# Sleds & Snowshoes: research notes

Date: 2026-10-25. Scope: recreational winter sliding sleds (toboggans, steerable runner sleds, saucers/discs, inflatable snow tubes) and recreational/backcountry snowshoes (trail, backcountry/mountaineering, running, and kids models). One register, two mechanically unrelated product families bundled under one taxonomy leaf (`supplies > active > sleds`, label "Sleds & Snowshoes"); `product_type` is the first split and does the dividing rather than two separate registers. Excludes: competition sleds (bobsled, luge, skeleton — a different, regulated sport, not sold as consumer winter-play gear), dog-sled and transport/pulk sleds (cargo/working gear, not recreational), and ski/snowboard bindings and poles, which belong to the sibling leaves `skis` (Skis & Snowboards) and `poles` (Hiking & Camping Poles). Snowshoe trekking poles are mentioned by sources only as an accessory and are left to the `poles` sibling, not swallowed here. `avalanche-safety` and `snow-boots-bindings` are separate sibling leaves; this register does not cover avalanche gear or boot/binding systems sold independently of a snowshoe.

Quote labels: **verbatim** = I read the actual page or abstract text myself, either via `curl` plus a Python regex strip of the HTML tags (printed to the terminal and read), or via NCBI's `efetch` plain-text API; **fetched excerpt** = the WebFetch tool's model summary of a page, used only where curl was not practical; **search excerpt** = a search-engine summary only. No search-excerpt-only number is in category.json; WebSearch itself was unavailable for this entire run (see below), so there are no search excerpts at all in this register — every claim is either verbatim-read or a WebFetch summary I later upgraded to verbatim by curling the same page myself.

## Seed

Ran `seed.py --category sleds`. Output, recorded verbatim:

> Example: bedsheets `weave` — "the expert split is often not the number everyone advertises. Thread count is the generic trap here. Weave is what changes how the sheet feels and sleeps."
> Persona: "An audio engineer, who thinks in measurable versus perceived quality and where marketing lives in the gap"

1. **The sleds-and-snowshoes equivalent of weave is frame-and-decking material on a snowshoe.** Snowshoe length (inches) is the number every listing leads with, often right in the model name ("Big Sky 32"), and it does matter — but two models at the same length can be rated for very different weights, and the length number says nothing about how the shoe feels or grips. What actually divides the snowshoe world the way weave divides sheets is the frame/decking choice: aluminum-and-nylon (traditional, grippy, heavier, noisier) versus plastic/composite (cheap, easy to add traction to, loud on hardpack) versus EVA foam (quiet, cushioned, little traction). Crescent Moon's own site literally splits its catalog into "Aluminum Snowshoes" and "Foam Snowshoes" as top-level categories, and Switchback Travel's buying guide runs the same trade-off in reviewer language. **Kept** as the `frame_decking_material` facet, explicitly written up as "the real split, not the length number everyone advertises."
2. **The audio-engineer persona's "measurable versus perceived, where marketing lives in the gap" lands on the snowshoe weight rating.** The measurable number is the stated weight capacity (or "recommended load"); the perceived gap is that manufacturers define it inconsistently — one Crescent Moon listing states "up to 260lbs" in its marketing copy and "160 to 265 lbs" in its spec line for the *same model*. The engineer would also notice that NEISS-based injury papers measure sledding risk by collisions and head injuries, not by any sled "performance" number at all — there is no measurable spec for sled safety, only epidemiology. **Kept** as judgement 3 (the weight-rating inconsistency) and judgements 1–2 (the safety-standard gap, sourced to two peer-reviewed NEISS studies rather than a spec).

Both survived the data check and are in category.json. Nothing from the seed step was dropped.

## Sources and what each supports

### Could not reach at all

**WebSearch** reported its session budget was fully used (200 of 200 calls) before I ran a single query for this category; every search attempt in this run returned the budget-exhausted message. The entire research pass therefore came from direct `curl`/WebFetch fetches of URLs I already knew or could guess, links found inside those pages, and NCBI's E-utilities API, never from a search engine. This is a real limitation: for a thin/niche category like this one, the brand and reviewer landscape is harder to discover without search, and I likely missed some sources (MSR/Tubbs's own current buying guide, a second specialist review, any ASTM/ISO document search) that a search engine would have surfaced quickly.

**Blocked or unreachable despite direct URL attempts** (recorded so a later run does not repeat the same dead ends): `cpsc.gov` (Akamai 403 on every path tried, including `/Safety-Education/Safety-Guides/Sports-Fitness-and-Recreation/Sledding-Safety`), `nsc.org` (403, "Access Denied... contact support"), `rei.com` (connection refused/timeout on `/learn/expert-advice/...` and `/blog/...` paths), `llbean.com` (WAF denial page), `msrgear.com` and `cascadedesigns.com` snowshoe product/category pages (site restructured since the guide URLs I knew; every guess 301'd to a 404), `healthychildren.org` and `safekids.org` and `chop.edu` sledding-safety pages (404 on every guessed slug), `nationwidechildrens.org` newsroom pages (302 redirect, page retired), `flexibleflyer.com` (pure client-side JS redirect, nothing in the raw HTML), `slipperyracer.com` and `airhead.com/collections/snow-tubes` (client-rendered, empty on curl), `crescentmoongear.com`'s own "Causes" and "Foam vs Aluminum" blog *link text* worked but the brand's actual how-to-buy guide page 404'd (I used its category and product pages instead, which were live). `web.archive.org` returned a 429 "suspected abusive bot traffic" block on the one snapshot I tried (an old `msrgear.com/snowshoes` capture) and was not retried.

**What this means for the register:** no CPSC or standards-body source appears in references, because none was reachable. The safety content (judgements 1 and 2) rests entirely on two peer-reviewed NEISS-based papers, read verbatim via NCBI's own plain-text API, which is a stronger source than a regulator's consumer-safety blog post would have been anyway, but I note the gap explicitly rather than implying a search for a standard came up empty.

### Mechanism and vocabulary (primer)

**Wikipedia, "Sled"** (curl, raw HTML stripped and read in full — verbatim)
https://en.wikipedia.org/wiki/Sled
- verbatim: "A sled, sledge, or sleigh is a vehicle that slides across a surface, usually of ice or snow. It is built with a smooth underside or a separate body supported by runners that reduce friction with the surface."
- verbatim, recreational sled list: "Airboard, a snow bodyboard, i.e. an inflatable single-person sled"; "Bum slider, a bum-shaped skid with a simple handle"; "Flexible Flyer, a steerable wooden sled with thin metal runners"; "Foam slider, a flat piece of durable foam with handles and a smooth underside"; "Inflatable tube sled, a plastic membrane filled with air to make a very lightweight round sled"; "Kicksled or spark, a human-powered sled"; "Saucer sled, a round saucer-shaped sled without runners, usually made out of plastic or metal"; "Toboggan, an elongated sled without runners, usually made from wood or plastic, but sometimes made from sheet metal."
- Supports: the sled_type values and vocabulary; which types have runners/steering and which do not; the gliding-friction mechanism (used qualified, since the article itself says the simple "meltwater film" explanation is "insufficient" — I kept the mechanism line soft, "part of why", rather than stating it as the whole explanation). Confidence high.

**Wikipedia, "Toboggan"** (curl, raw HTML stripped and read in full — verbatim)
https://en.wikipedia.org/wiki/Toboggan
- verbatim: "A toboggan differs from most sleds or sleighs in that it has no runners or skis (or only low ones) on the underside. The bottom of a toboggan rides directly on the snow."
- verbatim: "The traditional North American toboggan is made of bound, lengthwise parallel wood slats, all bent up and backwards at the front to form a recumbent 'J' shape. A thin rope is run across the edge of the curved front to provide rudimentary steering."
- verbatim: "Modern recreational toboggans are typically manufactured from wood, plastic or aluminum."
- Supports: toboggan construction, the "rudimentary" steering rope (why `steerable` for toboggans has a caveat in its backfill_note), sled_material values. Confidence high.

**Wikipedia, "Snowshoe"** (curl, raw HTML stripped and read in full — verbatim)
https://en.wikipedia.org/wiki/Snowshoe
- verbatim: "Their large footprint spreads the user's weight out and allows them to travel largely on top of rather than through snow. Adjustable bindings attach them to appropriate winter footwear."
- verbatim: "Snowshoes today are divided into three types: aerobic/running (small and light; not intended for backcountry use); recreational (a bit larger; meant for use in gentle-to moderate walks of 3–5 miles (4.8–8.0 km)); and mountaineering (the largest, meant for serious hill-climbing, long-distance trips and off-trail use)."
- verbatim: "Mountaineering shoes can be at least 30 inches (76 cm) long by 10 inches (25 cm) wide; a lighter pair of racing shoes can be slightly narrower and 25 inches (64 cm) or shorter."
- verbatim, bindings: "On modern shoes, there are two styles of binding: fixed-rotation (also known as 'limited-rotation') bindings, and full-rotation (also known as 'pivot') bindings... In fixed-rotation bindings, the binding is attached to the snowshoe with an elastic strap that brings the tail of the snowshoe up with each step. The snowshoe therefore moves with the foot and the tail does not drag... Fixed-rotation bindings are preferred for racing. Full-rotation bindings allow the user's toes to pivot below the deck of the snowshoe. They allow the crampon cleats that are under the foot to be kicked into a slope for grip in climbing, but are relatively awkward for stepping sideways and backwards as the tail of the snowshoe can drag. Fixed-rotation bindings often cause snow to be kicked up the back of the wearer's legs; this does not tend to happen with full-rotation bindings."
- verbatim: "Snowshoes function best when there is enough snow beneath them to pack a layer between them and the ground, usually at a depth of 8 inches (20 cm) or more."
- Supports: the three-category selection scheme (Switchback's four-category version, below, is the one actually used in the register, with "kids" as the fourth; this source is the second confirmation of the trail/backcountry/running split itself), length bands, and — the only source for this — the fixed-rotation versus full-rotation binding mechanics. Confidence high for everything except the binding-rotation claim, which is single-source (see judgement 4 and "Corrections" below).

### Brand spec pages (what two makers publish)

**Crescent Moon Gear, "Foam vs Aluminum"** (curl, raw HTML stripped and read in full — verbatim)
https://crescentmoongear.com/foam-vs-aluminum/
- verbatim: "Like a pair of your favorite running or walking shoes, foam bends and flexes in perfect synchronicity with your foot... They're quiet, cushioned, lightweight, and now, with the addition of ICESPIKE traction points, they're versatile and have great grip."
- verbatim: "Traditionalists will always find the tried and true aluminum framed shoes are the workhorses... They've got aggressive crampons that dig deep into steep and challenging terrain and with more surface area to provide additional float into the vast and un-tracked powder fields... Crescent Moon still bends its frames here in Boulder."
- Supports: the aluminum-vs-foam trade-off, in the maker's own words, independent of Switchback's reviewer framing. Confidence high (it is marketing copy, but the physical claims — crampon aggressiveness, flex, weight — match Switchback's independent description below, so I treat the trade-off itself as two-sourced).

**Crescent Moon Gear, "Aluminum Snowshoes" category page** (curl, raw HTML stripped and read in full — verbatim)
https://crescentmoongear.com/aluminum-snowshoes/
- verbatim: "Our aluminum snowshoe frame is made from lightweight aircraft-grade aluminum tubing for strength, durability and reliability. The decking is made from an ultra-durable material called TGS which won't tear or fail. These snowshoes have the only three-claw crampon traction system on the market, and features the only genuine toe claw found on any snowshoe. Each set of aluminum snowshoes come with a free pair of heel lifts."
- verbatim model names (each embeds a length in inches): "Big Sky 32", "Yellowstone 24.5", "Denali 37", "Leadville 29", "Sawtooth 27", "Vail 24.5", "Trailhawk 27".
- Supports: aluminum/nylon-equivalent construction detail (single-source for "TGS"/"three-claw crampon" specifics, so these exact material-trade-name claims are not in the JSON, only the general construction); the fact that brands embed snowshoe length directly in the model name. Confidence high for the model-name pattern.

**Crescent Moon Gear, "Big Sky 32" product page** (curl, raw HTML stripped and read in full — verbatim)
https://crescentmoongear.com/big-sky-32-mens-backcountry-snowshoes/
- verbatim, short description: "Conquer the backcountry with the rugged men's Big Sky 32 snowshoe... For men up to 260lbs."
- verbatim, features: "User Weight Range: 160 to 265 lbs"; "Fits M9 - M15"; "Flat to Mountainous Terrain"; "Durability: Built to last with Hypalon decking coupled with 5/8 inch diameter 6000 series aluminum frame."; "Single-Pull Binding System"; "Stainless Steel Crampon System: Each snowshoe has seventeen stainless steel crampon teeth for great traction when ascending, descending, or side-hilling."; "Traversing Claws: Two traversing claws per snowshoe add bite when side-hilling."; "Tear-drop shaped frame: Mildly tapered design keeps width front-of-foot for a natural stride while maximizing flotation."; "Integrated Heel Lifts."
- Supports: judgement 3 (the 260 lb vs 160–265 lb inconsistency, on the *same page*); the heel-lift-bar feature and crampon-count detail (single-source for the exact "seventeen teeth" and "Hypalon"/"5/8 inch 6000 series aluminum" figures, so those specific numbers are not committed to the JSON — see "Trimmed detail"). Confidence high for what the page says, medium for generalising any single number to the category.

### Specialist review

**Switchback Travel, "Best Snowshoes of 2026"** (curl, raw HTML stripped and read in full — verbatim; by Maddie Downie and John Ellings, last updated September 17, 2026)
https://www.switchbacktravel.com/best-snowshoes
- verbatim, categories: "Trail / Trail/recreational snowshoes are great for moderate hiking trails... built for flat terrain or rolling hills, but not for steep mountain ascents or the less predictable backcountry... shorter in length... traction designs are simpler and less aggressive." "Backcountry... more performance-oriented... longer, providing more float in deep snow and increased traction... most expensive of the bunch." "Running... a small but notable subcategory... place a premium on lightweight materials... most are tuned for hardpack." "Kids... smaller versions of their adult counterparts, with similar designs overall but shorter lengths and lower load limits... Atlas Spark (50 to 120 lb.) and Mini (30 to 80 lb.)."
- verbatim, sizing: "Most snowshoes range from 22 inches at the short end (MSR's Lightning Ascent and Evo Trail) up to 36 inches at the long end (the Tubbs Mountaineer)." "Atlas' Range-Trail is rated to 200 pounds for the 26-inch length, 250 pounds for the 30-inch variation, and 300+ pounds for the largest 35-inch model."
- verbatim, frame/decking: "There are two general types of snowshoes: an aluminum outer frame with flexible nylon decking, and a frameless, solid deck made of plastic or composite... The disadvantage is traction... plastic... it's louder and slightly more prone to damage... EVA... lack of a crampon underfoot does impact traction, but it has a nice cushioned feel that is quieter."
- verbatim, heel lifts: "Most high-end trail—and nearly every backcountry—snowshoes come with a heel lift... we don't find them necessary for most moderate terrain (they're useful for extended climbs)."
- verbatim, traction: "Low-cost models commonly have a small steel crampon under your toes, additional cleats around the middle of your foot... ultralight models like the Crescent Moon EVA and TSL Symbioz Hyperflex Racing have lightweight steel spikes rather than crampons."
- Supports: the `snowshoe_category` values and why_experts_care; the weight-to-length sizing table (Atlas); the frame/decking trade-off (second source, independent of Crescent Moon's own copy); the heel-lift-bar terrain dependence (used for the "trail"/"running" branch `makes_irrelevant: [heel_lift_bars]` legs); the traction ladder. Confidence high throughout — this is the single richest source for the whole snowshoe half of the register.

### Academic (injury epidemiology)

**Howell, Nelson & McKenzie, "Pediatric and adolescent sledding-related injuries treated in US emergency departments in 1997-2007," Pediatrics 2010;126(3):517-24** (NCBI `efetch` plain-text abstract — verbatim)
https://doi.org/10.1542/peds.2009-1499
- verbatim: "an estimated 229 023 patients ≤19 years of age were treated for sledding-related injuries in US emergency departments, with an average of 20,820 cases per year... The most frequent injury diagnoses were fractures (26.3%), followed by contusions and abrasions (25.0%). The head was the most commonly injured body part (34.1%), and injuries to the head were twice as likely to occur during collisions as through other mechanisms. Traumatic brain injuries were more likely to occur with snow tubes than with other sled types. A total of 4.1% of all cases required hospitalization."
- verbatim conclusion: "The use of sledding products that may reduce visibility (such as snow tubes) should be discouraged."
- Supports: `sled_type`/`steerable` why_experts_care; judgements 1 and 2. Confidence high (peer-reviewed, NEISS national sample).

**Evans, McAdams, Roberts & McKenzie, "Sledding-Related Injuries Among Children and Adults Treated in US Emergency Departments From 2008 to 2017," Clinical Journal of Sport Medicine 2021;31(6):e460-e466** (NCBI `efetch` plain-text abstract — verbatim)
https://doi.org/10.1097/JSM.0000000000000864
- verbatim: "An estimated 220 488 patients... were treated in EDs for sledding-related injuries... Most patients were injured by a collision (63.2%). Snow tube and disk riders were more likely to sustain a concussion or CHI compared to sled and toboggan riders [relative risk (RR): 1.53 (95% CI: 1.18-1.99)]."
- verbatim: "Sledding should be performed in obstacle-free areas to reduce the risk of collision and while wearing helmets to mitigate head injuries."
- Supports: second, independent confirmation (different decade, same NEISS methodology) of the tube/disk-versus-sled/toboggan concussion-risk gap; the 63% collision figure. Confidence high. The two studies together are the "two sources" behind judgement 1; I did not find a third, non-NEISS-based source, which is why judgement 2 (no governing design standard) stays at medium confidence rather than high.

## Data points used for derived metrics

- `load_margin_lb`: Switchback Travel's description of "recommended load" as including "your body weight plus anything you'll be carrying" is the only source for framing this as a subtraction; it is arithmetic (max_user_weight minus your own total), not a category-specific formula with a derived constant, so no second number-source was needed beyond confirming the concept.
- Snowshoe length-vs-weight pairing (the `two_axis_map` visual grammar and `max_user_weight`/`snowshoe_length_in` ranges): two brands' own numbers — Crescent Moon's Big Sky 32 (32 in, 160–265 lb) and Switchback's quoted Atlas Range-Trail table (26 in/200 lb, 30 in/250 lb, 35 in/300+ lb) — agree on the direction (longer shoe, higher rated weight) even though neither states a formula; the register's `range.typical` text ("trail shoes run 21 to 27 in; backcountry models run 25 to 36 in") comes from Switchback's "22 inches... up to 36 inches" plus its per-model category assignments, cross-checked against Crescent Moon's own model lengths (24.5 to 37 in across its lineup).

## Against the existing add-on schema

Read `zamana-expertise/data/schema-parts/supplies/active/add-ons/sled.jsonc` (the entire file is one field: `sizing.snowshoe_length`, value + unit `[in, cm]`), `active/active-standard.jsonc` (one field: `sizing.max_user_weight`, value + unit `[kg, lb]`), `supplies/supplies-standard.jsonc` (`high_signal_markdown` plus generic `sizing`), and `AGENTS.md` for the field-kind vocabulary (vertical anchor, category quality currency, hard filter, sizing; visually-derived fields stay out; fields derivable from kept fields are cut). This is the thinnest add-on I have seen in this series — one field — which matches the brief's "if the category is thin, write a short register" instruction on the research side too. Proposals only; nothing in `zamana-expertise` was edited.

| Facet in this register | In add-on today | Proposed kind under AGENTS.md | Note |
|---|---|---|---|
| `snowshoe_length_in` | yes, `sizing.snowshoe_length` | sizing (already correct) | Keep as is. Brands embed it in the model name, so it is cheap to backfill from the title alone when the spec field is missing. |
| `max_user_weight` | yes, `active-standard.jsonc` `sizing.max_user_weight` | sizing (already correct, shared) | Keep as is; this register just documents how it is read differently per branch (a hard ceiling for sleds, a judgement-needs-length number for snowshoes). No change needed. |
| `sled_type` (toboggan / runner_sled / saucer_disc / inflatable_tube) | no | vertical anchor / hard filter | Proposed new enum on the sled add-on. This is the one fact every sled listing states and that NEISS categorises injuries by (product codes 1217/1273/1274/1299), so it is as close to a vertical anchor as this leaf has. |
| `snowshoe_category` (trail / backcountry / running / kids) | no | hard filter | Proposed new enum, parallel to `sled_type`, for the snowshoe half. Backfill cost low: inferable from length, price tier and the brand's own collection, so worth adding even though it is rarely a literal retail-listing field. |
| `sled_material` (wood / plastic / aluminum / inflatable_vinyl) | no | hard filter | Proposed. Usually in the title or a materials bullet. |
| `frame_decking_material` (aluminum_nylon / plastic_composite / eva_foam) | no | hard filter / category quality currency | Proposed. Brands already split their own catalogs this way (Crescent Moon's top-level nav is literally "Aluminum Snowshoes" vs "Foam Snowshoes"), which is the strongest signal in this whole register that a field belongs in the schema, not just the markdown. |
| `steerable` | no | **not a new field** | Derivable from `sled_type` (true for runner_sled, true-with-caveat for toboggan, false for saucer_disc/inflatable_tube); AGENTS.md rule 5 says fields derivable from kept fields should be cut, so this stays a register-only annotation, not a schema proposal. |
| `crampon_traction` (light_spikes / toe_crampon / toe_and_heel_rails) | no | **markdown, not a filter** | Partly visible in the product photo (AGENTS.md rule 1: visually-derived fields stay out of schemas), and judging "aggressiveness" needs a reviewer's comparison, not a typed value most listings carry. Worth a sentence in `high_signal_markdown` for the top N models, not a structured field. |
| `binding_pivot_type` (full_rotation / fixed_rotation) | no | **markdown, not a filter** | High backfill cost (researched, single-sourced in this pass — see judgement 4) and essentially never stated on a retail listing. A `high_signal_markdown` line for the top N backcountry models, same treatment as `crampon_traction`. |
| `heel_lift_bars` | no | markdown bullet, borderline filter | Low priority (priority.narrows_set 1): most backcountry shoes have it, most trail/running shoes do not need it, and it is a minor feature rather than a hard filter buyers search by. A `high_signal_markdown` bullet is enough; not worth a dedicated boolean column. |

Nothing in the current one-field add-on counts as noise — there simply is almost nothing there yet to flag.

## Branches

Every leg below is traced to a sentence in category.json or in these notes, as the brief requires; no leg needed a two-fact chain.

- `product_type = snowshoe` → ask `snowshoe_category`, `snowshoe_length_in`, `frame_decking_material`; makes irrelevant `sled_type`, `sled_material`, `steerable`. Traced to mechanism.plain: a snowshoe "spreads body weight over a wide deck" while a sled "rides a slick, low-friction underside on gravity alone" — the sled facets describe a sliding device and do not apply to a flotation device. SUPPORTED.
- `product_type = sled` → ask `sled_type`; makes irrelevant `snowshoe_category`, `snowshoe_length_in`, `frame_decking_material`, `crampon_traction`, `binding_pivot_type`, `heel_lift_bars`. Same mechanism split, reverse direction. SUPPORTED.
- `sled_type = saucer_disc` or `inflatable_tube` → ask `max_user_weight`; makes irrelevant `sled_material`. Traced to Wikipedia's own definitions: a saucer sled is "usually made out of plastic or metal" and an inflatable tube is "a plastic membrane filled with air" — the type name already fixes the material on this branch, so asking `sled_material` separately would be the generic-trap kind of redundant question. SUPPORTED.
- `sled_type = toboggan` or `runner_sled` → ask `sled_material`, `max_user_weight`. Traced to Wikipedia's "Toboggan" page: "typically manufactured from wood, plastic or aluminum" — material genuinely varies within this branch, unlike the saucer/tube branch. SUPPORTED.
- `snowshoe_category = backcountry` → ask `frame_decking_material`, `crampon_traction`, `binding_pivot_type`. Traced to Switchback: backcountry models are "more performance-oriented... increased traction for handling a variety of conditions, including steep terrain and ice," and its named backcountry picks (MSR Lightning Ascent, Tubbs Mountaineer) are both aluminum-framed — all three follow-up questions are live choices on this branch. SUPPORTED.
- `snowshoe_category = trail` → ask `frame_decking_material`; makes irrelevant `heel_lift_bars`. Traced to Switchback verbatim: for trail terrain, "we don't find them necessary for most moderate terrain (they're useful for extended climbs)." SUPPORTED.
- `snowshoe_category = running` → makes irrelevant `heel_lift_bars`. Traced the same way: running snowshoes are built for "hardpack or small amounts of fresh snow," not sustained climbing, so the heel-lift feature (explicitly a climbing aid, per Switchback and Crescent Moon's "Integrated Heel Lifts" on backcountry/trail models only) does not apply. SUPPORTED, same evidence as the trail leg.
- `snowshoe_category = kids` → ask `max_user_weight`; makes irrelevant `frame_decking_material`, `crampon_traction`, `binding_pivot_type`, `heel_lift_bars`. Traced to Switchback verbatim: kids' models are "smaller versions of their adult counterparts, with similar designs overall but shorter lengths and lower load limits" — the construction questions collapse because the shoe is just a scaled-down adult design, and the one thing a parent actually needs to check is whether it fits the child's weight. SUPPORTED.

No branch leg needed a two-fact chain and none was dropped as weak.

## Trimmed detail

- `sled_type.why_experts_care`: Wikipedia's full list of minor sled types (airboard, bum slider, foam slider, kicksled) is folded into the `values` enum only as the four that actually show up as retail categories (toboggan, runner_sled, saucer_disc, inflatable_tube); the others are real but not common enough as distinct *purchasable* categories to carry their own value.
- `frame_decking_material.why_experts_care`: Crescent Moon's own brand-name materials ("TGS" decking, "Hypalon" decking, "ICESPIKE" traction points) are single-sourced to one maker and are not generalised into the register as category-wide facts; they stay here as examples of what a spec sheet calls the same underlying materials.
- `crescent_moon_bigsky32`: the exact figures "seventeen stainless steel crampon teeth," "5/8 inch diameter 6000 series aluminum frame," and "Fits M9 - M15" are single-source (one model, one brand) and are not committed to the JSON as category ranges; they are kept here as illustrative detail for a future backfill pass on this exact model.
- `switchback_best_snowshoes`: per-model weights (Evo Trail 3 lb 10 oz, Lightning Ascent 4 lb 5 oz, Tubbs Mountaineer 5 lb 4.8 oz) are single-source and not added as a `snowshoe_weight` facet for that reason — dropped rather than included with weak backing (see "Dropped").
- `wiki_snowshoe`: the full history (origins 4,000–6,000 years ago, Cree/Huron/Ojibwe/Iroquois regional shapes, the 1950s Tubbs Green Mountain Bearpaw redesign, "mal de raquette" snowshoe-sickness, the 1994 step-in binding, the United States Snowshoe Association) is genuinely interesting but not expert-buyer-relevant; kept here rather than in `summary`/`mechanism` to respect the word caps.
- `max_user_weight.why_experts_care`: the Crescent Moon 260 lb vs 265 lb discrepancy itself lives in judgement 3, not repeated in the facet's own capped text.

## Dropped

- A `snowshoe_weight_lb` facet (weight of the shoe itself, not its load rating). Real and reviewers mention it, but every figure I found (Switchback's per-model weights) was single-sourced to one review, with no second source to confirm a category-level range. Kept only as illustrative text above.
- A `rider_count` or `tandem` facet for sleds (how many people a toboggan seats). Plausible — Wikipedia's toboggan illustration shows multiple riders — but no source gave a sourced number of riders per length, so this stayed a plain-language note inside `sled_type.why_experts_care`'s absence rather than an invented facet.
- A formal "governing safety standard" facet (the house rule's instruction for products with a safety risk). I looked for one and could not find it: CPSC's site blocked every fetch, and nothing in the reachable sources named a mandatory US design standard for recreational sleds (unlike, say, UL 153 for lamps or UN 38.3 for batteries in earlier runs). Rather than invent a facet with no real values, I wrote this up as judgements 1 and 2 instead, which the schema explicitly allows ("the governing standard is a facet or a judgement").
- `binding_closure` (strap/webbing vs single-pull vs Boa dial) — Switchback describes this in detail (Atlas/Tubbs webbing, MSR's Paragon/Paraglide mesh systems, Crescent Moon's single-pull, Boa dial systems from several brands), and it is genuinely a buyer-facing comfort choice. Dropped from the register for thinness: it is a comfort/ease-of-use spectrum with no safety or performance consequence I could source (unlike `binding_pivot_type`, which trades climbing grip against tail drag), and the category is already not generating a fourth split; it is mentioned here so a future pass does not have to re-discover it.

## Corrections

None — this is the first pass for this category, built directly against sources, not revising an interrupted run. The `sleds/` folder held no prior files when this run started.

## Corrections after the blind check (2026-10-25)

Re-fetched `switchback_best_snowshoes` directly (curl, HTML stripped, read verbatim) to verify every number touched below before editing; quotes matched the checker's report exactly, including the Atlas Range-Trail weight/length table, the Trail/Backcountry per-model length lists, and the TSL Symbioz Hyperflex Racing figure ("Category: Running Length: 21 in"). No other source needed fetching; the remaining fixes were wording and classification corrections against material already in this file.

1. **Fixed** (most severe, CONTRADICTED) — `max_user_weight.why_experts_care`'s "same length rated for 200 lb and for 300 lb" example claimed a same-length comparison that Switchback's Atlas Range-Trail table does not show (weight rises with *increasing* length: 200 lb/26 in, 250 lb/30 in, 300+ lb/35 in). Dropped the invented same-length pairing and rewrote the sentence to the general, now-supported point: a brand's own length-to-weight table shows capacity rising with length, not fixed per model.
2. **Fixed** — `snowshoe_length_in.definition` misattributed the 22 in short end to "racing shoes"; Switchback names the 22 in examples as a Trail model (Evo Trail) and a Backcountry model (Lightning Ascent), not a racing shoe. Re-keyed the 22-36 in span to trail/backcountry models, and re-sourced the racing-length claim to two sources that agree in direction: Switchback's own TSL Symbioz Hyperflex Racing at 21 in (re-fetched and confirmed verbatim) and `wiki_snowshoe`'s "25 inches (64 cm) or shorter." Added `wiki_snowshoe` to this facet's evidence array.
3. **Fixed** — Judgement 2's citations (`evans_2021`, `howell_2010`) are pure injury-epidemiology papers and do not address regulatory design standards. Dropped the evidence array rather than attach the archived NSC page as a second wrong-genre citation, and reworded the note to say plainly this is an unconfirmed gap, not a sourced claim.
4. **Fixed** — `snowshoe_length_in.range.typical`'s stated bands ("trail 21-27 in; backcountry 25-36 in") had direct counter-examples inside Switchback's own table (Helium-Trail reaches 30 in on the trail side; Lightning Ascent's 22 in option sits on the backcountry side). Re-fetched and confirmed both; widened to "trail mostly 22-30 in; backcountry mostly 22-36 in, overlapping at the short end."
5. **Fixed** — `sled_type`'s cold-start answer "steerable sled" reused the same steerability axis as the separate `steerable` facet, which also marks toboggans steerable (via rope), so the answer could not reliably separate `runner_sled` from `toboggan`. Replaced with a construction-based label, "sled with runners," in both `splits[1].question` and the facet's own `cold_start_question`. No branch `when`/`then_ask`/`makes_irrelevant` logic changed, so the Branches trace in this file is unaffected — only the question wording changed.
6. **Fixed** — `frame_decking_material` was marked `source: listing` on evidence that was actually a brand *category* page (Crescent Moon's "Aluminum Snowshoes" vs "Foam Snowshoes" nav), not a per-product retail listing — the exact pattern `check.md` calls the most common defect across this series. Reclassified to `researched`, raised `backfill_cost` from `low` to `medium`, reworded `listing_field_hint` to say plainly it is not a typical listing field, and added a `backfill_note` recommending a top-N pass, matching how `binding_pivot_type` and `crampon_traction` are already handled.
7. **Fixed** — Judgement 3's (index 2) note said the Big Sky 32 page gives "two different weight ceilings" (260 lb / 265 lb) and mislabeled which block says what. The page actually gives three figures: 260 lb in the marketing blurb, 160-265 lb in a Features bullet, and 150-260 lb in the Specifications tab. Rewrote the note with the correct count and correct block labels.
8. **Fixed** — `snowshoe_category`'s `source: researched` was inconsistent with its own `backfill_note`, which describes a derivation ("inferable from length plus price tier... without a human lookup"), exactly what the schema defines as `derived`. Reclassified to `derived` and `backfill_cost` from `low` to `none` to match. This facet keeps its `cold_start_question` for interactive use; the `source` field only governs backfill when the question wasn't asked directly, and derivation is cheaper than the old `low`/`researched` pairing implied.
9. **Fixed** — `frame_decking_material`'s cold-start question ("quiet and cushioned, or light and grippy?") was binary for a three-value facet, and "light and grippy" did not cleanly resolve to either `aluminum_nylon` (grippy but heavy) or `plastic_composite` (light but not inherently grippy per this register's own trade-off text). Rewrote as a three-option question — "Quiet and cushioned, light and affordable, or heavy-duty grip?" — mapping cleanly to `eva_foam`, `plastic_composite`, and `aluminum_nylon` respectively.
10. **Fixed** — `visual_grammars[0].origin` claimed Atlas and Crescent Moon "already publish this pairing" as a chart; Atlas's own sizing chart was never fetched directly (only paraphrased through Switchback's review copy) and Crescent Moon's pages give single-model specs, not a cross-model chart. Softened to "Implicit in Atlas's and Crescent Moon's own spec sheets, not drawn as a cross-model chart today," matching the phrasing already used for the Traction ladder's origin.

No findings rejected; all ten required fixes were applied as the checker specified or with the checker's own stated alternative. No branch leg's `when`/`then_ask`/`makes_irrelevant` values changed as a result of these fixes (only facet-level prose, sourcing, and classification), so the "Branches" section above still traces correctly and needed no edits beyond this log entry.

## Note on scripts

`validate.py`, `render.py` and `viewer.py` all ran clean on the first attempt that had correct JSON (after two small prose fixes made during my own STYLE.md pass: a doubled-period in two `visual_grammars[].origin` strings caused by a trailing period meeting the renderer's own added period, and an awkward "bite into ice and climbs" phrase in `mechanism.plain`). No script bug to report.