Product
Everything HoopStruct computes, and why nobody else computes it this deeply.
Two pillars: advanced stats built for evaluating players, and high-quality scouting reports built for preparing games, backed by two ways to get a box score in, purpose-picked charts, and Shot Chart Entry for logging shot locations by hand.
How data gets in
Two ways in. One unmatched output.
Both paths produce a real box score. One of them also unlocks the analytics no photo or manual entry can give you.
Photo upload
The easy, AI-powered path: unlimited on the Photo plan.
- Upload a photo of a printed or screenshotted box score
- Claude AI reads it via OCR and extracts a full stat line for both teams
- Fast: works from almost any box score you can point a camera at
Play-by-play import
The differentiator: free, local, and more accurate than either photo or manual entry.
- Import a EuroLeague-style play-by-play Excel export
- Parsed locally: zero API cost, unlimited on the Photo plan
- Exact free-throw splits, exact shot types, fouls drawn, shots rejected
- Real, measured minutes and plus/minus, reconstructed from substitution timestamps and score deltas
As far as we've found, no other consumer-facing tool in this space reconstructs real minutes and plus/minus from a spreadsheet import. Everyone else estimates or leaves it out.
Manual entry is always there too, as a baseline. The principle holds across all three: the more data you give HoopStruct, the deeper it can go: a manually-typed line gets the standard box score and headline metrics, and play-by-play is what unlocks real plus/minus and the Impact Rating.
Advanced stats, unmatched
The analytics stack nobody else in this space provides.
PIR, PER, PIE, a from-scratch Impact Score, the Four Factors, and a real RAPM-style Impact Rating. These are the same formulas real analytics teams use, computed individually and at team level, every game, benchmarked against the league.
PIR
Performance Index Rating
EuroLeague-style production score from a single box score line.
PER
Player Efficiency Rating
Hollinger-style, pace-adjusted per-minute production.
PIE
Player Impact Estimate
NBA's real formula: a player's share of everything that happened on court.
IS
Impact Score
A from-scratch, BPM-style composite built specifically for HoopStruct's data.
DOE
Four Factors + Ratings
Dean Oliver's Four Factors, weighted, plus offensive and defensive rating.
NET
Net Rating
Exact ORtg minus DRtg, for players, built from real +/- when play-by-play exists.
USG%
Usage Rate
Share of a team's possessions a player used while on court.
RAPM
Impact Rating
A real, from-scratch adjusted plus-minus: box score plus lineup data, the same technique behind LEBRON and EPM.
Plus the full taxonomy, not a token few stats
Scoring
Shooting
Rebounding
Ball handling
Four Factors, weighted the way they actually predict winning
Each combo score blends the core factor with real sub-metrics such as Assisted FG%, live-ball turnover rate, opponent OREB%, and more.
One real Impact Rating. Not two borrowed ones.
Built from actual play-by-play lineup data (the same core technique behind LEBRON and EPM), reconstructed straight from your play-by-play import. It ships with a visible confidence level (Very Low → Low → Medium → High) based on how many games actually back the number, instead of two arbitrarily-different commercial numbers dressed up as certainty.
Insights nobody else generates
Scouting reports, written instantly for every game, team, and player.
The kind of breakdown a scout would spend hours building: real comparisons against real season averages, ready the moment a game ends.
Game insights
Pick a saved game and get what went well and what went less well, for both teams, at team and player level.
- · A normally-high scorer going cold
- · A low-volume player breaking out
- · A non-rebounder suddenly crashing the offensive glass
Team scouting report
Strengths and weaknesses versus the league average, plus key players ranked by real PIE.
Player scouting report
The same strengths/weaknesses framing, focused on how one player's stat line shifts between wins and losses.
“How to beat them”
Sharpest insightTeam scouting reports compare a team's own numbers in losses versus wins (turns it over more, gets outrebounded, allows more points, whatever the real pattern is), so an opposing coach knows exactly what to force.
Visualization nobody else offers
Six chart types built for basketball, not spreadsheets.
Each one opens from clicking the relevant stat, with a plain-English line explaining what it shows.
Opta-style radar
A player's percentile profile against the league, across every key stat category.
Shot chart
Every make and miss, plotted exactly where it happened on the floor.
Win probability
How the game swung, possession by possession, from tip to final buzzer.
Ridgeline
A stat's full distribution across a season, layered so the shape is unmistakable.
Bump chart
How rankings (scoring, efficiency, standings) shift and cross over time.
Gravity & movement
Where the offense pulls defensive attention, and how the ball actually moves.
Scroll to follow the data
01. Input
One photo or one spreadsheet
AI reads a box score photo, or a play-by-play export goes straight in. Both produce a full stat line.
02. Processing
Zero guesswork
Every stat is recomputed from the actual data, not estimated, not padded.
03. Advanced stats
8 real metrics, automatically
PIR, PER, PIE, DOE, Net Rating, and a from-scratch RAPM Impact Rating, every game.
04. Insights
Scouting reports, instantly
Game, team, and player breakdowns written the moment the final buzzer sounds.
05. Charts
6 ways to see it
A purpose-built chart for every question, not the same bar chart, six times.
Shot Chart Entry
Click anywhere on a realistic half-court diagram and HoopStruct silently classifies the zone (at-rim, mid-range, corner three, wing three, top-of-key three), then you type makes and attempts for a player or the whole team. No shot-tracking software or stats feed needed, just a coach and a few minutes per game. Builds a per-player-per-zone breakdown and a team total automatically, correctly, even logged over several sessions.
Multi-competition support
The same team or player can play across multiple leagues and cups in one season (domestic league, national cup, and continental competition), including the Greek Cup, Copa del Rey, Coppa Italia, Turkish Cup, Coupe de France, BBL-Pokal, LKL Cup, and Israeli State Cup. Stats are correctly isolated per competition, and also viewable combined across all of them.
Primary, context, and strategic: the honest way
The Four Factors breakdown for any team is organized in three tiers: Primary (what actually happened: ORtg, DRtg, Net Rating, eFG%, TOV%, ORB%, FTr), Context (pace, opponent quality, lineup combinations from play-by-play data), and Strategic (why it happened). The Strategic tier is shown as an explicit N/A with the reason: it needs film and tracking data no box-score-based tool can produce, and we'd rather say so than fake it.
Built for players too
From your desk to every player's pocket.
The first part of HoopStruct built for a player to use directly, not just a coach or analyst.
Coach publishes
A finished scouting report PDF, sent straight from the desktop app to the whole roster.
Player opens the app
Each player opens the companion mobile app and always sees exactly the current report, never a stale one, never next week's early.
Coach sees who's read it
A simple view back in the desktop app shows who's actually opened the report, and who hasn't yet.
See it applied to your own team.
Start with whichever import fits what you already have: a photo, or a play-by-play export.