How to use the model's projections and tiers

The model projects a number for each prop: what it expects the player to do tonight. The gap between that projection and the line is the edge. A projection of 27.1 against a line of 24.5 is an edge of +2.6 toward the over.

The tiers

Picks are grouped by how big the edge is compared with the line:

• A: the strongest edges of the night.
• B: solid edges.
• C: small edges, close to a coin flip.

Bigger edges have hit more often over time, and every pick is graded in public on the record page: by tier, by stat and by week, losing weeks included. Check how each tier and stat has done before you trust it.

What the model knows, and what it doesn't

It knows recent form, minutes, the opponent, and how players do against each position. It is slower with news it can't see in the numbers yet: a late scratch, a coach's decision, a player managing a sore ankle. That's where your own research adds the most.

How to use it well

1. As a second opinion. If your research and the model agree, that's a stronger spot. If they disagree, find out why before betting.
2. With the minutes. An edge built on 34 minutes means little if he plays 26 tonight. Check projected minutes first.
3. With the price. An edge on the line is worth less at a bad price. Shop the line.
4. With patience. Judge it over weeks, not nights. The weekly model status every Monday shows how it really did.

The model is a tool for doing the research, not a substitute for it.