This page is the whole method, in public. If you disagree with a score, you should be able to see exactly how we got there.
The four criteria
Every product is scored on the same four criteria, each out of 100. Nothing gets a criterion that others do not, which is what makes scores comparable across the site.
Specs for the price – 35%
Raw capability measured against what it costs, and against everything else at that price.
For wheel bases: peak torque, slew rate, encoder resolution, motor type, build materials. For pedals: sensor type, load cell rating, adjustability, pedal face construction. For rigs: profile dimensions, wall thickness, adjustment range, tested rigidity.
A product does not score well here for being powerful. It scores well for being powerful for the money. A 21 Nm base at $1,800 can score lower than a 9 Nm base at $600.
Compatibility – 25%
What actually works, today, with what you own.
Platform support (PC, PlayStation 5, Xbox Series X|S), quick-release standards, mounting patterns, adapter requirements, driver and firmware maturity, and whether the manufacturer’s software is stable on both Windows and macOS.
We weight this heavily because it is the single most common cause of buyer’s remorse in this hobby. A base that cannot talk to your console is worth zero to you regardless of its torque figure.
Community consensus – 25%
What people who actually own it report, over time.
We read owner reports across community forums, subreddits, manufacturer support channels and verified purchase reviews, then filter for signal: repeated failure modes, thermal behaviour after long sessions, build variance between units, problems that only appear after six months.
We always publish the sample size. A consensus score built on 40 reports is not the same as one built on 400, and you deserve to know which you are reading.
We discount launch-week enthusiasm and single dramatic complaints equally. What we are looking for is the pattern that repeats.
Support and warranty – 15%
Warranty length and what it actually covers. Spare parts availability. Typical RMA turnaround by region. Whether support answers.
Almost no review site scores this. It is also the thing that decides whether a $1,000 purchase is a five-year investment or a two-year one.
The formula
Overall = (Specs x 35 + Compatibility x 25 + Consensus x 25 + Support x 15) / 100 / 10The result is a score from 0 to 10, shown to one decimal place.
There is no manual override. We cannot type a score we like into a box. The number is computed from the four criteria every time a review is saved. If we think a score is wrong, we have to argue with the criteria, which is the point.
If a criterion cannot be assessed for a particular product, its weight is redistributed proportionally across the criteria we can assess, so an incomplete review is not unfairly penalised.
What the scores mean
| Score | Verdict | What it means |
|---|---|---|
| 8.5 – 10 | Highly recommended | Best in its class. Buy with confidence. |
| 7.5 – 8.4 | Recommended | Strong choice with known trade-offs we spell out. |
| 6.5 – 7.4 | Think twice | Works, but something specific should give you pause. |
| 5.0 – 6.4 | Better options exist | Not broken. Just beaten at the same price. |
| Below 5.0 | Don’t buy this | We would not spend our own money here. |
We expect a healthy spread across this range. A site where everything scores above 8 is not reviewing, it is recommending.
Evidence tiers
Every review is labelled with what it is built on:
Spec analysis. We have not handled this product. Built from manufacturer specifications, technical documentation, teardowns where available, compatibility records and filtered owner reports. These reviews never describe how something feels in the hand, because we do not know.
Hands-on tested. We have used the product ourselves. The review states how long and on what. These reviews carry photographs we took.
The four criteria and their weightings are identical for both tiers. Hands-on testing changes how confident we are in the inputs, not what we measure. This is deliberate: it keeps every score on the site directly comparable, which is what lets you sort a category by rating and get a meaningful list.
Where products come from
When we test hands-on, the product is either bought by us or supplied by the manufacturer for review. Where a manufacturer supplied it, the review says so.
Supplied products are not returned in exchange for a score, and no manufacturer sees a review before it publishes. We do not send drafts for approval, and we do not accept conditions attached to a loan. A manufacturer who does not like their score is welcome to say so in public.
Affiliate links do not affect scores
Some links on this site earn us a commission. The score is computed from the four criteria before any link is added, and commission rates are not an input to any criterion. Some of the lowest-scored products on this site pay the highest commissions. See our affiliate disclosure.
When we update a review
We revisit a review when a firmware or driver update changes compatibility, when a hardware revision ships, when a price change moves a product into a different value bracket, or when enough new owner reports accumulate to shift the consensus score.
Updated reviews show the revision date at the top. Where a score changes, we say what changed and why.
If you think we are wrong
Tell us. Specifically:
- A specification is incorrect – send the source and we will correct it, usually within a day.
- A compatibility claim is out of date – firmware moves fast, we will not always catch it first.
- A consensus score does not match your experience – this one is genuinely useful, since it is the criterion most likely to drift.
Corrections are made openly, not quietly. Email corrections@wheelverdict.com.