How We Rate Hosts
The five categories every review scores, how the overall score is calculated, and what we do not do.
Five categories, one to five stars
Every review scores the same five things. There is no weighting and no secret adjustment — a provider's overall score is the plain arithmetic mean of every approved review's own mean.
| Category | What reviewers are asked about |
|---|---|
| Performance | Tick rate and frame stability, CPU clock, disk speed, how it behaves under a full load |
| Support | How fast someone replied, whether they understood the problem, and how they treated you |
| Value for money | What you actually got for the price, including the renewal price |
| Control panel | Whether you can do your own job without a ticket: mods, backups, console, files |
| Uptime & reliability | Unplanned downtime, node stability, how DDoS attacks were handled |
What the numbers mean
- 4.5 – 5.0 — Excellent. Consistently recommended by people who stayed.
- 4.0 – 4.4 — Great. Minor complaints, no pattern of serious ones.
- 3.2 – 3.9 — Good. Works, with real trade-offs worth knowing about.
- 2.2 – 3.1 — Mixed. Experiences differ sharply; read the reviews, not the number.
- Below 2.2 — Poor. A consistent pattern of problems.
What we do not do
- We do not weight recent reviews more heavily. A pattern over two years is more informative than a good week.
- We do not remove low scores at a provider's request.
- We do not publish an aggregate score until a provider has real reviews. An unrated provider is shown as unrated, not as zero.
- We do not let our own editorial opinion move the community score. Where we have written an assessment, it sits beside the community score, never inside it.
Specifications
A provider page only shows specifications once a researcher has verified them against the provider's own published information and dated the check. Until then the page says so plainly rather than showing a plausible-looking blank.
Last updated