When a match is scored in Padlo, the app reads the whole point history and produces Findings: per-match coaching insights, each with a headline, a plain-language explanation, a recommendation for the next session, and links to the exact balls it came from.
How a Finding earns its confidence
A Finding is never a hunch. Each one is backed by the tagged balls of the match, and its weight comes from how many rallies actually evidence the pattern. A Finding can carry up to around thirty supporting balls, with up to six of them marked as critical, the rallies where the pattern decided the point. All of them are linked, so “second-serve returns broke down in the third set” is not a claim you have to take on faith. It is a claim with twenty-four rallies attached, six of which you should watch first.
A pattern with three supporting balls and one with twenty-eight read very differently, and the count is right there. That is the whole trust model: the evidence is the same point data you tagged courtside, and you can open it.
What you do with it at the next practice
This is the part that matters. Say the Finding above arrives after a Sunday match. On Monday you open the six critical balls, watch what actually happened, and decide whether you agree. If you do, the next session starts with fifteen minutes of returns against high-kicking serves, before anything else. The Finding already told you which player, which set, and which serves, so the drill designs itself.
Take one recommendation into a session, not six. A player told six things fixes none of them. Over a training block the Findings accumulate, and a headline that repeats across five matches is a genuine pattern rather than a bad Tuesday.
Sometimes you open the critical balls and disagree, and four of the six turn out to be tagged sloppily. That is useful too: now you know to fix the tagging, and the next match’s Findings get sharper. An insight you cannot audit is not worth reading. These you can.
Filter by player
A padel match has four players and you usually care about one or two. Findings filter per player, so a coach preparing a session with a single student pulls only what concerns them, and a player reading their own Findings gets their own game, not a quarter share of a match report.
What Findings are not
They see what was tagged: the shape of the point, not footwork or body language, and nothing about matches nobody scored. They answer one narrow question well: given everything that happened in this match, what is worth changing at the next practice? That question is worth automating. The rest of coaching is not.