A score you can explain line by line.
The matching engine is deterministic: same profile, same data, same ranking. It never outputs a probability of recruitment or admission, and it tells you what it could not check.
1. Hard constraints
Applied before any score. A program is never ranked above one that meets the athlete's non-negotiables.
Competition category, excluded states, required competition levels, an active program, maximum budget, a required major and English minimums. Each check returns pass, fail, or unknown when the data needed is missing.
- Meets hard constraintsEvery critical constraint passed.
- Needs verificationNone failed, but at least one could not be checked. Shown separately so you can verify it.
- ExcludedA constraint failed. Listed with the reason, never scored.
2. Seven weighted components
Default weights below. Agencies can change them, and each athlete's importance levels adjust them further.
| Component | Default weight | Based on |
|---|---|---|
| Sporting compatibility | 25% | The athlete's competition level and preferred divisions against the division's tier (a ScoutPath heuristic, open to agency review). |
| Roster opportunity | 20% | Projected depth at the athlete's positions after expected departures, compared with a target depth, or a need the program confirmed. |
| Academic compatibility | 15% | Intended major offered, English requirements against test scores, GPA against admission rate, and completion rate. |
| Financial compatibility | 15% | Estimated annual cost against the preferred and maximum budget, and whether the association permits athletic aid. |
| Geographic compatibility | 10% | Preferred states or regions, campus setting and campus size. Excluded states are a hard constraint, not a score. |
| Development potential | 10% | Current depth at the position, last season's record, head coach tenure and, when wanted, a two-year transfer route. |
| Language and integration | 5% | Players from the athlete's country, international share of the roster and of undergraduates. Nationality is never treated as a language. |
Importance levels: Not important, Low, Medium, High, Critical. A component set to “Not important” is ignored for that athlete.
3. Missing data is never filled in
This is the part most tools get wrong.
- Left out, not guessed
- If a component cannot be computed (no roster loaded, no published cost), it is excluded and the remaining weights are renormalized.
- Data coverage
- Every result shows the share of the total weight that was actually computed from data.
- Ranking order
- Results are ordered by a coverage-adjusted score, so a program with little data cannot outrank a well-documented one by luck. The displayed score is the unadjusted one.
- Confidence
- High, medium or low, from data coverage, data freshness and how the roster facts were obtained (verified versus inferred).
4. Where AI is used
Narrowly, and only on facts the engine already computed.
When an Anthropic API key is configured, Claude rewrites the engine's evidence into a short paragraph a family can read. The model receives only the computed facts and must return structured output. ScoutPath checks that every number it mentions appears in the evidence and that it does not use words such as “guaranteed” or “will be offered”.
If the check fails or no key is configured, a template explanation built from the same evidence is shown instead. Scores and rankings never come from a language model.
See it with your own athletes.
Start the free trial with one of your athletes, or explore a complete fictional example first.
Free trial: one athlete profile, the best university match shown in full, the others locked until you choose a plan, no end date set, no card asked.