The bookmaker takes 5% through the vig. Your own brain takes more. The cognitive biases that affect every punter — confirmation bias, the gambler's fallacy, overconfidence, loss aversion, recency bias, the sunk cost effect — are not minor psychological quirks. They are the reason smart people make consistently bad betting decisions. They operate below conscious awareness. They feel like rational analysis when they are happening. And they cost you more money than the vig ever will.
This piece covers the six most destructive biases in sports betting, how to recognise them in your own behaviour, and — most importantly — the practical countermeasures that work. Not "be more rational." Actual structural defences that prevent the biases from reaching the decision point.
1. Confirmation bias: finding what you already believe
What it is: the tendency to search for, interpret, and remember information that confirms your existing beliefs — and to ignore or dismiss information that contradicts them.
How it shows up in betting: you have a feeling Collingwood will cover the line. You open the AFL stats page and your eyes go straight to the stats that support Collingwood covering — Collingwood's strong recent form, the opponent's injuries, Collingwood's record at this venue. You skim past the stats that cut the other way — Collingwood's poor record as a favourite, the opponent's strong defensive rating, the weather forecast that favours low scoring. At the end of five minutes of "research," you have convinced yourself the bet is good. You have not analysed the bet. You have gathered evidence for a conclusion you already reached.
Countermeasure: write the case against the bet. Before placing any bet, write down three reasons the bet could lose. Not vague rationalisations — specific, evidence-based reasons. If you cannot produce three, you have not thought about the bet properly. This exercise forces your brain to engage with disconfirming evidence. It is uncomfortable. That discomfort is the bias fighting back. Do it anyway.
2. The gambler's fallacy: believing in streaks
What it is: the belief that past independent events influence future independent events — that after a streak of one outcome, the opposite outcome becomes "due."
How it shows up in betting: a team has lost five in a row. "They are due for a win." A coin has landed heads five times. "Tails is more likely next flip." But sports outcomes are not independent coin flips — and even if they were, the probability of the next flip is always 50% regardless of the previous five. In sports, the losing streak might actually signal that the team is worse than the market believes — the streak could be information, not noise. Betting on a team because they are "due" is betting against the information the streak contains.
Conversely, the hot hand fallacy: believing a player who has scored well recently is more likely to score well next game. The statistical evidence for hot hands in sports is weak — most apparent streaks are within the range of normal random variation. A player averaging 25 points who scores 35, 38, 32 in three consecutive games is probably just experiencing normal variance around their true mean, not a "hot streak" that predicts continued elevated scoring. Betting on streaks to continue is usually betting on variance to persist, which it does not.
Countermeasure: ask "what would I think if the streak had not happened?" If Collingwood had lost last week instead of winning, would you still want to bet them this week? If the answer is no, the streak is driving the decision, not the underlying analysis.
3. Overconfidence: the punter's universal bias
What it is: the systematic tendency to overestimate your own knowledge, skill, and accuracy of predictions.
How it shows up in betting: every punter thinks they are better than average. In a 2024 survey of Australian sports bettors, over 60% rated themselves as "above average" at picking winners — a mathematical impossibility. The overconfidence is strongest among the most engaged punters — the ones who watch the most sport, read the most analysis, and have the most confidence in their knowledge. Knowledge increases confidence faster than it increases accuracy. The gap between the two is where the money is lost.
Overconfidence produces specific betting behaviours: betting too large (overestimating edge size), betting too frequently (overestimating the number of +EV opportunities), and failing to price-shop (overestimating the value of your selection relative to the importance of getting the best price). A punter who is 20% overconfident in their edge and bets 20% more than optimal stake will eventually go broke even if their underlying picks are genuinely +EV. The stake sizing error amplifies into a bankroll destruction error.
Countermeasure: track everything and believe the numbers.The only corrective to overconfidence is data. Track every bet. Calculate your actual return on turnover. Compare your actual results to your pre-bet confidence. If you rated a bet as "high confidence" and your hit rate on high-confidence bets is 52%, your confidence calibration is off by a large margin. The numbers do not care about your feelings. Trust them over your gut every time.
4. Loss aversion and the sunk cost fallacy
What it is: loss aversion is the tendency to feel losses more intensely than equivalent gains — losing $100 hurts roughly twice as much as winning $100 feels good. The sunk cost fallacy is the tendency to continue investing in a losing position because you have already invested in it — the money is gone but it does not feel gone.
How it shows up in betting: the most common manifestation is loss chasing — increasing bet size after losses to recover the lost money faster. "I am down $200 for the day. If I bet $200 on this next one at $1.90, I will be roughly even." The next bet is selected for its potential to recover losses, not for its expected value. It is typically a worse bet than you would have placed if you were even or ahead. And when it loses — which it often does, because it was selected poorly — the hole is now $400 deep and the chasing impulse is stronger.
Another manifestation: refusing to cash out or hedge a position because you "have come this far." You placed a futures bet on a team to win the premiership at $15. They have made the grand final and are now $2.50. You could lay them on Betfair and lock in a profit regardless of the result. But it feels wrong to bet against "your team" — you have been on them all season. The $15 ticket feels like a commitment, not a financial position. The refusal to hedge costs you guaranteed profit because the emotional attachment to the original bet overrides the rational financial decision.
Countermeasure: hard stop-loss and pre-committed hedging rules. Set a maximum daily loss before the session starts. When you hit it, stop. No exceptions. No "one more to get back." The rule exists precisely because you cannot trust your in-the-moment judgment. For futures bets: decide your hedging plan when you place the bet. "If this reaches $3.00 at any point, I will lay off 50% of the potential profit." The decision is made cold, before emotional attachment forms.
5. Recency bias: the last game is the only game
What it is: the tendency to overweight recent events relative to the full body of evidence.
How it shows up in betting: a team plays one outstanding game — a 60-point win where everything clicked — and the market overcorrects, shortening their price for the next game. Punters pile in because the performance is fresh. The team regresses to their mean the following week and loses as a short-priced favourite. The pattern repeats every round of every season in every sport.
The bookmaker knows about recency bias and prices for it. After a team's big win, the bookmaker shortens the price by more than the underlying probability shift because they know recreational punters will bet the team at almost any price. The post-big-win game is systematically overpriced (from the punter's perspective — the price is worse than fair value). Betting against teams after unusually good performances and on teams after unusually bad performances is a simple strategy that exploits recency bias in the market. It is not a guaranteed edge — the bookmaker is aware of this too — but it is a better starting point than the reverse.
Countermeasure: use a weighted average, not the most recent data point. A team's true strength is better estimated by their last 10-15 games, weighted toward recency but not dominated by the most recent game. A simple approach: weight the last 5 games at 50%, the previous 5 at 30%, and the 5 before that at 20%. The most recent game gets 10% of the weight in this scheme, not 100%. The weighting forces your brain to incorporate the full evidence base.
6. The illusion of control: betting like you have agency
What it is: the tendency to overestimate your ability to control or predict outcomes that are substantially determined by chance.
How it shows up in betting: punters develop rituals and superstitions — "I always win when I bet on Friday night," "I have a feel for this team." The rituals feel like they provide an edge. They provide comfort, not edge. The comfort is dangerous because it increases confidence without increasing accuracy.
More damaging: the belief that more information equals better decisions. Punters consume pre-game analysis, watch panel shows, read match previews, check stats — and feel more confident in their bets as a result. But the information is publicly available and already priced into the market. Consuming it does not give you an edge over the market. It gives you an edge over other recreational punters who consumed less information, which feels like progress but is irrelevant — you are not betting against other punters. You are betting against the bookmaker's margin. Information that is already in the price does not help you beat the margin.
Countermeasure: distinguish between information that is in the price and information that is not. Publicly available stats, team news, and analysis are in the price. The market has already incorporated them. If your entire analytical process is reading things that everyone else can read, you are not generating an edge — you are consuming content. An edge requires either information the market does not have (rare and usually inside information, which is illegal to bet on in Australia) or a better interpretation of publicly available information than the market consensus (possible but much harder than it sounds). Most punters do neither. They consume content, feel informed, and bet with unjustified confidence.
Building structural defences
You cannot think your way out of cognitive biases. The biases operate at a level below conscious reasoning. Willpower fails. "Being more disciplined" fails. What works: structural defences that prevent the biased decision from being made.
- Pre-commit to rules. Stake sizing, maximum bets per day, stop-loss limits, minimum odds thresholds. Set them when you are calm and not in a betting session. The rules are non-negotiable during the session. If you find yourself wanting to override a rule, that is a bias talking. The bankroll management guide covers stake sizing frameworks in detail.
- Track everything. Every bet. Odds, stake, outcome, net result. The aggregate numbers are the only objective feedback on whether you are profitable. Your feelings about how you are doing are wrong. The tracker is right. See the bet tracker setup guide.
- Write the case against every bet. Three specific, evidence-based reasons the bet could lose. If you cannot produce them, you have not done the analysis.
- Use a checklist. Before placing a bet: have I checked at least three bookmakers for the best price? Is the stake within my pre-committed limits? Have I written the case against? Is this bet consistent with my strategy or is it an impulse? The checklist forces analytical engagement at the moment of decision.
- Separate research from betting sessions. Analyse during the week when no money is at stake. Place bets during a defined window based on the week's analysis. Never analyse and bet in the same session. The presence of money degrades analytical quality.
Frequently asked questions
Can I train myself to not have these biases?
No. Cognitive biases are features of human cognition, not bugs. They persist even when you know about them — knowing about confirmation bias does not prevent you from confirming your existing beliefs. The goal is not to eliminate the biases. It is to build systems that prevent them from reaching the decision point. A pre-committed stake sizing rule prevents overconfidence from affecting bet size. A hard stop-loss prevents loss aversion from triggering chasing. The system works when you follow it even — especially — when your biased brain is telling you to override it.
Do professional bettors have these biases too?
Yes. The difference is that professionals have structural defences in place and years of tracked results that provide objective feedback. When a professional's tracker shows a 3% ROI over 5,000 bets, they trust the number over their gut. When a recreational punter's gut says they are winning but they have no tracker, they trust the gut. The professional still feels the biases — the urge to chase, the overconfidence after a winning streak, the despair after a losing one. They just do not act on those feelings because the system prevents it.
Sarah covers the sport-by-sport pricing landscape and the wider betting culture. Reports on tipster schemes, social-media betting scams, and the specific market inefficiencies that show up in AFL, NRL, and NBL player props.