How to win at sports betting: what holds up over hundreds of bets

Why the price matters more than the pick

At -110 you risk 110 to win 100, so you need to win 52.38% of the time to break even. A book posting -110 on both sides of a market keeps 4.5% of everything wagered on it if the action is balanced, and a bettor picking at random at that price loses 4.5% of every unit staked.

The same picks at a better price are a different business. Win 52% of your bets at -110 and you lose 0.7% per unit staked; win the same 52% at -105, where break-even is 51.22%, and you make 1.5%. Nothing about the selections changed. This is why the rest of this guide is mostly about prices.

Estimate a probability, then compare it with the price

Every bet is a claim that an outcome is more likely than its price implies. Turn the price into a probability, remove the book’s margin so you are comparing against the market’s real opinion, and bet only when your own estimate is higher. That gap is the edge, and expected value turns it into an amount per unit staked.

Your estimate can come from a model, from a sharper book’s devigged price, or from information the market has not absorbed yet. Wherever it comes from, write it down before you look at the price. An estimate formed after seeing the odds tends to drift toward them.

Shop every line

The -110 versus -105 gap above is the whole case for line shopping: 1.2 percentage points of break-even win rate, available by checking a second book. In football, a half-point through 3 or 7 can be worth more than that. Hold accounts at several books and place each bet where the number is best for your side.

What is a unit, and how big should one be?

A unit is the standard stake you choose for one bet, usually set as a small fixed share of the money you have set aside for betting. Records are kept in units so that results compare across bettors and bankroll sizes: +10 units means the same thing whether a unit is $5 or $500.

Size the unit so that a normal losing run does not end you. A bettor who wins exactly the break-even 52.38% at -110 has a 27% chance of at least one run of ten straight losses somewhere in 1,000 bets. At 1% of the current bankroll per bet, ten losses in a row leave 90.4% of it. The Kelly criterion sizes stakes to the edge instead; in practice most bettors who use it stake a fraction of what it says.

Track closing line value, not wins

The closing line is the market’s last and most accurate price before the game starts. If the number you took is consistently better than where the market closed, you are finding real value, and that shows up long before profit and loss settles down. A winning month with prices that lose to the close is a warning sign; a losing month with prices that beat it is often just variance.

Record every bet with its price, its stake and the closing price on the same side. That one habit answers the question of whether your process works faster than any profit-and-loss total.

How many bets before a record means anything?

Far more than most people think. A bettor with no edge at all, one who wins exactly the 52.38% break-even rate staking flat at -110, finishes 95% of the time with a return on investment between −18.7% and +18.7% after 100 bets, between −8.4% and +8.4% after 500, and between −3.7% and +3.7% after 2,500. Any short record, good or bad, is mostly that spread.

This is the yardstick for your own results and for anyone selling theirs. A record means something when it is long, complete, kept at the prices actually taken, and has its losses still visible.

Where a paid service fits

Services sell one of three things: selections, prices and tools, or education. None of them changes the arithmetic above, so judge any of them by the same tests you would apply to your own betting: a full graded record, the price each play was recorded at, and enough bets for the number to mean something.

On this site every staked play is posted with its line, price, book and stake, and stays on the public Track Record whether it wins or loses. The comparison guide linked below lays out what other services publish about their prices and trials, in their own words.

Bets is $29 a month or $250 a year, and All Access is $49 a month or $399 a year. Any monthly plan opens with 7 days for $1.

Bet legally, and only what you can afford to lose

Sports betting is legal only in some places and is age-restricted where it is: at least 18, and 21 in many US states. Check the law where you are before you place a bet. Set the money aside in advance, treat it as spent, and stop if betting stops being something you choose. In the US, call or text 1-800-GAMBLER for confidential help.

Common questions