Every player wants to know the same thing, and almost nobody can find out: am I any good?

Club tennis gives you a partial answer. You know you beat Dave and lose to Sarah. What you do not know is whether your forehand is unusually good for your level, whether your serve is the thing holding you back, or how you would stack up against the player on the next court who you have never played.

Leaderboards on OnCourtAI exist to answer that, using technique rather than match results.

Four Tiers, Starting Very Local

You are ranked at four levels of geography, and the most interesting one is the smallest.

Nearby ranks you against every player within 50 miles of you, measured as actual great-circle distance from your location rather than by town names. This matters more than it sounds: a player in a village outside Harrogate and a player in Harrogate itself are neighbours in reality but would never share a "town" label. The nearby tier ignores naming entirely and just measures distance.

Town ranks you within your town or city. Region ranks you within your county or region. Country ranks you nationally.

In the app you will see these as plain statements rather than a wall of tables: #4 best player within 50 miles, #12 best player in Yorkshire, #58 best player in United Kingdom. Tap any line and it expands to shows how far you are away from the top player in that tier.

You Are Ranked on Your Best, Not Your Average

Ranking uses your peak score for each stroke — your best-ever analysis, not your average.

This is a deliberate choice and worth explaining, because the alternative is worse. If leaderboards ranked on your average, every video you ever filmed would be a risk. Film on a windy evening when you are tired, score badly, and your rank drops. Players would quickly learn to only film when conditions were perfect, which defeats the entire purpose of a tool meant to show you where you are weak.

Ranking on your peak means filming more can only ever help you. A bad session costs you nothing on the leaderboard. That removes the conflict between "use the tool honestly" and "protect my rank", and it is the single most important design decision in the whole feature.

The Overall Rank Is a Rank of Ranks

This is the part that is genuinely unusual, and it exists to stop a specific kind of gaming.

Your overall position is not the average of your raw scores. It is the average of your ranks across the strokes you have data for. If you are #3 in your area on the serve and #1 on the volley, your overall standing is built from those placings — 3 and 1 — rather than from the underlying numbers.

Why do it this way? Because averaging raw scores rewards specialists and punishes all-rounders. Technique scores are not calibrated to be equally generous across strokes: it is easier to score well on a volley than on a serve. A player who only ever films volleys would float to the top of a raw-score leaderboard while being, in any meaningful sense, a worse player than someone solidly good at four strokes.

Ranking the ranks fixes that. It asks "how well do you place against your peers on each stroke you play?" and rewards the player who places well across the board.

There is no minimum stroke count. You are ranked on what you actually have data for, so a player with one stroke analysed still appears — they are simply being judged on a narrower slice of their game.

Tennis and Padel Are Ranked Separately

Leaderboards are scoped by sport. Padel players are ranked against padel players on padel strokes; tennis players against tennis players. A strong tennis serve has no bearing on your padel standing, which is correct — they are different motions with different checkpoints.

In practice the tennis boards are far more populated. The tennis side of the platform has processed 2,602 analyses from 891 players, so a national ranking there is a real statement. Padel is earlier and the boards are correspondingly sparse, which we would rather tell you than let you discover.

Per-Stroke Boards Are Where the Insight Is

The overall rank is the headline, but the per-stroke rankings are the useful part.

A fairly common pattern looks like this: a player sits around the middle of their region overall, but is in the top ten locally on the forehand and nowhere on the serve. That is not a vague sense that the serve "could be better". It is a specific, comparative statement that the serve is the thing dragging the whole game down, and it points straight at what to work on.

This is the same argument we make about component scores within a stroke. Aggregates hide things. The detail is where the instruction lives.

On Leaderboards and Motivation

Technique scores measure how closely your movement matches patterns associated with efficient, repeatable strokes. They do not measure competitiveness, tactics, consistency under pressure, or whether you win. Plenty of players with unorthodox technique beat better-looking opponents every weekend, and a scoring system that pretends otherwise would be lying to you.

What a leaderboard is genuinely good for is scale. "Your forehand scores 61" means very little in isolation. "Your forehand scores 61, which puts you 9th of the players within 50 miles of you" means something immediately. It converts an abstract number into a position you can actually interpret.

And because ranking uses your peak, the way to climb is simply to keep filming until you catch yourself on a good day — then work out what you did differently.

Where to Find It

Rankings appear on your Stats page in the OnCourtAI app, underneath your overall player rank in "My Stats". You need at least one completed analysis and a detected location. Film a second stroke and you get a second set of boards, and your overall position starts reflecting your range rather than a single shot.