◇  01 · THE FIT ENGINE

Nine signals.
Zero star ratings.

Every shoe in our database is scored against nine biomechanics signals derived from published lab measurements. Those signals are then weighted against your personal profile — not against how popular the shoe is or whether a brand paid for placement.

Take the quiz See our data sources

How the score is calculated

Each of the nine signals produces a sub-score on its own scale (e.g. 0–25 for category fit, 0–18 for toebox width). Those sub-scores are summed to produce a match score out of 100 for each shoe, specific to your profile. The same shoe will have a different match score for different users — a Hoka Clifton 9 might be a 91/100 for a 65kg runner with plantar fasciitis and a 72/100 for a 100kg trail runner.

The match score is separate from the core score shown in the catalogue (which is a general quality rating based on comfort, durability, performance, value, and fit — averaged across all user types). The match score is personal; the core score is universal.

Match score

Personal — changes per user profile. Drives the quiz shortlist.

Core score

Universal — quality rating across all user types. Shown in catalogue.

The nine signals

Weights are fixed across all shoe types. Detailed breakdowns below.

25%

Category fit

signal 01

Does this shoe match your stated use case?

18%

Toebox width

signal 02

Measured toebox mm vs your foot width preference.

15%

Cushioning

signal 03

Midsole hardness (HC scale) + stack height vs your body weight.

12%

Arch support

signal 04

Neutral / stability / motion-control classification.

10%

Budget

signal 05

Current lowest verified price vs your stated budget.

08%

Injury history

signal 06

Per-condition rules mapped to shoe construction.

05%

Durability

signal 07

Durability score weighted by your weekly mileage.

04%

Shoe weight

signal 08

Grams vs your activity intensity.

03%

Heel drop

signal 09

mm of heel-to-toe drop vs your foot strike pattern.

Total: 100% · scored against every shoe in the database

Signal deep dives

Why each signal was chosen and how it is scored.

25%weight
signal 01

Category fit

The single biggest predictor of satisfaction is buying the right type of shoe. A max-cushion daily trainer scores close to zero for a track 5K; a racing flat scores close to zero for recovery runs. We classify every shoe against 14 categories (road running, trail, sneaker, gym, basketball, hiking, walking, work/safety, and more) and score full match at 25%, partial overlap at 10–20%, and a near-zero score for a clear mismatch. This one signal alone eliminates most bad purchases.

Example

You say 'trail running'. A road shoe scores ~5/25. A trail shoe scores 22–25/25.

18%weightmm
signal 02

Toebox width

We record the internal toebox width in millimetres for every shoe (from Running Warehouse's physical measurements). We map your width preference — narrow, regular, wide, extra-wide — to a target mm range. The score peaks when the shoe's measurement falls inside your range and drops linearly as the gap grows. Width mismatch is the number-one cause of blisters, black toenails, and bunion flare-ups.

Example

Wide preference = ~98–104mm target. A shoe at 92mm scores ~8/18. A shoe at 100mm scores 18/18.

15%weightHC + mm
signal 03

Cushioning

Heavier runners need softer, higher-stacked midsoles to prevent joint overload. Lighter runners are better served by firmer, more responsive foam — excess softness creates instability. We combine the midsole hardness measurement (Shore C scale, sourced from Running Warehouse lab tests) with heel stack height and compare against your bodyweight to assign a cushioning match score. The relationship is non-linear: a 90kg runner who needs plush foam and buys a firm racing shoe is a far worse mismatch than the reverse.

Example

80kg runner + firm 38HC shoe → ~7/15. Same runner + 22HC max-cushion shoe → 14/15.

12%weight
signal 04

Arch support

We classify every shoe as neutral, stability, or motion-control based on the brand's design spec and independent lab assessments. We ask you whether your arches tend to overpronate, are neutral, or are high-arched. A neutral runner in a motion-control shoe loses about half their score on this signal; the reverse (pronator in a neutral shoe) loses most of it. This is a categorical signal — there's no partial credit for 'close enough' since arch mismatch compounds with distance.

Example

Neutral arches + neutral shoe → 12/12. Overpronator + neutral shoe → 3/12.

10%weight$
signal 05

Budget

We use the current lowest verified price across tracked retailers (not MSRP). Your budget gets a 20% grace window — a £140 shoe against a £120 budget still scores 8/10. Beyond 20% over budget, the score drops linearly. At 50%+ over budget, it floors at 2/10 (we still show it so you can see what you're missing). Going under budget is never penalised.

Example

$120 budget. Shoe costs $119 → 10/10. Shoe costs $145 → 7/10. Shoe costs $200 → 2/10.

8%weight
signal 06

Injury history

We maintain a rule set for eight common conditions: plantar fasciitis, achilles tendinopathy, knee pain, shin splints, bunions, Morton's neuroma, IT band syndrome, and metatarsalgia. Each condition maps to construction requirements — high heel drop for plantar fasciitis, wide toe box for bunions/Morton's, lower drop and more flexibility for achilles issues. Shoes that conflict with your conditions lose points; shoes that specifically address them gain points. If you report no injuries, every shoe scores the full 8/8 on this signal.

Example

Plantar fasciitis: high-drop shoe (12mm) → 8/8. Low-drop shoe (4mm) → 1/8.

5%weightkm/wk
signal 07

Durability

A shoe's overall durability score (sourced from aggregated tester reports and outsole compound data) is weighted by how many kilometres you run per week. A high-mileage runner (50km+/week) taking a racing flat with a thin outsole will wear it out in weeks — the durability penalty is severe. For a casual walker doing 20km/week, durability matters less and the penalty is mild. This is the only signal where low weekly mileage actually helps your match score on lower-durability shoes.

Example

50km/wk + low-durability shoe (score 60/100) → 1.5/5. Same shoe for 10km/wk → 3.5/5.

4%weightg
signal 08

Shoe weight

Weight matters most for speed — a 320g shoe will cost you time in a race. It matters least for hiking or casual wear, where stability and durability outweigh grams. We map your use case to a weight sensitivity level and score accordingly. Road racers and speed-session runners have high weight sensitivity; hikers and casual walkers have low sensitivity. The signal is capped at 4% because even a very heavy shoe in a speed context rarely makes or breaks a purchase decision on its own.

Example

Road racer + 340g shoe → 1/4. Hiker + 340g shoe → 3.5/4.

3%weightmm
signal 09

Heel drop

Heel drop affects where your foot lands and how load is distributed across the Achilles, calf, and knee. Heel strikers generally tolerate higher drops (8–12mm) better. Forefoot strikers and minimalist runners prefer low drop (0–6mm). Midfoot strikers sit in the middle. We score drop mismatch on a sliding scale — it's the lowest-weighted signal because the evidence for strict drop matching is weaker than for width or cushioning, and most runners can adapt across a moderate range.

Example

Heel striker + 10mm drop → 3/3. Forefoot striker + 10mm drop → 1/3.

What this engine does not do

01It does not account for personal biomechanical quirks that only a gait analysis can catch. If you have an unusual strike pattern or asymmetry, see a specialist.
02It does not predict long-term fit perfectly — feet swell, shoes break in, and preferences change over time.
03It does not incorporate user reviews or social proof. A shoe with 50,000 five-star reviews scores no higher than one with ten reviews if the data matches you better.
04Measurements are sourced from third parties and may have tolerances of ±2–3mm for width and ±1mm for drop. We flag when a measurement is estimated rather than directly measured.
Find my shoe Data sources