Scoring Methodology
Every number on this site is calculated the same way, every time. This page explains the four formulas we use — what they measure, why we chose the weights we did, and exactly how the math works.
1Rating Scale Conversion
Ratings are stored in the database as whole numbers on a 1–10 scale (one integer per half-star, where 1 = 0.5 stars and 10 = 5 stars). Every value you see on the site — star displays, score callouts, JSON-LD structured data — is converted to the familiar 0.5–5 star scale before display.
Formula
display_score = stored_value ÷ 2
| Stored value | Displayed as |
|---|---|
| 1 | 0.5 ★ |
| 2 | 1.0 ★ |
| 5 | 2.5 ★ |
| 7 | 3.5 ★ |
| 10 | 5.0 ★ |
2Per-Category Averages
Each of the six rating categories is averaged independently across approved reviews for a provider. No weighting is applied at this stage — it is a straight arithmetic mean, rounded to one decimal place.
Formula (per category)
avg_category = SUM(latest approved ratings in that category) ÷ unique_reviewer_count
The six categories are: Pricing, Performance, Reliability, Support, Ease of Use, and Value for Money. These averages feed directly into the Weighted Score below.
3Weighted Score
The Weighted Score combines all six categories into a single quality number for a provider. Categories are not treated equally — Reliability carries the most weight because server downtime has immediate, measurable business impact for customers. Performance and Value share the next tier because they directly affect day-to-day usability and cost efficiency. Support and Pricing follow, then Ease of Use.
Formula (on the internal 1–10 scale)
overall = (reliability × 0.25)
+ (performance × 0.20)
+ (value × 0.20)
+ (support × 0.15)
+ (pricing × 0.12)
+ (ease_of_use × 0.08)
The weights sum to 1.00 (100%), so the result stays on the same 1–10 scale. Divide by 2 to get the 0.5–5 display value. On the providers page, sorting by Weighted Score ranks providers by this value descending.
| Category | Weight | Rationale |
|---|---|---|
| Reliability | 25% | Downtime costs money. This is the most critical factor for any production workload. |
| Performance | 20% | CPU, disk I/O, and network speed affect every application running on the server. |
| Value for Money | 20% | Price-to-performance ratio matters as much as raw speed for most buyers. |
| Support | 15% | When something breaks, response time and quality of support are critical. |
| Pricing | 12% | Absolute cost is a factor, but Value for Money captures it better as a combined metric. |
| Ease of Use | 8% | Control panel and setup experience matters, but most VPS buyers are technical users. |
| Total | 100% |
4VHR Rank
VHR Rank is the default sort order on the providers page (the VHR Rank option). It solves a real problem: a provider with a single 5-star review should not outrank a provider that has collected hundreds of balanced, real-world reviews.
The formula multiplies the Weighted Score by the natural logarithm of (review count + 1). The logarithm grows quickly at first — the first review makes a big difference — then levels off so additional reviews keep mattering, but never dominate the signal. A genuinely poor provider cannot climb the rankings simply by accumulating reviews.
Formula
rank_score = overall_score × ln(review_count + 1)
ln = natural logarithm (base e ≈ 2.718).
The + 1 prevents ln(0) when a provider has no reviews.
Example scores
| Provider | Weighted Overall (1–10) | Review count | ln(count + 1) | VHR Rank |
|---|---|---|---|---|
| A — one perfect review | 10.0 | 1 | 0.693 | 6.93 |
| B — solid, well-reviewed | 8.5 | 50 | 3.932 | 33.42 |
| C — great, lots of reviews | 9.0 | 200 | 5.303 | 47.73 |
| D — average, high volume | 6.0 | 500 | 6.216 | 37.30 |
Provider A's single perfect review cannot compete with sustained quality at volume. Provider D has many reviews but an average score — volume alone does not win.
Why these choices?
Review sites fail in two common ways. The first is treating every category as equally important: a host with great marketing collateral and a slick control panel can score as highly as a genuinely reliable host, which misleads buyers chasing real-world performance. Weighted scoring fixes that.
The second failure is letting small sample sizes dominate rankings. A freshly-launched provider with two enthusiastic early reviews should not appear above a host with years of verified customer feedback. The logarithmic volume adjustment addresses that without completely discounting small providers — they still rank, just more conservatively until the review base matures.
The weights for the Weighted Score were chosen based on the most commonly cited priorities in VPS hosting purchase decisions: uptime and reliability first, raw performance and value second, then support responsiveness. They are not arbitrary — but they are not permanent either. If review data consistently shows that buyers weight categories differently, we will update the formula and document the change here.
What we do not do
- We do not manually adjust scores. Every number is produced by the same formula run against the same review data.
- We do not include unverified or unapproved reviews in any score or ranking. Only moderated reviews count.
- We do not count a reviewer's historical opinions against their current one. Only the most recent approved review per reviewer per provider influences the aggregate score — so if a provider improves, returning reviewers can reflect that change.
- We do not accept payment to alter a provider's score or ranking position. Affiliate relationships, where they exist, are disclosed in the footer and have no effect on the formulas.
- We do not hide the methodology. This page exists so you can verify every number you see on the site.
Questions about the methodology? Something look off? Browse the raw reviews and verify the numbers yourself — every approved review that feeds a score is public.