ScoutRoom's base tier is built from public box scores. Saying that out loud is a commercial risk: "just box scores" sounds thin next to tracking data and film platforms. So this piece is the honest audit: what a season of box scores supports, what it never will, and why the line between the two is where a useful scout lives.
What a season of boxes supports
Take a full season of box scores for one opponent, every game, both teams, and you can establish more than most coaches expect.
Who they actually are. Not the roster on the website; the rotation on the floor. Who starts, who closes, whose minutes are trending, which bench player has quietly become a twenty-minute regular. Availability patterns show up too: the connector who has missed four straight, the import whose minutes dipped after the bye.
How the offence is wired. Usage concentration falls straight out of the numbers. One alpha carrying the scoring with everyone else spotting up reads completely differently from five players between eight and thirteen points, and the defensive plan for each is different. Assist patterns say whether the ball moves or stops. Volume from three says who genuinely lets it fly and who just stands out there.
What beats them. This is the underrated one. Line up a team's losses and the box scores of those games, and a blueprint usually falls out: every team that beat them reached a certain score, or won the turnover battle, or held their alpha under her season line. In the sample teardown on this site, one look at the losses showed all four winning teams scored 90-plus, and the three heaviest defeats came against teams that sped the game up. That's a game plan founded on nothing more exotic than reading the box scores of four losses properly.
The economies. Turnover pressure, free-throw traffic, rebounding share, foul counts. Each one is a lever you can pull or a leak you can attack, and each is counted, not inferred.
What it will never tell you
The box score records outcomes, not mechanisms. It says she scored 22; it doesn't say whether the twos came off ball screens she'll get again on Saturday or broken-play scraps she won't. It doesn't know where shots came from, who was guarding, what coverage was played, or what the coach calls after a timeout. A poor defensive number can't distinguish bad scheme from bad effort from one unplayable matchup. Anyone who claims scheme detail from a box score alone is inferring, and should say so.
That's the boundary film exists to cross, and it's why the confidence stamps matter. In a ScoutRoom teardown, a claim the box score can carry is stated flat. A claim it can't is either checked on film, flagged as inference, or left out. The reader always knows which kind of claim they're holding.
Thin data, read properly
Here's the position I've landed on after a season of building on this: the box score is undervalued at exactly the levels that can't afford anything else. The women's and lower-league game mostly doesn't have tracking data, and won't soon. The choice isn't box scores versus tracking. It's box scores read properly versus box scores skimmed, and the gap between those two is enormous. Most of a workable game plan, the rotation, the wiring, the blueprint, the economies, is sitting in public data that nobody reads systematically because reading it systematically is hours of work per opponent.
That grind is what got automated here. Not to replace the richer sources, but to wring everything the honest source actually contains, label its limits, and leave the coach's attention free for the part that needs eyes. Box data tells you more than you think and less than a vendor claims. A teardown's job is to know that line and say which side every claim sits on.