How we rank
No order gets set first and justified afterward. Products are scored axis by axis, then sorted by the average of those axes. The whole calculation is on the page.
How the total is calculated
A total is the plain average of the axis scores its category uses. No axis is weighted more than another. The moment you weight them, that ratio becomes the editor's taste wearing a number as a disguise — better to lay the axes out and let you apply your own priorities.
Which means totals don't travel across categories. An 8.2 in Appliances and an 8.2 in Books were computed from entirely different axes.
What each axis measures
- Performance
- How well it does the core job
- Value
- What the same money buys elsewhere
- Durability
- Failure and fade reports past the one-year mark
- Ease of use
- Setup, daily operation, cleaning, maintenance
- Quietness
- Operating noise — a higher score means quieter
- Design
- Build quality and how it sits in a room
- Support
- Warranty handling, parts availability, software updates
- Depth
- How much ground it covers, and how deep
- Freshness
- How current the information is
Which axes apply where
Categories don't share one axis set. Asking a vacuum how current its information is makes no sense, and neither does asking a book how quiet it is.
| Category | Axes |
|---|---|
| Appliances | Performance · Value · Durability · Ease of use · Quietness |
| Tech | Performance · Value · Durability · Ease of use · Design |
| Home | Value · Durability · Ease of use · Design |
| Kitchen | Performance · Value · Durability · Ease of use · Quietness |
| Outdoors | Performance · Value · Durability · Design |
| Health (not yet stocked) | Depth · Value · Ease of use · Support |
| Books (not yet stocked) | Depth · Value · Freshness |
Why there are no prices on this site
Amazon prices move by the hour, and Amazon's terms only permit displaying prices pulled live from their Product Advertising API, which we don't run. So prices are gone from the data model entirely — there is no field to fill in, and the build fails if someone adds one. Buying buttons open the live listing instead.
How the head-to-head pages get made
Pairing everything with everything balloons into thousands of matchups nobody is torn between. Pairs are only generated within the top 4 of a single ranking — the comparisons people actually weigh.
What we don't do
- We don't say we tested it. Nothing here is measured by us. Figures are attributed — manufacturer spec sheets, independent measurements like RTINGS, or the consensus among owners.
- We don't aggregate customer ratings. We collect no user ratings, so no aggregate-rating markup leaves this site. Every number you see is an editorial score, not an Amazon star average.
- We don't leave the weakness box empty. Every product carries a "Skip it if" list, affiliate link or not. The schema refuses to accept a product without one.
- We don't use Amazon's logo or trade dress. No marks, no orange-button lookalikes, nothing that implies Amazon endorses this site.
Editorial principles
- 01
Products get recorded first. Ranks come after.
Every product is entered as a standalone record — specs, axis scores, buying links — independently of any ranking. A rank is that data sorted. We never fix the order first and backfill reasons for it.
- 02
The axes are published.
You can see which axis earned which score on every page. We don't show a single total and hide the arithmetic behind it.
- 03
Weaknesses stay in.
Every product carries a 'Skip it if' list, affiliate link or not. If there's a situation where you shouldn't buy it, that situation is named.
- 04
We don't claim tests we didn't run.
Nothing here is measured in a lab of ours. Numbers are attributed — manufacturer spec sheets, independent reviewer measurements like RTINGS, or the consensus in owner communities.
- 05
We don't print prices.
Amazon prices move hourly, and a price copied into a page is wrong by the time you read it. We link you to the live listing instead of quoting a number that has already expired.