The Gambler's Fallacy: Why 'It Must Rise This Time' Drains an Account | NOONOO TRADING
After four consecutive losses, you approach the fifth trade thinking, “A win must be due,” and want to bet more than usual. The gambler's fallacy is the mistaken belief that earlier outcomes change the next outcome's probability. In trading, it commonly appears as increasing size after losses, a combination that can drain an account quickly.
What is the gambler's fallacy?
The gambler's fallacy assumes that past results influence the next result in independent trials. After six heads, thinking tails is now due is a typical example. A coin has no memory; the next toss still has a 50% probability of tails.
The term comes from gambling, but trading can make it especially expensive. Casinos impose betting limits, while traders choose their own leverage and position size. A feeling that a win is due can translate directly into quantity.
The error misapplies the law of large numbers to a short sequence. Across thousands of tosses, the proportion tends toward 50:50. This happens because many new outcomes dilute the initial imbalance, not because later outcomes reverse what already happened. Markets do not repay a debt.
Three forms in trading
“After four losses, the next trade is more likely to win” → increase usual risk from 1% to 3%.
2. Decrease size after wins
“After five wins, a loss is due” → reduce size at a valid opportunity.
3. Reverse the direction
“Five bearish candles mean the next must be bullish” → enter against the move without other evidence.
The first two act in opposite directions but share the same assumption: the sequence contains predictive information that has not been established. Their consequences differ. The second may miss profits; the first directly expands exposure during losses.
The third can resemble a justified countertrend entry based on support or resistance. The source offers a distinction: if the only reason for entry is the streak count, it is the gambler's fallacy. “Five bearish candles” without further evidence is a feeling rather than a rule.
What is the win probability after losses?
Consider a strategy with a 45% win rate. After four losses, what is the probability that trade five wins?
Strategy win probability: 45%, assuming independent trades.
Probability of four losses = 0.554 ≈ 9.2%.
Probability the fifth trade wins after those losses = 45%.
Probability the fifth also loses = 55%.
→ Probability increase attributable to being “due” = 0%.
The source describes four-loss sequences as ordinary over 100 trades. As discussed in losing-streak probability, it also treats eight consecutive losses as within a normal range for a 55% win-rate strategy. A streak's length alone does not determine the next outcome.
The financial damage comes from tripling the bet in response.
Account: ₩10 million. Normal risk: 1%, or ₩100,000.
After four losses: approximately ₩9.6 million.
Increase fifth-trade risk to 3%, about ₩290,000.
Win probability remains 45%; loss probability remains 55%.
A loss leaves about ₩9.31 million.
Increase again to 5% on trade six: approximately ₩470,000.
The source warns that climbing this ladder another three or four steps can push
the required recovery return above 20%, turning decisions into gambling.
Required recovery returns accelerate as losses deepen. A 20% loss requires 25% to recover; a 50% loss requires 100%. Increasing exposure during a growing drawdown confronts this asymmetry directly.
Martingale: turning the fallacy into a system
Martingale says to double after each loss so one win recovers the losses. The source describes it as a formalized version of the fallacy. Its idealized logic requires unlimited capital and unlimited betting capacity, neither of which exists in practice.
Starting bet ₩100,000, doubled after each loss:
Trade 1: ₩100,000; 2: ₩200,000; 3: ₩400,000.
4: ₩800,000; 5: ₩1.6 million; 6: ₩3.2 million.
7: ₩6.4 million; 8: ₩12.8 million.
Total committed through eight losses: ₩25.5 million.
At a 45% win rate, eight-loss probability = 0.558 ≈ 0.84%.
The original guide states that a ₩10 million account cannot keep up by the sixth step,
and describes the destructive sequence as roughly a one-in-119 event.
Although 0.84% sounds rare, five trades a day produces more than 100 trades in a month. The source argues that encountering such a sequence should be planned for. The core problem is that one failure can exceed all previously accumulated profits. Averaging down can share this risk: the issue is expanding exposure during losses, rather than reducing average entry cost by itself.
The opposite belief: the hot hand
The hot-hand fallacy grows in the opposite direction from the same root: “I have won three in a row, so I have the touch,” followed by a larger bet.
Gambler's fallacy: after losses, “A win is due” → increase size.
Hot-hand fallacy: after wins, “I am on a roll” → increase size.
Shared assumption: recent outcome order predicts the next result.
Shared consequence: predefined sizing rules change with circumstances.
Both can coexist in one person. Increasing after losses and after wins means inventing a new reason to increase every time. The actual sizing rule becomes mood, undermining meaningful performance evaluation. Add outcome bias, and a lucky oversized winner can validate the habit.
Are markets actually independent?
Unlike coin tosses, market returns are not perfectly independent. Volatility clusters, and trending periods can contain consecutive candles in the same direction.
This does not justify the fallacy. The source argues that observed autocorrelation more often supports continuation than reversal, and is often too small to cover fees and slippage. “Five bearish candles mean a rebound” is not supported merely by that observation.
Trade outcomes can also cluster because a market regime does not suit the strategy. In that case, the source favors reducing or pausing exposure, since the next trade may face the same regime. The gambler's fallacy prescribes the opposite response.
Response rules
1. Fix risk per trade before trading; do not change it because of the preceding result.
2. In this framework, size adjustments depend on account balance, such as reducing risk after each 10% balance decline.
3. During losing streaks, stop the day with a daily loss limit instead of increasing size.
4. Include “previous outcome” and “current size” together in the trading journal.
5. Skip an entry if its only rationale is a streak count.
The fourth rule can expose the bias directly. Compare average size by previous outcome. Figures such as “average risk after a loss: 2.4%; after a win: 1.1%” reveal that size responds to outcomes, regardless of a belief that the rules are being followed. Win rate or PnL alone does not show this.
The balance-based adjustment in rule two moves opposite to the fallacy. When the balance falls, the cash amount represented by 1% also falls. This is a mathematical property that reduces risk of ruin. Increasing size after losses manually removes that property.
Common misconceptions
“A 45% win rate means 45 wins in 100, so missing wins must arrive later.” A probability is a long-run proportion, not a quota. Later wins do not compensate for earlier shortages; larger samples merely reduce the early imbalance's relative weight.
“Increasing after losses speeds recovery.” Only when the larger bet wins. Increasing size during negative expectancy magnifies the negative expectancy too. Ruin accelerates along with possible recovery.
“Reducing after a winning streak is safe, isn't it?” It lowers exposure, but the source questions the reason. Fixed-fraction risk naturally grows in cash terms with a rising balance. An unexplained manual reduction changes the rule and can invite equally arbitrary changes in the other direction.
“Should losing streaks simply be ignored?” Ask a different question: not whether a win is due, but whether the streak is within the strategy's expected range. Repeated Monte Carlo simulations can estimate that range. A result outside it calls for reviewing the strategy, not merely increasing size.
Recap
2. After four losses, the next win probability gains 0%; the market owes nothing.
3. Martingale systematizes the belief, risking a loss larger than accumulated gains in one eight-loss sequence.
4. Increasing after wins through a hot-hand belief has the same root when size reacts to outcomes without evidence.
5. Recording previous outcome alongside current size reveals the bias numerically.
The source's central rule is: size follows the account balance, not the preceding result. Losing streaks are part of a strategy's normal behavior. The response is to keep risk small according to established rules. Survival allows a larger sample to accumulate and the strategy's long-run probabilities to be observed.
NOONOO TRADING invites you to follow live trading in our free chat.
Start in the bot📈 OKX trading fee discount for new registrations
Register for the OKX Fee Discount →