Most basketball played on earth has no tracking data, no play-type logging, and no synced film library. It has a box score, a schedule, and whoever was in the gym. That is the condition almost every coach reading this actually works in, and it is entirely possible to scout well inside it.
What it is not possible to do is scout well by accident. Without a data layer doing the noticing for you, the method has to supply the discipline the platform would have supplied.
Decide what you are counting before you watch anything
This is the whole method and everything else is detail.
With a tracking platform you can ask questions afterwards, because it recorded everything. Without one, you can only answer questions you decided to ask in advance. So the first job is not watching. It is choosing four or five counts, writing them down, and committing to them for the whole scout.
Choose them from what would change the plan. If the plan hinges on whether their five can pass out of a double, then count catches inside the elbows and what happened next. If it hinges on whether their point guard goes left, count that. Nothing else.
The temptation is to count everything and stay flexible. Resist it. A scout with five reliable counts beats a scout with twenty unreliable ones every time, and the twenty are always unreliable because attention is finite.
The box score is a better instrument than people think
It carries more than the four numbers everyone quotes.
Distribution tells you the shape. Who takes the shots, who gets to the line, who gets the assists. A team where one player has a third of the attempts is a different scouting problem from one where four players share them, and both are visible without a single tracked possession.
Free throw rate is the closest thing to a free tracking stat. Attempts relative to field goal attempts tells you who lives inside and who settles, per player, over a season, from public data.
Rebounding splits tell you their crash rule. A team with a high offensive rebound rate is sending bodies, and a team sending bodies has fewer bodies going the other way. That is a transition finding derived from a rebounding number.
Game logs beat averages. The average is what they usually do. The logs tell you what they do against teams like you, which is the actual question. Pull the three most similar opponents on their schedule and read those three games instead of the season line.
Your eyes are the tracking layer, so treat them like one
Two rules make live watching produce data rather than impressions.
Count in a fixed unit. Possessions, not "a lot". Twelve of their nineteen half-court possessions started with a side ball screen is usable. "They ball screen constantly" is not, and it is the same observation.
Write the number before you write the conclusion. Otherwise the conclusion arrives first and the number becomes a search for evidence. Every scout I have got badly wrong started with a read I liked.
What you cannot do, said plainly
You cannot get reliable play-type efficiency by eye over one viewing. You cannot get defensive matchup data at any useful precision. You cannot get shot quality, only shot location, and only roughly.
So do not build a plan that depends on those. The absence of the data is not the problem. Building a plan that quietly assumes you have it is the problem, and it happens most often to coaches who have previously worked somewhere with a platform and have not adjusted their instincts.
What actually gets lost, and what does not
What gets lost is breadth. A platform lets you check twenty things cheaply, and by hand you can check five. That is real and it is the honest cost.
What does not get lost is the read. The judgement about which two things decide the game, and what we do about them, is not a data problem and never was. Coaches with a legal pad have been making that call correctly for fifty years.
A league with no tracking data does not make you a worse scout. It makes you a slower one, and it removes your ability to be lazy, because every number on the page had to be chosen and counted by somebody who understood why it mattered. There are worse constraints to work under.