Most sports bettors ask one question: who do I like tonight?
The sharper question is different: is the price right?
That's what moneyline edge is. And once you understand it, you'll look at every pick you see — ours included — in a completely different way.
What a moneyline actually says
A moneyline is a sportsbook's price on which team wins. The format in the US is American odds:
- +150 means you risk $100 to win $150 — the book thinks this team wins less than half the time.
- -130 means you risk $130 to win $100 — the book thinks this team wins more than half the time.
But the number on the board isn't the "true" probability. It's a price. The book builds in a margin (the vig), so the implied probabilities of both sides always add up to more than 100%.
Strip out the vig and you get the book's implied probability for each side.
The math (it's simple)
Underdog (positive line):
p = 100 ÷ (ml + 100)
So +150 implies: 100 ÷ 250 = 40% win probability
Favorite (negative line):
p = |ml| ÷ (|ml| + 100)
So -130 implies: 130 ÷ 230 = 56.5% win probability
What edge means
Edge is the gap between your model's probability for a team and what the line implies.
Edge = model probability − implied probability
If our model says a team wins 62% of the time and the line implies 56.5%, the edge is +5.5%.
A positive edge means the book is pricing that side cheaper than your model says it should be. You're getting more than fair value.
A negative edge means you're paying too much for the side you like.
Why edge matters more than the pick
Two picks can both be "correct" — meaning the right team wins — but have completely different expected value.
Example:
- Pick A: team wins 55% of the time. Line implies 52%. Edge: +3%.
- Pick B: team wins 55% of the time. Line implies 60%. Edge: -5%.
Pick A and Pick B are both bets on teams that win more than half the time. But Pick A is profitable long-term. Pick B is not — you're overpaying for that side every time.
Over hundreds of picks, this difference compounds dramatically.
Early season edge is noisy
Early in a baseball season, sample sizes are small. A team that goes 7-3 in its first 10 games looks very different from one that goes 7-3 after 70 games.
Our model applies early-season shrinkage — meaning we weight current results less and lean more on expectation when samples are thin. This makes edges less extreme in April than they are in August. That's intentional.
It also means early edge readings move more day to day. Don't over-index on a single +10% edge in the first two weeks of the season.
How we use it
Every pick we post includes the line we had when the card ran and the model edge vs that number. If there's no line yet, we note that.
We track edge over time the same way we track W/L — because a positive-edge approach that loses short-term isn't broken, but a zero-edge approach that wins short-term is just variance.
The record matters. The edge calculation is what makes the record meaningful.
For informational use only. Past results don't guarantee future performance. Bet responsibly.