Outcome bias in sports betting: judge the bet, not the result

You bet a team at -110. They lead by 10 with two minutes left, give up two late scores and fail to cover. The next night you take a long shot on a whim, and it cashes at +400.

Which bet was better? Most people answer by checking the scoreboard. That habit has a name: outcome bias.

What outcome bias is

Outcome bias is judging a decision by how it turned out instead of by what you knew when you made it.

In 1988, psychologists Jonathan Baron and John Hershey published a study called "Outcome bias in decision evaluation." People read about a doctor who recommended heart bypass surgery that carried an 8% risk of death. Some readers learned the surgery worked. Others learned the patient died. The facts and the choice were the same. Readers still rated the doctor's decision as better when the outcome was good. Many of them said outcomes shouldn't matter, and they showed the bias anyway.

Former professional poker player Annie Duke calls the same habit "resulting": using the outcome as the only test of whether a decision was good or bad.

In betting, it's everywhere. A win feels like proof you were right. A loss feels like proof you were wrong. On its own, neither one proves much.

Good bets lose all the time

Take a bet at -110 that you've priced as a 55% chance to win. The breakeven win rate at -110 is 52.4%, so on paper this bet has an edge.

Here's the expected value on a $110 stake:

• Win 55% of the time: +$100
• Lose 45% of the time: -$110
• EV = (0.55 x $100) - (0.45 x $110) = +$5.50 per bet

That's a 5% return on what you risk. It also loses 45 times out of 100. If you call it a mistake every time it loses, you'll call a positive bet wrong almost half the time.

Now flip it. A bet at +150 needs to win 40% of the time to break even. If the true chance is 35%, the EV on a $100 stake is (0.35 x $150) - (0.65 x $100) = -$12.50. That bet still cashes about one time in three. When it does, it feels smart. It wasn't.

Short runs lie

The problem gets bigger over a weekend or a week, because small samples swing hard.

Say every bet you make is that same 55% bet at -110. Here's how often you'd still be down money:

• After 10 bets: about 50% of the time
• After 50 bets: about 39% of the time
• After 200 bets: about 22% of the time

Read that again. A bettor with a real edge on every single wager is still in the red after 200 bets more than one time in five. A losing month doesn't mean your process is broken. A winning month doesn't mean it works.

How to grade the decision instead

Ask these questions before you look at the score:

• What price did I get? Write down the odds and the implied probability.
• What did I think the true chance was, and why? One or two sentences is enough.
• Did the line move my way? Closing line value tells you whether the market agreed with you after you bet. If you took a team at -3 and it closed at -4.5, you beat the closing number, win or lose.
• Did I size it right? Same unit as always, or did a bigger stake sneak in after a loss?
• Did anything happen that I couldn't have known? A mid-game injury or a fluky bounce isn't a judgment error.

Then record the result in a separate column. Over hundreds of bets, results start to reflect the quality of your process. Over ten, they barely tell you anything.

Two traps to watch

Learning the wrong lesson from a win. A hunch bet cashes, so you start betting more hunches. That can feel great right up until the math catches up.

Dropping a sound method after a loss. A cold week makes a good approach feel broken. Before you change anything, check two things. Were you beating the closing line? Did your bets meet your own price? If both answers are yes, the week was likely just variance.

Keep the log honest

The simplest fix is a record. Before each bet, note the price, your estimate and your reason. After the game, add the result and the closing line. Review in batches of 50 or more, not one night at a time.

Results still matter. Over time, they show whether your estimates hold up. They just shouldn't be the only grade you give any single bet.

The DARIMARKET board ranks every play by confidence and shows the price it needs to beat.

If it stops being fun, call or text 1-800-MY-RESET (1-800-697-3738). 21+.