The concepts our models are built on, written for humans. These are the same ideas the how we grade apply in production — closing line value, vig removal, bet sizing, simulation, and calibration — without the site-specific plumbing.
Closing line value (CLV) measures whether you beat the market’s final number. Why it predicts long-term profit better than short-term results, and how to measure it.
Sportsbook prices contain a built-in margin (the vig). Devigging removes it to recover the market’s implied probabilities — the first step in any serious betting analysis.
The Kelly criterion sizes bets in proportion to your edge and odds. Why full Kelly is too aggressive in practice, and why fractional Kelly (quarter, eighth) is the working standard.
Instead of predicting one score, a Monte-Carlo engine simulates a game thousands of times, plate appearance by plate appearance, producing full outcome distributions.
A calibrated model’s 60% claims come true about 60% of the time. How calibration is measured (Brier score, MAE), and why it beats accuracy talk in sports betting.
The Locked-In Index is a 0–100 measure of how well an MLB hitter is currently seeing and striking the ball. Its three components, their weights, and how to read the number.
The Matchup Score ranks tonight’s projection for a hitter against every start this season, 0–100. What it measures, the per-market breakdown, and how to read it next to the Locked-In Index.
Expected value is the average result of a bet repeated thousands of times. How to calculate it from American odds, and why a +EV bet still usually loses.
A realistic ROI benchmark for sports betting, why published returns above 10% are usually small samples, and how many bets a return rate needs to mean anything.
How an opening number is made, why lines move, sharp books versus retail books, and why "the book wants balanced action" is mostly a myth.
Line shopping means checking the same bet across several sportsbooks and taking the best number or price. On standard markets it is worth more than most handicapping.
Implied probability is the win rate a betting price assumes. How to convert American, decimal and fractional odds, why a market’s probabilities add to more than 100%, and what to do about it.