Methodology: anatomy of a snow blower rating

Most methodology pages describe a process in the abstract. This one does something more useful: it takes a single comparison card and dissects it from top to bottom, explaining where each visible element gets its data and, just as importantly, what that element is not. If you have a card open in another tab, read them side by side.

The rank number in the corner

The small numeral is a position within one comparison list, nothing more. It is assigned after sorting by the raw editorial score described below, descending, with ties broken alphabetically by product slug so the order is reproducible rather than arbitrary. It is not a ranking of quality across the whole category, it is not a ranking against machines we did not include, and it changes if you use the sort controls — which is the point. Position one means highest documented score among the listings in that specific comparison, on that specific day’s data.

The OUR RATING badge and the number inside it

This is the element people read first and misread most often. The number is a calibrated editorial index out of ten. It is not a test result, not a customer satisfaction figure and not a measurement of anything we did with a machine.

Behind it sits a raw score on a 100-point scale, produced by one of two published formulas. When the exact Amazon listing documents a clearing width, the spec-fit formula applies:

Documented spec fit = 100 × [0.40 × min(clearing width / 28 in, 1) + 0.30 × (two-stage or three-stage ? 1 : 0.5) + 0.30 × (documented throw distance ? min(throw distance / 50 ft, 1) : documented power source ? 0.5 : 0)], rounded to a whole number. Clearing width must be stated in the exact Amazon listing with recorded provenance; stage count and power source come from the same listing text. This index summarises what a listing documents about snow-clearing capacity — it is not a clearing test, a throw measurement, a noise reading or a durability judgment.

When it does not, we refuse to estimate one, and the card is scored on how completely the seller documented the listing instead:

Documentation completeness = 100 × [0.5 × min(feature bullets / 8, 1) + 0.3 × min(listing images / 5, 1) + 0.2 × (brand stated ? 1 : 0)], rounded to a whole number, computed from the stored Amazon listing record when the listing does not document a clearing width. It measures how completely a seller documents a listing — nothing about how the machine clears snow.

The badge label tells you which of the two produced the number, so a machine is never quietly credited with specifications its listing never stated.

Why the number is not the raw score

The raw 100-point score is not what you see. It is mapped onto a display scale, and the reason is honesty about precision: a two-point difference in a documentation score does not deserve to look like a meaningful gap.

The displayed editorial rating is the raw blended score divided by 10, shown to one decimal on a 0-10 scale. There is no catalog normalization, rank interpolation, forced winner, forced floor or artificial gap. Equal scores remain equal: the available data does not justify a distinction. Ranking uses the unmodified raw score descending, then listing slug for exact ties; slug is a stable identifier, not a quality signal. The same raw score always has the same display rating in every shortlist. This is not a physical test result or an Amazon customer rating.

Within a single comparison, displayed values strictly decrease wherever raw scores differ, so the visible order always tells the same story as the underlying arithmetic. Two listings with identical raw scores show identical displayed values rather than an invented tiebreak.

The star row beneath the badge

The stars are a visual restatement of the same displayed rating, on a five-point scale, and they carry no independent information. They are drawn from our editorial index, not from Amazon’s customer star average — a distinction worth holding onto, because the two sit close together on the card and mean entirely different things.

The review count link

Where it appears, this element is the one piece of the card that is genuinely other people’s opinion. It shows the public review count captured from the listing’s own Amazon product page and links there. When no public review data was captured for a listing, the card says so plainly rather than showing a zero or omitting the line.

Review data, where present, also feeds the rating:

Where Amazon publicly displays customer reviews for a listing (average star rating and review count captured from the public product page), the raw editorial score is 100 × [0.45 × (Amazon average / 5) × min(review count / 50, 1) + 0.35 × min(log10(1 + review count) / log10(1 + 5000), 1) + 0.20 × documentation score / 100]. The confidence factor min(review count / 50, 1) damps averages based on very few reviews, and the logarithmic volume term rewards large public review volumes, so widely and highly reviewed listings rank ahead of thinly reviewed ones. Listings with no captured public review data use the documentation score alone. This combined index reflects the public Amazon reputation and the listing documentation — it is not a physical test and not our own product testing.

The availability and price pills

These small pills are rendered by your browser, not baked into the page. They appear only when the underlying offer snapshot is still within its freshness window, and they disappear entirely when it is not — which is why you may see a card without them. We publish no price or stock claim in the static page, because a price we wrote last week is a price that is wrong today. The pills link you to the exact listing to check the current figure yourself.

The product image

Listing images are time-limited by the terms under which they are supplied, and our pages honour that deadline: an image past its window is removed by the same browser-side guard that handles the pills, leaving a placeholder. An image on a card is therefore current by construction, never an old screenshot.

What the whole card is not

Taken together, the card summarises what one Amazon listing documents about one machine, weighted by published formulas and public review data. It is not a clearing test, a noise measurement, a durability assessment, a reliability record or a value judgment. We have not run these machines. Stated throw distances are manufacturer figures obtained under conditions favourable to a long throw; your snow will be denser and your result shorter.

What we do when data is missing

Nothing. That is the whole policy. A missing clearing width is not inferred from a similar model, a missing throw distance is not estimated from engine size, and a shared brand name transfers no specifications between products. Missing data moves a listing to the documentation-completeness formula and is labelled; it is never filled in.

Corrections

Every documented specification records where it came from, which makes corrections fast to verify. Send the page URL, the exact listing identifier and the contradicting listing or manual text to [email protected]. Our scope and limits are set out on the about page.