Test the tactical change before match day does.

Tactical decisions are frequent, and the chosen response is observable within days — a pressing scheme, a shape change, a substitution plan. The unchosen alternatives still require a named proxy or comparative design. That makes football the right first proving ground: the case can be inspected in public on real tactical questions without confidential club data.

A coaching staff frames the current plan and the proposed change as a Strategy Case. The engine branches the opponent responses worth worrying about, shears the assumptions the plan rests on, and compares candidate actions across those Scenario Branches. The intended output is the action that holds up best within the declared case, with an explicit trigger and downside. The coach decides.

The public proof begins with a worked tactical case whose assumptions are authored and labelled illustrative. Its current comparative-evidence tier is D — unresolved until a valid evaluation exists. It shows the intended loop: current plan, opponent adaptation, failure mode, recommended adjustment within the authored case, trigger, and downside. No club data is involved.

Great Bear Football is a flagship prototype. The public walkthrough now exists on the homepage at /#football-case. The interactive workspace — where an analyst edits the case, reruns the branches, and saves versions — and the outcome-resolution tooling are in development.

Soccer-q is our bettor-facing match-futures experiment. It serves a different audience and it does not share private accounts, Strategy Cases, or Decision Graph data with the coaching tool.

See it in Labs→