Cognitive Biases in Stake Decisions

## The Biases That Attack Staking Specifically Many cognitive biases affect selection analysis (discussed in the Betting Psychology topic). A subset of biases specifically distort stake sizing decisions — often in ways that selection biases do not. ## Denomination Effect The denomination effect: people treat larger nominal amounts more carefully than smaller ones, even when the relative value is identical. For bettors: a £100 stake feels significant; ten £10 bets feel trivial. The cognitive cost of each £10 decision is lower than a single £100 decision — but the total exposure is identical. This can lead to under-analysis of small bets and proliferation of low-quality small bets. **Correction:** Evaluate every bet by its expected cost (EV amount), not its nominal stake size. ## Sunk Cost in Betting Once money is staked, it is gone (win or lose). The sunk cost fallacy treats existing losses as a reason to bet more (to recover). The right framework: each new bet is evaluated from zero — what is the EV of this specific bet, independent of all previous results? ## Availability Heuristic in Staking Bettors remember large wins at high stakes more vividly than consistent small wins at normal stakes. This availability makes high stakes feel "worth it" — even when the expected value of the high-stake bet does not justify the increase. **Correction:** Base staking decisions entirely on the Kelly calculation. Vivid memories of large wins are irrelevant. ## Mental Accounting in Betting Treating "house money" (from recent winnings) differently from "original bankroll" money is a dangerous cognitive error. All money in your bankroll is equal — whether won recently or deposited originally. A bet with 2% of house money has identical risk to 2% of original capital. **Correction:** Track total bankroll as one number. Never create mental sub-accounts of "winnings" vs "deposits."