No stats degree required. LineHype does three things: shows you the honest price of every game, compares it to our own model, and keeps a public record of every read. Here's the whole thing, start to finish — and every underlined word explains itself if you tap it.
Every game has a number attached to it — the line. That number is the betting market's best guess at what will happen, and the market is sharp: it moves on real information, not TV narratives. But the number you see is never quite honest, because the sportsbook builds in its own profit first. LineHype takes that number apart, cleans it up, sets it next to our own independent model, and shows you where the two disagree. Then we write down what we thought — publicly — so you can check our work later.
That's it. The rest of this page is just those steps, slowly.
Every betting line quietly includes the sportsbook's profit margin — the . It's why the two sides of a game add up to more than 100%: the book wins either way. We strip that margin out, so the number you're left with is a fair, honest probability instead of one tilted in the house's favor. This clean-up is called .
// the number a book shows you is inflated. de-vigging normalizes both sides back to a true 100% so the market's real opinion is visible.
Separately from the market, we run our own model. At its core is an — one number for how good each team is right now, borrowed from chess. From the gap between two teams' ratings (plus a nudge for home-field), the model produces a : our independent estimate of who wins tonight, with no idea what the betting line says.
// elo_v2 updates after every game; margin- and opponent-adjusted. it never sees the market, so it can disagree with it honestly.
Now we have two honest guesses: the market's cleaned-up and our model's win probability. We line them up side by side — — and measure the distance between them. That distance is the . A big edge means the math and the money disagree, and that's the game worth a second look. It is not a tip to bet — it's a flag that something interesting is happening.
// we flag any gap of 5 percentage points or more. edge ≠ a lock. it's a disagreement between two smart estimates.
Every read we publish goes into a public ledger — win or lose, with the date and the number we called. Nothing gets quietly deleted after a bad night. The whole point is that you judge us on the full record, not the highlight reel. That's what separates a data desk from a guy yelling locks on the internet.
// every call timestamped and scored against the closing line. the ledger is the accountability layer.
It's not a tipsheet. We don't sell picks or promise you winners. We show you the honest numbers and where they disagree; what you do with that is your call.
It's not financial advice. Nothing here is a recommendation to wager money. It's data and analysis for fans who like the numbers.
It's not narrative. No "they wanted it more," no power rankings by vibe. If we can't put a number on it, we don't pretend to know it.
One email at 7am ET: the day's biggest market moves, the data behind them, and the one read worth reading — across every sport we cover, in under five minutes. It's the easiest way to see all of the above in action.