🎓 Learn Sports Betting Math

Visual explanations · Interactive demos · Real methodology
← Hub 📐 Methodology 💰 Bankroll

📚 The Curriculum

Betting profitably requires understanding 5 concepts in order. Each builds on the last:

  1. Vig — why markets are hard to beat in the first place.
  2. EV — what makes a bet actually worth taking, vig and all.
  3. Distributions — how to estimate probabilities (not as single numbers but as ranges).
  4. Kelly — how much to bet once you know edge + probability.
  5. CLV — how to know whether you're actually improving.

Skip ahead at your own risk. Kelly without EV will ruin your bankroll. EV without distributions is wishful thinking. CLV without all four is a number you don't understand.

Each lesson ~5 min. After the five core lessons, the capstone framework ties them together and the backtest shows what happens when the framework is applied to real data.

1
Vig & Fair Odds
How sportsbooks make money. Why ~52.4% is break-even at -110 odds. The hidden margin you have to beat before any strategy matters.
🔍 Vig calculator · no prerequisites
2
Expected Value (EV)
The single most important concept. Why "high confidence" doesn't always mean "good bet." How a 65% win rate at -200 still loses money.
📊 Interactive EV calculator · requires Lesson 1
3
Probability Distributions
How to estimate confidence as a range, not a single number. Beta for binary outcomes (wins/losses), Gaussian for continuous (margins, totals), Tiers for grouping similar confidences.
📈 Beta visualizer · Score histograms · Tier groupings · requires Lessons 1–2
4
Kelly Criterion (Bet Sizing)
How much to bet once you know edge + probability. Why full Kelly bankrupts you. Why Quarter Kelly works. Why we cap at 2% of bankroll.
📊 Bankroll growth simulator · requires Lessons 2–3
5
Closing Line Value (CLV)
The metric that predicts long-run profit better than win rate. Why our backtest's negative CLV detected failure inside 500 bets. How to track it from bet #1.
📈 Line movement chart · requires all of 1–4
— Beyond the core five —
🎯
Capstone — The Strategy Framework
Putting it all together — the 7-gate decision process, the alternatives we don't use, what we deliberately don't bet on, and when to skip even when the math says yes.
📈 Bankroll growth chart · decision flowchart
📉
Backtest — 13,903 NBA Games
The simplified model run against real historical odds: −95% ROI over 4,615 bets. CLV detected the failure inside 500 bets. The cautionary tale that justifies every gate in the framework.
🔬 Real data · honest results
⊕
Bonus — Tier Clustering (Boris Chen method)
A practical application of Lesson 3's distributions: grouping picks into confidence bands so we don't act on differences that aren't statistically meaningful.
🎯 Tier visualization
⊕
Bonus — Score Distributions (Gaussian)
A deeper dive on the Gaussian tool from Lesson 3, applied specifically to point spreads and game totals.
📊 Score histogram
📅 Coming when we have data:
Lesson 8 — Track Record Analysis (after 20+ bets logged)
Lesson 9 — What Actually Predicts Wins for OUR Picks (after 50+ bets)
Lesson 10 — Bayesian Updating Your Own Model

Right now we have 1 bet logged. Doing performance analysis with N=1 would be theater — patterns from 1 data point are noise. Real learning comes after the sample is big enough.