The Lab
Betting math and market mechanics, explained from the ground up — the same concepts the model runs on. No picks, no hype. Start at Foundations and work down; each level assumes the one before it.
New here? Start here — every term in plain English, a reading path for how much you already know, and every record number we've published, dated, beside today's.
Foundations
| Start here — the Lab in plain English
New to betting, a regular, or a numbers person — here's where to begin, every term the Lab uses in one place, and the record we've published, dated, beside today's. |
2026-10-22 |
| What a betting line really says
A moneyline isn't a prediction — it's a price. Learning to read -157 as a probability is the first skill in evaluating any betting market. |
2026-10-22 |
| The vig — what betting actually costs
The bookmaker's fee is invisible, automatic, and the reason most bettors lose. Here's how to see it and what it does to a season. |
2026-10-22 |
| Is it better to bet favorites or underdogs?
Neither. At fair odds both break even exactly, and at real sportsbook prices both lose at the same rate. Here's the math with $100 bets, and what actually creates an edge. |
2026-10-22 |
| How to track your sports bets (and what to actually record)
A spreadsheet beats no tracking, an app beats a spreadsheet, and the field almost everyone leaves out is the one that decides whether you're winning. Here's the minimum that works. |
2026-10-22 |
| How to win your NFL pick'em pool
Picking the favorites every week gets you a respectable score and almost never wins the pool. Here's the math on why, and what actually separates the winner — for pick'em, confidence and survivor formats. |
2026-10-22 |
Reading the market
| The closing line — the number that grades everything
The last price before a game starts is the sharpest number in betting. Here's why every call should be measured against it, and what we got wrong doing exactly that. |
2026-10-22 |
| Does line movement mean anything?
We measured it across 122 games. Line movement genuinely predicts where the line goes next — and following it still loses money. Here's the number, the two traps we nearly fell into, and what's actually going on. |
2026-10-22 |
| Why books disagree — and whether the softest line is the best bet
Sportsbooks differ because they run inventory, not forecasts. Here's what a soft price actually tells you, worked through one of ours that won and still failed the test. |
2026-10-22 |
| Which sportsbook has the best odds?
We compared eight sportsbooks across 1,440 moneyline markets on our own recorded closing prices. The book that wins most only posts the best number 27% of the time — which changes the question from which book to use to how many to have. |
2026-10-22 |
| Same-game parlays — the correlation is real, and the book charges for it
Bets from one game win and lose together, so multiplying their odds gets the price wrong. We measured what actually moves together — and why a same-game parlay is only worth it at its fair price, which books rarely pay. |
2026-10-22 |
| Player props — the price knows his role before his stats do
We built a stats model for NFL player props and tested it on 592 FanDuel lines before trusting it. It lost to the price. Here's why, what a prop really costs, and how much a hot streak is worth. |
2026-10-22 |
The math
| How often do favorites win? Four sports, measured
Across 20,360 games, favorites won 58.7% in baseball, 62.2% in hockey, 68.0% in basketball and 67.1% in football — each within a point and a half of what their prices said. Why the sports differ, and why betting the favorite still loses. |
2026-10-22 |
| Expected value in sports betting — why the biggest edge is always a longshot
Expected value is one line of arithmetic, and the price term inside it is why a longshot tops almost every +EV screen, including ours. |
2026-10-22 |
| How many graded games until a win rate means anything?
A win rate needs thousands of games to mean much. Here's the arithmetic, our board's record as it has grown from 519 games, and the faster number to check while you wait. |
2026-10-22 |
| Reading a null result properly — the form signal, in full
A p-value of 0.373, an R-sq of 0.41%, and a negative adjusted R-sq. Here is the complete output on our strongest shadow signal, and why "not significant" is the least interesting thing in it. |
2026-10-22 |
| The Brier score — how to grade a probability, not a pick
A pick is right or wrong. A probability is something subtler, and grading it takes a different tool. Here's the Brier score, the market's own score in four sports, and why it settles in hundreds of games what a win–loss record needs thousands for. |
2026-10-22 |
| Signals that predict the line, not the winner
Two of our signals predicted where the closing line would go, at more than three standard errors. Adding either one to our probabilities made them worse. Here's how both can be true, and what a line-predicting signal is actually worth. |
2026-10-22 |
Our method
| Why "no pick" is a feature
Most days the model flags nothing. That silence is the most honest output a betting model can produce — here's the math behind it. |
2026-10-22 |
| We tried seven ways to beat the market. Here's what happened.
Line shopping, sharp books, stale prices, unders, line moves, spreads and arbitrage — tested on 26,000 games with the rules written down first. Six didn't work, one only paid for itself, and one lead survived a later dig. |
2026-10-22 |
| Why we stopped posting long shots
On October 17 our feed posted a +650 college underdog as a "19.7% edge." It lost — but that isn't why we changed the rules. Every pick we had ever published was an underdog, and the arithmetic said it would be. |
2026-10-22 |
Questions we haven't answered yet
Every article starts as a question we couldn't answer honestly in one sentence. These are still open — not knowing yet is part of the method. Send yours through the feedback link below.
- Does closing-line value actually predict long-term profit, or is it just the market grading our homework?
- When a prediction market (Kalshi, Polymarket) disagrees with every sportsbook, who's right more often?
- When the books disagree with each other more than usual, does their consensus get less accurate — or is a wide spread of opinions just as reliable as a tight one?
- Does a sportsbook's same-game parlay price ever beat the fair price of legs that move together — and how often? We're logging FanDuel's prices by hand to find out.
- The rested hockey team beat its price against teams on the second night of a back-to-back in five past seasons. Does it hold on games nobody had seen when the rule was written? Scoring forward since October 10.
- Do props on a hot streak — cleared tonight's line in 7+ of the last 10 — beat their price, or has the market already priced the streak? Promoted only past two standard errors over 100 bets.
- If we trusted any one number we use 10% more, or 10% less, for two weeks, would any version call the games better? Frankenstein runs October 22 – November 4, with its forecast written down first.
- Our pre-rewrite tracker showed +6.9% ROI across 143 bets. We can't yet tell whether that was an edge or the fact that a person chose which 143 of two thousand flagged games to actually bet. How would anyone tell those apart?
Model output is informational and entertainment content, not betting or financial advice. If you bet, bet what you can afford to lose.