How TDC Bet works

The model

Every spread starts with The Depth Chart's power ratings: each team's projected 2026-27 roster, player by player, priced from real box-score production, transfers and development curves. Once games start, results move the ratings every night, opponent- and venue-adjusted, with the preseason projection counting as eight games.

The spread is the rating gap plus home court (worth more against weak visitors, and measured for each arena). The total comes from tempo × efficiency: how fast both teams play and how many points each scores and allows per 100 possessions. The moneyline is our win chance written as a price, with no vig.

Reading the board

Each game is a ticket: away team on top, home team below, with Spread, Total and Moneyline boxes. Before books post, the boxes hold our numbers and our favourite's boxes are blue. Once a book posts, the book's number is big and ours sits underneath. A green box is the side our model likes by a real margin. Tap any box to add it to your slip.

Best bets

The edge board (the top of Tonight) prices every market the books post: spreads, totals, moneylines, team totals, first-half and alternate lines, and player props. Each one is priced at every sportsbook, and only the bets worth more than they cost are kept. For each bet you get the best price and where it is, our chance next to the market's, the expected value per $1, and a quarter-Kelly stake for your bankroll. Pick your books under My books to see only prices you can actually bet.

Our chance is blended with the market before anything is ranked. Sportsbooks are sharp: a model that ignores them finds an "edge" on every game. A bet makes the board only when our number and the market's together still beat the price. A line that's far off the other books gets flagged as possibly stale instead of recommended.

Player props run on a model calibrated against two seasons of game logs: when it says 60%, it happened about 60% of the time. It uses our projection, this season's games, minutes, pace and the opponent's defense.

The Angles section is the older card: backtested spots, before you check the price. A bet only makes the Angles card when there's evidence behind it: a spread 3+ points off the book, a total 4+ off, or a prop spot our backtest likes and our projection agrees with. Confidence (one to five bars) comes from the size of the edge or the backtest's hit rate. Some nights nothing qualifies, and we'd rather show nothing than force it.

Every line is logged the morning of the game and graded on the Record page. Losses stay there.

Props

Player production per minute drifts back toward a player's own career. A player whose last season ran well below his career pace tends to bounce back (overs); one coming off a spike tends to regress (unders). We tested that on two full seasons of game logs. The line we show is last season's pace × this season's projected minutes, the number a book anchored on last year would hang.

Futures

We play the whole season 2,000 times: every regular-season game, every conference tournament on its real bracket, the committee's selection and seeding, then March. A team's title chance is the share of seasons it wins. Win totals use the same seasons. Award odds come from projected lines and team records, calibrated on past voting.

Glossary

Spread

Points the favourite gives. Duke −4.5 has to win by 5 or more; the other side covers if they lose by 4 or fewer, or win.

Total

Combined points scored by both teams. Bet over or under the number.

Moneyline

Just pick the winner. −150 means bet $150 to win $100; +130 means $100 wins $130.

Vig (juice)

The book's cut. At −110 on both sides, you need to win 52.4% just to break even.

Fair odds

Our chance written as a price with no vig. A bet is worth it only when the book pays more than the fair price.

Edge

How far our number is from the book's: points on spreads and totals, percentage points of win chance on moneylines and futures.

ATS

Against the spread: a team's record when you count covering the line, not winning the game.

CLV

Closing line value: getting a better number than the line closes at. The most reliable sign a bettor has an edge.

Unit

A standard bet size, usually 1–2% of a bankroll. Records in units compare bettors who stake different amounts.

Kelly

A formula that sizes a bet by its edge. Most bettors use a quarter to a half of it. Calculator →

Parlay

Several bets that all have to win. Payouts multiply, and so does the vig.

Push

The result lands exactly on the number. The bet is refunded.

Play it safe