Analysis · Basketball scouting

He scored 37. What would he give your team?

Tyler Dorsey’s career high against Real Madrid tells you what happened on one night, while his season answers a different question.

By TeamKeys Analytics, Data deskPublished 2026-09-06Updated 2026-10-173 min read

Which basketball stats matter before signing a player?

The most useful scouting stats are those that describe the job you need a player to do. That question comes before the leaderboard.

Would the player you are watching give your team the scoring it needs? Tyler Dorsey’s career-high 37 points against Real Madrid in 2025/26 give you a reason to look.

But a club recruiting a wing might need someone who defends two positions and shoots off the catch. The points total alone cannot answer that brief.

What does a career-high game tell a scout?

A career high shows what a player produced in one game, while a season gives a broader view of his output. Neither answers every recruitment question.

Dorsey’s 37 against Real Madrid sits alongside 15.9 points per game and 59.5% true shooting across 43 games. The night and the season are different pieces of evidence.

If that night first caught your attention, the season gives you somewhere to take the investigation. You still need to examine his role and minutes before judging the fit.

EuroLeague BasketballDorsey drops 37 points in a statement performance: Olympiacos vs Real Madrid

How do you compare basketball players across leagues?

Cross-league comparisons need to account for the conditions behind the numbers. TeamKeys shows an adaptation estimate beside the raw figure, so the original remains visible.

Pace, physicality, refereeing and defensive quality can change what a player’s output means. His role and minutes also matter when considering what you would ask him to do.

For a club, the useful comparison keeps the recorded performance and the estimated adaptation together. The estimate supports a judgement about a different league; it does not settle it.

Can player similarity help you find the right signing?

Player similarity helps scouts narrow a search to candidates worth watching. It cannot establish that those players will suit the same team.

Two players can resemble each other across every measure a model uses and still fit a dressing room differently. A close statistical match leaves that question open.

Use the shortlist to decide whose video deserves attention, then return to the role you need to fill. Similarity gives the search direction without making the decision.

How can you check an AI scouting answer?

An AI scouting answer should identify its source, comparison group and the assumptions that could change its conclusion. A staff needs evidence it can question.

That database covers 200 leagues and more than 65,000 players, with updates every four hours. Its breadth supports the search, but the question still determines which comparisons matter.

Before signing, the question remains what this player could give your team in the role available. The report should help you investigate that fit and make the remaining uncertainty visible.

Tyler Dorsey’s figures refer to the supplied 2025/26 data: 37 points against Real Madrid, 43 games, 15.9 points per game and 59.5% true shooting. TeamKeys database figures reflect the information supplied for this article.

Questions people ask

What basketball stats should scouts look at?

Scouts should choose statistics according to the role they need to fill, rather than start with a leaderboard. Scoring and shooting figures need to be read alongside league, role and minutes, then examined with video before informing a signing decision.

Does a higher shooting percentage mean a better shooter?

A higher shooting percentage does not automatically identify the better shooter, because volume, shot difficulty and defensive quality affect the comparison. Sample size matters too, so scouts should examine the conditions behind the percentage before judging whether it fits their needs.

How do you compare basketball stats from different leagues?

Start with the original figures and examine differences in pace, physicality, refereeing and defensive quality, alongside each player’s role and minutes. TeamKeys displays an adaptation estimate beside the raw figure, keeping the recorded performance separate from the estimate.

Can AI choose which basketball player a club should sign?

AI can help narrow a search and retrieve evidence, but statistical similarity cannot decide whether a player suits a team.

← All articles