Published on February 22, 2026 | 3 min read
Most money lost on sports betting is not lost to bad luck or to a missing statistic. It is lost to a handful of well-documented reasoning errors that feel like judgement while they are happening, and each of them has a name.
None of what follows makes anyone a winning bettor — the margin built into the price does not move. What it does is stop a bankroll from being spent on decisions nobody would defend the next morning.
The four that cost the most
Chasing losses is the sunk-cost fallacy with money on it. The loss has already happened and cannot be recovered by the next bet; the expected value of that next bet is exactly what it would have been after a win. Every staking plan that increases the stake after a loss is this error formalised.
The gambler’s fallacy is the belief that an independent sequence owes a correction. It is named after a night at the Monte Carlo casino in August 1913 when a roulette wheel landed on black twenty-six times in a row and the house took a fortune from players betting ever larger sums on red. The twenty-seventh spin had the same probability as the first.
Overconfidence after a win is outcome bias: a bet that came in is treated as evidence that the reasoning was sound, when a single result says almost nothing about a process. And recency bias prices the last match too heavily — a market has already absorbed that result, usually faster than you have.
| Bias | What it feels like | What it costs | The counter |
|---|---|---|---|
| Chasing losses | Getting it back | Stake size grows as judgement falls | Stake decided before the odds are opened |
| Gambler’s fallacy | It is due | Betting against an unchanged probability | Treat every event as independent |
| Overconfidence after a win | I read that correctly | Larger stakes on weaker reasoning | Record the reason, not the result |
| Recency bias | They are in form | Paying for information the market already has | Compare your price with the closing price |
| Confirmation bias | The research agrees | Only the supporting facts get counted | Write the case against the bet first |
What actually counters them
Four habits do most of the work, and all of them are boring. Decide the stake before opening the odds, so the number is not chosen by the market. Write down, in one line, why the bet is being placed — a reason that cannot be written down is usually not a reason. Set a session limit in money and in time before the first bet. And review bets against the closing price rather than against the result: beating the closing line is the only cheap signal that a process has any edge at all.
The arithmetic that sits under all of it is in the house edge of whatever is being played — the games where that number is genuinely lower is the useful companion piece, and how odds and random number generators actually work explains where the margin lives. For the money side, bankroll sizing and what the operator’s own terms do to it matters more than any single bet, and the history of the slot machine shows how long these same instincts have been designed for. The vocabulary is in our betting glossary.
Common questions about betting psychology
Does recognising a bias stop it?
Not by itself. What stops it is a rule made in advance — a stake fixed before the odds are seen, a session limit, a written reason — because the rule does not need judgement at the moment judgement is worst.
Is chasing losses ever rational?
No. The loss is sunk and the next bet’s expected value is unchanged by it. Raising the stake after a loss only concentrates risk into fewer, larger outcomes.
How do I tell skill from luck in my own results?
Compare the price you took with the closing price. Consistently beating the close is weak but real evidence of an edge; a positive balance over a few dozen bets is not.