Cognitive biases in betting
Cognitive biases are systematic errors in human reasoning that cause judgments to deviate from rationality. In betting, these biases distort the player's estimation of probabilities, his evaluation of expected value, and his staking decisions. Because profitable betting depends on accurate probability estimation, understanding and mitigating cognitive biases is essential.
The biases described here are not character flaws — they are features of human cognition that affect all people, including experienced players and experts. Awareness of their existence does not automatically prevent them; deliberate, systematic processes are required to counteract their influence.
Gambler's fallacy
The gambler's fallacy is the belief that past random outcomes affect future random outcomes. For example, after observing five consecutive losses, a player might believe that a win is "due" — that the probability of winning the next bet has increased because of the preceding losses.
This is incorrect. If the events are independent, the probability of each outcome is unaffected by previous outcomes. A coin that has landed on heads ten times in a row still has a 50% probability of landing on heads on the eleventh flip. The law of large numbers guarantees convergence of the average over many trials, but it does not operate by "correcting" individual outcomes — it operates by diluting the impact of any finite sequence as the total number of trials grows.
In betting, the gambler's fallacy can lead to increasing stakes after losing streaks, expecting a reversal that probability does not support. This behavior is a common cause of bankroll depletion.
Favorite-longshot bias
The favorite-longshot bias is the empirically observed tendency for odds on low-probability outcomes (longshots) to be less favorable than odds on high-probability outcomes (favorites), relative to the actual probabilities. In other words, the bookmaker's margin tends to be distributed unevenly: longshot options carry more margin than favorites.
This bias has been documented across many sports and markets. Several explanations have been proposed:
— Recreational players disproportionately bet on longshots because the potential payoff is large and exciting, allowing bookmakers to offer worse odds on these options.
— Risk-averse bookmakers charge more margin on longshots because the potential payout per bet is high.
— Players systematically overestimate the probability of low-probability events.
For the informed player, this bias has a practical implication: value may be more frequently found on favorites than on longshots, because favorites tend to carry less margin. However, this is a statistical tendency across markets, not a rule that applies to every individual bet.
Confirmation bias
Confirmation bias is the tendency to search for, interpret, and recall information in a way that confirms one's existing beliefs, while disregarding information that contradicts them. In betting, this manifests in several ways:
— A player who believes a team will win may selectively focus on statistics that support this belief (recent wins, home advantage) while ignoring contradicting data (key player injuries, poor performance against similar opponents).
— After placing a bet, a player may seek out analysis and opinions that confirm his choice, reinforcing his conviction rather than critically evaluating the decision.
— A player may remember his successful predictions more vividly than his unsuccessful ones, creating a distorted perception of his accuracy.
Confirmation bias is particularly dangerous because it undermines the objectivity of probability estimation. The remedy is to actively seek disconfirming evidence and to use systematic, data-driven methods for probability estimation rather than relying on subjective assessment alone.
Anchoring
Anchoring is the tendency for initial information to disproportionately influence subsequent judgments. In betting, a player who sees the bookmaker's odds before forming his own probability estimate may unconsciously anchor his estimate to the implied probability of the odds, rather than deriving an independent estimate.
For example, if the bookmaker offers an odd of 2.50 (implying 40%), a player who might otherwise have estimated the probability at 50% may unconsciously adjust his estimate downward toward 40%, reducing the perceived edge and potentially causing him to pass on a profitable bet.
To mitigate anchoring, the player should form his probability estimate before looking at the bookmaker's odds. This ensures that the estimate is based on analysis of the event rather than on the bookmaker's pricing.
Overconfidence
Overconfidence is the tendency to overestimate the accuracy of one's own judgments. A player may believe his probability estimates are more precise than they actually are, leading him to stake too aggressively. If the Kelly criterion is applied with an overestimated probability, the resulting fraction is larger than optimal, increasing the risk of substantial drawdowns or bankruptcy.
Overconfidence is particularly prevalent after winning streaks, when the player attributes his success to skill rather than to variance. It can also manifest as an underestimation of the difficulty of beating the market — the belief that the player's edge is larger or more consistent than it actually is.
The use of fractional Kelly staking is, in part, a structural defense against overconfidence: by wagering less than the theoretically optimal fraction, the player builds in a buffer against the possibility that his probability estimates are wrong.
Sunk cost fallacy
The sunk cost fallacy is the tendency to continue an action because of previously invested resources (money, time, effort), rather than evaluating the action on its current merits. In betting, this manifests as a reluctance to stop a losing strategy because of the money already lost, or as increasing stakes to "recover" previous losses.
Each bet should be evaluated independently on the basis of its expected value. The money lost on previous bets is irrelevant to the expected value of the current bet. A player who increases his stakes not because the expected value justifies it, but because he wants to recover past losses, is making a decision driven by the sunk cost fallacy rather than by rational analysis.
Recency bias
Recency bias is the tendency to give disproportionate weight to recent events when estimating probabilities. A team that has won its last five matches may be perceived as more likely to win its next match than the full body of evidence would support. Conversely, a team that has lost its last three matches may be underestimated.
Recent results do contain information — they may reflect changes in form, injuries, or tactical adjustments. However, giving excessive weight to a small number of recent results at the expense of a larger statistical sample leads to inaccurate probability estimates. The appropriate approach is to incorporate recent results as one factor among many, weighted in proportion to their statistical significance.