How Fuzz reviews
Most of our reviews are research-led: we haven't played with the product ourselves. We gather the best available evidence, check it and explain what it means for you. Here's exactly how.
The sources we use
- Manufacturer specs. Head size, weight, balance, string pattern, beam, recommended tension.
- Independently measured specs. Retailers such as Tennis Warehouse measure strung weight, balance, swingweight and stiffness. We use single values with credit, and show both figures when sources disagree.
- Published playtests. Reviews by testers who actually hit with the product. We summarise them in our own words, quote a few words at most and link to the original.
- Owner reviews. We read them and count recurring themes ("more power than the last version: 4 of 8"). We say how many reviews there are, where they're from and when, so you can judge the sample.
- Our readers. Reviews posted on Fuzz, checked under our reader review guidelines.
The Fuzz Score
Each product is scored from 1 to 5 in the categories that matter for that kind of gear, and the overall score is the average, rounded. Every category shows the evidence behind it.
| Product | Categories |
|---|---|
| Tennis racquets | Control, comfort, stability, spin, power, forgiveness |
| Pickleball paddles | Control, power, spin, sweet spot, durability, value |
| Padel rackets | Control, power, comfort, sweet spot, durability, value |
| Shoes | Support, comfort, durability, fit, weight, value |
| Ball and stringing machines | Performance, features, ease of use, build, value |
A score reflects the product for its intended player. A 2 for forgiveness on a control racquet isn't a flaw for an advanced player; it's a warning for a beginner.
"What the specs mean for you"
These sliders are our estimate, worked out by placing a product's specs (for racquets: head size, weight, balance, swingweight, stiffness, string pattern) against every other product in our database. They're a guide, not a measurement.
Similar products
"Closest alternatives" are the products with the nearest specs, calculated automatically, then checked by an editor so the list makes sense.
Research-led vs hit-tested
When we do get a product on court, the review says Hit-tested and tells you who tested it, how long, at what level and with which strings. Until then it says Research-led. We never blur the two.
How we use AI
- We use AI tools to help collect sources, sort owner reviews into themes and draft summaries.
- An editor checks every fact, number and quote against the original source before publishing.
- Fuzz's audio verdicts are read by an AI-generated voice. The script is the reviewed text.
- We never use AI to invent reviews, testers, quotes or product photos.
Keeping reviews current
Every review shows when its evidence was collected and when it was last updated, with a changelog. We recheck a review when a new version is released, when the price moves a lot or when new playtests and owner reviews appear.