Every pick we publish starts the same way: with a number.
Before we look at narratives, before we consider who's hot or cold, before any of the noise — we generate a projected spread and total for every game. That projection becomes our anchor. Everything else is measured against it.
This post breaks down how that process works and what we're actually looking for when we release a pick.
The Projection Comes First
We build forward-looking projections based on how teams actually perform, not how they've done against the spread historically.
The inputs include offensive and defensive efficiency, pace, shooting trends, rebounding rates, and how those numbers shift in specific matchup types. For totals, we're looking at combined scoring environments and pace interactions.
None of this is revolutionary. What matters is that it's consistent. The same inputs, weighted the same way, every game. That consistency is what lets us measure whether we're actually finding value or just getting lucky.
Why We Wait
You'll notice we release picks closer to game time than most. That's intentional.
Injury news changes everything. A starting point guard being ruled out two hours before tip can swing a spread by 3-4 points. A key defender being downgraded from probable to questionable matters. Back-to-back fatigue hits some teams harder than others, and the impact compounds when rotation players are already out.
A projection that doesn't account for the actual personnel on the floor isn't complete. We'd rather be late and accurate than early and stale.
What We're Looking For
Edge isn't just disagreeing with the number. It's disagreeing with the number for reasons we can explain and have tested.
When our projection differs from the market, we ask:
- Is this within normal variance, or is there a specific factor driving the gap?
- Have we accounted for injuries, rest, and travel correctly?
- Is the market likely reacting to something we've already priced in?
If we can't articulate why we're different, we don't have edge. We have a guess.
Confidence and Sizing
Not every edge is the same size, and not every edge carries the same certainty.
A two-point discrepancy in a clean matchup where our projection inputs are stable is different from a two-point discrepancy in a volatile spot with multiple injury unknowns. We size accordingly.
Our confidence tiers reflect how much conviction we have in the projection itself, not just the size of the gap. A large edge in a murky situation might warrant less exposure than a smaller edge in a spot we understand well.
Losses Are Part of the Record
We track every pick — wins and losses — with the same level of detail.
There's no selective memory here. If we go 1-4 on a Tuesday, that's in the results. If a model version underperforms over a two-week stretch, we evaluate whether it's variance or something structural that needs adjustment.
Transparency isn't a marketing angle. It's how you learn what's actually working.
The Bottom Line
We're not trying to predict the future. We're trying to identify spots where the market price doesn't reflect the most likely outcome, then bet those spots consistently over time.
Some weeks that works. Some weeks it doesn't. But the process stays the same either way.
That's how you find edge. Not by chasing trends or reacting to last night's box score — by doing the work before the number moves and trusting the methodology you've built.
All content is for informational and entertainment purposes only. Past performance does not guarantee future results. Please gamble responsibly.