Why models matter: finding and evaluating bets
How models and data automation help us evaluate thousands of bets, with research and judgment shaping each decision.
Fourth & Value · Learn
Evergreen fundamentals: how sportsbooks set prices, what a bet really costs, and how to compare an opinion with the market. For dated game analysis, see the Blog.
Three reads, in order. Together they cover the core idea behind everything on this site: the price matters as much as the pick.
What the vig is, why −110 on both sides adds up to 104.76%, what the book really keeps, and how to strip it out to get fair probabilities.
Take the best price on each side across books and a 4.5% hold drops to about 3%. How to calculate it, and when it goes negative.
A model disagreement shrank to almost nothing, while the same bet paid at one book and lost at another.
Concepts that don't expire with a game. Examples may come from a past slate; the idea still applies.
How models and data automation help us evaluate thousands of bets, with research and judgment shaping each decision.
Compare averages, hit rates, push protection and prices, then check the context before betting.
Overtime, shootouts, regulation-only lines and scratched players. Why matching rules comes before comparing prices.
Does overtime count, what a tie does to your moneyline, and when an injured player’s prop still stands.
Rain-shortened games, action vs. listed pitchers, extra innings and the one-plate-appearance rule.
Overtime in totals and props, half and quarter bets, and what happens when a player sits or leaves early.
Why minutes move props more than matchups, how back-to-backs and the injury report work, and what happens when a star sits.
The summer’s biggest moves, the NBA markets on our board, and what books charge on props compared with game lines.
Break-even rates, overround versus hold, and how to devig a price.
Best price on each side, fair odds from the whole market, arbitrage and its catches.
How markets are made, why sharp prices matter, and how to check your book’s price against a no-vig fair price.
Why we shrink model projections toward the market, and why line shopping can matter more than the pick.
Why sharp bettors bet early, what it costs them, closing line value and middles.
Ratio screens, the low-volume trap, NBA combined props, and how to check a gap with probabilities.
Every outcome of the prop pair that looks free, the math, and what real arbitrage looks like.
Measured on 2012–2025 games: about 1.4% for yardage, about 5% for touchdowns, and why personal splits don’t predict.
Calculators that run in your browser. Nothing you enter is sent anywhere.
Enter each book's price on both sides of the same line. See every book's hold, the best price per side, the synthetic hold, consensus fair odds and the stake split when a market goes negative.
Convert American and decimal odds to implied probability, remove the vig to get fair odds and hold, and check a bet’s break-even and expected value.
Compare your price with the closing line: the closing no-vig probability, your CLV in points and the return it implies.
Turn a Kalshi price in cents into American odds and a break-even win rate with the fee included, or find the Kalshi price that beats a sportsbook line.
Short narrated explainers, about two minutes each.
The break-even point at −110, overround, and devigging.
Building the best market across books, fair odds, and why a negative hold is rarely free money.