handling laps, because they matter most for safety.
The overall score only applies to our own tests. When we publish a third-party test, its finishing order is the publication's own.
The problem with the old method
Until now, each result was expressed as a percentage of the best tyre in the test. A tyre that stopped in 46 m when the best stopped in 44 m scored 95.7%. Those percentages were multiplied by the weights and added up.
The weakness is that a 1% gap means very different things in different disciplines. Across our published tests, the middle half of a field typically spans about 1% on a dry handling lap, but about 6% in wet braking, 7% in rolling resistance and 16% in tread life. Disciplines that naturally spread widely therefore produced large percentage gaps and counted for more than their weight, while lap times produced small gaps and counted for less.
We measured this across 36 of our published tests by comparing each discipline's weight with its real impact: how much of the difference between the tyres it actually created.
The old method, 36 of our published tests
How much each discipline really decided, as a multiple of its weight
At category level, Value had 1.8 times the influence its weight gave it, while Dry had three quarters of its share. This had two consequences.
Neither fits a scoring system that is meant to be transparent and fair.
An example
Take two tyres in a test where wet braking carries a 20% weight and tread wear 5%, so wet braking is meant to count four times as much.
In our tests both are about a typical gap for their discipline.
Old method
New method, in a typical field
What we considered
Simple, but a 0.1 m and a 5 m braking deficit would cost the same, which throws away the size of the difference.
Some magazines grade each discipline on a fixed scale. That needs fixed grade boundaries for every discipline, vehicle, surface and season. Our conditions and fields vary from test to test, so the boundaries would need constant revision, and a tiny difference at a boundary moves a tyre a whole grade.
Each discipline is converted at a published rate, so that a typical gap counts the same everywhere. This fixes the problem in a typical test and is the base of the new method. On its own, though, a test in which one discipline spreads unusually widely would still let that discipline decide more than its weight, which is the problem we set out to remove.
This makes every weight hold in every test. Without a limit, it would also magnify disciplines in which the tyres finished almost level, so measurement noise could decide a result. We adopted it with a limit.
How the new method works
- Percentage of the best
As before, each result is expressed as a percentage of the best non-reference tyre in the test.
Lower is better
such as braking distance or lap timebest ÷ result × 100Higher is better
such as aquaplaning speedresult ÷ best × 100 - Discipline score
Each discipline scores:
100 - rate × (100 - percentage)and never less than zero. At a rate of 2, a tyre at 98% of the best scores 100 - 2 × 2 = 96.The best tyre scores 100 and every 1% behind it costs the rate in points. Because a discipline score cannot go below zero, no discipline can ever cost a tyre more than its weight.
- The rate
Each discipline starts from a published typical rate, set from how widely it spreads across our archive. The narrower its typical spread, the higher its rate, so that a typical gap counts the same in every discipline.
For each test the rate is then adjusted to how widely that discipline actually spread among the tyres tested, so that its share of the separation between the tyres equals its weight. A rate can be lowered without limit, so a wide spread can never make a discipline count for more than its weight. It can be raised at most to double the typical rate, so a near tie cannot be turned into a decider.
Wider spread than usual
Dry handling laprate 4 2Suppose that in one test the dry handling lap times spread twice as widely as usual, because the field mixes very fast tyres with much slower ones. The dry handling rate for that test is halved, and dry handling decides its weight rather than twice its weight.
Closer finish than usual
Wet brakingrate 0.75 at most 1.5If the same tyres finished unusually close together in wet braking, the wet braking rate is raised so that wet braking still counts for its full weight.
- Overall score
The discipline scores are multiplied by their weights and added up, and the finishing order follows the overall score.
overall score = sum of (discipline score × weight) - Joint places
Tyres within 0.4 points of the best tyre in their group share its place, and the next tyre takes its own position, so three joint winners are followed by fourth. The group is measured from its best tyre rather than from the tyre directly above, so a run of small gaps cannot chain a whole field together.
Illustration, with made-up scores
Tyre Score Place Tyre Abest in its group 87.4 Joint 1st Tyre B0.1 behind A 87.3 Joint 1st Tyre C0.3 behind A 87.1 Joint 1st Tyre D0.5 behind A, so a new group 86.9 Joint 4th Tyre E0.2 behind D 86.7 Joint 4th Tyre F0.5 behind D, so a new group 86.4 6th Tyre F is only 0.3 points behind Tyre E, but 0.5 points behind Tyre D, the best tyre in that group, so it does not share 4th.
In a typical summer test
0.4 pointsin the overall score is worthabout 0.6 son a 75-second dry handling lapor about 0.8 mon a 30 m wet stopWe judge differences that small too close to call, so we do not separate tyres on them. The threshold is part of each rate set and is published with it.
Spread, and with it each discipline's share of the result, is measured on the middle half of the field, so one exceptionally good or poor tyre cannot move the rates.
Typical rates
The current rates, set 2026-10, published on 10 October 2026, were calibrated from our own published tests wherever at least ten of them include a discipline, and otherwise from all published tests in our database with at least five tyres. They are rounded to a short fixed scale, and a rate of 1 corresponds to a typical spread of 4.6%.
Because each test's rates are then adjusted to its own field, the typical rate mainly sets how far a discipline's rate can be raised, which is at most double. The rates each test actually used are shown on its page, rounded to two decimals, and the scores are calculated with those rounded figures.
| Discipline | Typical spread | Typical rate |
|---|---|---|
| Dry | ||
| Dry braking40 of our tests | 3.5% | 1.25 |
| Dry handling lap41 of our tests | 1.1% | 4 |
| Subjective dry handling35 of our tests | 5.7% | 0.75 |
| Wet | ||
| Wet braking40 of our tests | 5.9% | 0.75 |
| Wet handling lap40 of our tests | 2.3% | 2 |
| Subjective wet handling33 of our tests | 7.5% | 0.6 |
| Straight aquaplaning37 of our tests | 2.9% | 1.5 |
| Curved aquaplaning27 of our tests | 7.9% | 0.6 |
| Snow | ||
| Snow braking28 of our tests | 3.0% | 1.5 |
| Snow handling lap27 of our tests | 2.5% | 2 |
| Snow traction27 of our tests | 4.3% | 1 |
| Comfort | ||
| External noise25 of our tests | 1.8% | 2.5 |
| Subjective comfort37 of our tests | 5.0% | 1 |
| Value | ||
| Rolling resistance36 of our tests | 7.0% | 0.6 |
| Tread wear57 tests from all publications | 15.7% | 0.3 |
| Running cost per 1,000 km44 tests from all publications | 14.9% | 0.3 |
| Fuel consumption27 tests from all publications | 1.9% | 2.5 |
A set of rates is never changed once tests have been scored with it. When we recalibrate, the new rates are published as a new set and apply to new tests only, so published results do not move.
Weights for each kind of test
Each test starts from the weight profile for its kind of tyre. If a test does not include a category, such as snow, that share is spread across the others in proportion. The weights used in each test are shown on its page.
Category weights in each profile
How Value works
Value works differently, because what a test measures for it varies. Each Value discipline in a profile carries its own part of the Value weight: in the summer profile, tread wear, running cost per 1,000 km and rolling resistance each carry 30% of it and abrasion 10%.
A discipline the test did not run takes its part away, and the other categories are scaled up to fill the gap. A test that measured rolling resistance alone therefore gives Value only rolling resistance's part, so rolling resistance on its own never carries the weight of a full running-cost assessment.
The figures for rolling resistance alone are after the other categories have been scaled up. Within each category the profile also sets how the disciplines share its weight. In every road-tyre profile, braking carries more weight than handling laps in the dry and the wet.
Limitations
Each discipline is measured against the best tyre in that test, and each rate is adjusted to that field. A score of 95 in one test is not directly comparable with 95 in another. To compare tyres across tests, use the Tyre Reviews Score on each tyre's page, which compares results across tests through the tyres those tests share.
Because rates adjust to each field, 1 second on a lap or 1 metre in braking is worth a different number of points in different tests. Each test lists the rates it used, so its scores can always be reproduced.
When we re-scored our published tests with one tyre removed, the order of neighbouring tyres swapped in about 8 cases in 100, against fewer than 1 in 100 with fixed rates. That is the cost of making the weights hold in every test, and we think it is the right trade.
A tyre 0.29 points behind a group's best shares its place and one 0.31 points behind does not, although the difference between them is tiny. A fixed, published line is still fairer than deciding ties case by case.
A discipline in which the tyres finished almost level is raised at most to double its typical rate. In our published tests that limit applies to about one discipline in seven, which then counts for a little less than its weight, and the other disciplines for a little more.
That is where a discipline's share of the result equals its weight. Measured across the whole field, including the very best and worst tyres, a category's share differs from its weight by about 5 percentage points on average in our published tests, because an exceptionally poor result at the bottom of the field counts in full. The floor at zero still means no discipline can cost a tyre more than its weight.
They are made by our tester on a scale that is recalibrated each year in co-driving sessions with heads of tyre testing from premium manufacturers, as described in our methodology. The adjustment to each field also stops an unusually wide spread of subjective ratings in one test from counting for more than its weight.
For disciplines we have run in fewer than ten of our own tests, such as tread wear, running cost, fuel consumption and snow slalom, the typical spread comes from other publications' tests in our database. Only the spread is used, never their scores or weights. These rates will move to our own data in a later set once we have enough tests.
What stays the same
- How we test: the procedures, equipment, reference tyres and controls described in our methodology.
- Every measured result is published in full on each test, in its charts and its results explorer.
- Tests published before this change keep the scores and finishing order they were published with.
- Third-party tests show the publication's own finishing order.
Our commitments
- The typical rates, the joint-place threshold and the weight profiles are published on this page, and every test shows the weights and rates it used.
- A set of rates is never edited after tests have been scored with it. Any change to the method is published here as a new version before it is used, and applies to new tests only.
- If you want to check how any result was scored, contact us and we will go through the numbers with you.