THE MATFORGE METHOD

A statistic you can't defend is worse than no statistic. MatForge never renders a rate without a real denominator, never averages what wasn't measured, and always shows you the sample size. If your wrestler is 1-for-2 on single legs, we say “1 of 2” — not “50%,” because 50% off two shots is a coin flip wearing a suit. A move with no recorded misses has not been measured — which is not the same as never missing.

Every stat is only as honest as the hand that entered it. That's why MatForge explains events at the moment you record them, why volunteer-entered data is marked for coach review, and why your club has a Data Health page instead of a false sense of precision. We would rather slow you down for one sentence at entry than hand you a season of confident nonsense.

When we don't know an athlete's age, we treat them as the youngest. When we don't know their division, we derive it from what we do know — and say so — or apply the strictest protections. The AI never guesses about a child, never invents a birthday, and will refuse a task before it will fabricate a fact about a kid. This is not configurable. It never will be.

Live scoring is for scoring — fast, offline-proof, undo-friendly. Understanding happens afterward, in the film room, where you can pause, rewind, and be right. Attempts, techniques, and the story of a match get tagged against the footage, not guessed from the corner of your eye between whistles. That's how real programs break down film, and it's the only way the numbers stay true.

MatForge's AI reads what actually happened — the events you recorded, the patterns across a season, the memory of every match — and sharpens your judgment. It never replaces it. It computes nothing it can't ground, tells you when its inputs are thin, and shows you what it knew. You approve every plan before an athlete sees it.

The seventh time a kid goes flat on bottom, MatForge says “seventh time” — not “kids sometimes struggle on bottom.” Season-long memory per athlete is the difference between analysis and advice.

Score the match → repair the film → tag the truth → read the analysis → build the practice → drill it → measure the next match. Every screen in MatForge points to the next step of that loop, because data that doesn't change Tuesday's practice is trivia.

Every toggle is a decision we refused to make. MatForge ships with the defaults we believe in, and we add a setting only when two right answers genuinely exist. If the app feels like it has an opinion — it does.