Hand-maintained — see footer
Chaos Hockey cover art

Vision Training

The drill library for Chaos Hockey's training mode — the half of the project aimed at teaching hockey vision: reading space, recognising passing lanes, judging shot quality, and making better decisions under pressure. Each drill asks a different hockey question rather than being the same exercise at a higher speed.

Scores are deliberately large and open-ended. Nobody wants to finish a drill on 8/10 when they could finish on 18,450 — and an unbounded score means there is always room to improve rather than a ceiling to hit.

Foundations

Shared technology that has to exist before most of the library can be built. These are not drills — they are the things the drills stand on.

Loading data…

Current Drills

Everything installed in the game right now, generated directly from the game's own DrillCatalog — so this list cannot disagree with what the app actually offers. Confirmed in play means a human has played it and said it works; built, not yet played means it exists and passes its tests, which is a weaker claim and is labelled as one.

Loading data…

Future Plans

Designed but not built in any form — no code, no catalog entry, nothing in the game. These come from the original design document and are listed separately so what exists is never confused with what is imagined.

Loading data…

Difficulty Tiers

Difficulty is not simply "faster." Each tier tightens the timing windows and the positional tolerances, works the opponents harder, and pays more — and the multiplier is the only thing that touches the score, so a score is always comparable within its own tier. Leaderboards are kept per tier for the same reason.

These are the real values the game runs on, exported from DrillDifficultyCatalog — the same colours and letters shown in the app.

Skills

Every drill trains one primary skill, colour-coded identically in the app and on this page so a colour means one thing everywhere.

Data Model

1. Raw performance data

What actually happened: time, distance, speed, position, angle, puck velocity, possession, reaction time, accuracy, errors, attempts, successes, turnovers, interceptions, contact events.

2. Evaluated skill data

What those measurements mean: skating efficiency, puck control, passing and shooting quality, defensive positioning, decision quality, reaction, accuracy, deception, awareness, recovery. This is where the existing evaluators already live.

3. Presentation data

What the player actually sees: 42,850 POINTS, 8 STREAK, 96% CONTROL, FASTEST REACTION 0.41 SEC, NEW PERSONAL BEST. Keeping this separate means presentation can improve without touching evaluation.

Rewards, Badges & Honours

Loading data…

A Deliberate Design Rule

Not every drill has a correct answer. A player can make an excellent decision — open teammate, good lane, right weight, right moment — and still have the receiving AI fumble it. They keep the high score anyway. Equally, a technically perfect pass into an obviously terrible situation scores badly. The system is built to ask "did you make a good hockey decision?" rather than "did the computer let you win?"