Gambler’s Fallacy — also called the Monte Carlo fallacy — is the cognitive error of believing that a run of independent random outcomes makes the opposite outcome more likely. In trading, it surfaces when a trader sizes up after four straight losses because they feel a win is “due,” or fades a trending stock after six down sessions because “it can’t keep falling.” Both behaviors ignore the core statistical reality: if your edge is constant, each trade is independent — past outcomes do not rebalance the probability of the next.
Key Takeaways
- A losing streak of any length does not increase the probability of winning on the next trade — streaks carry zero predictive value for independent events.
- With a 45% win rate, five consecutive losses have a ~2.5% probability per window and occur roughly every 40 trades — a normal event, not a reversion signal.
- Retail traders who overtrade — driven by overtrading behaviors including streak-chasing — underperform buy-and-hold by 6.5% annually (Barber & Odean, 2000).
How Gambler’s Fallacy Works
The fallacy originates in what Tversky and Kahneman (1971) called the “law of small numbers” — the tendency to expect small samples to mirror the statistical properties of the entire population. A coin that lands heads five times in a row is not “due” for tails; the probability of tails on flip six is still 50%. The same logic applies to trade outcomes.
In trading, the fallacy appears in two distinct forms:
Form 1 — Oversizing after a losing streak. A trader with a 45% win rate suffers four consecutive losses and interprets this as evidence that a win is imminent. They double position size on trade five. Mathematically, the probability of winning trade five is still 45% — unchanged by the prior streak.
Form 2 — Fading a trending instrument. A trader sees NVDA down six sessions in a row and buys the dip purely because “it can’t keep dropping.” This confuses the gambler’s fallacy with legitimate mean reversion analysis. Mean reversion strategies are grounded in structural market factors — not streak counts.
The difference between the fallacy and legitimate pattern recognition is the basis for the decision. Reading volume divergence or order flow to time a long entry is analysis. Changing size or direction because of how many losses appeared in a row is the fallacy.
Streak Probability by Win Rate
For a system with a 45% win rate (55% loss rate), the probability of N consecutive losses:
P(streak of N) = (1 - win_rate)^N
N=3: 0.55^3 ≈ 16.6% (expected every ~6 trades)
N=4: 0.55^4 ≈ 9.2% (expected every ~11 trades)
N=5: 0.55^5 ≈ 2.5% (expected every ~40 trades)
For a 40% win rate system, the probability of six consecutive losses is 0.60^6 ≈ 4.7% — nearly one in twenty 6-trade windows. Many traders treat this as a system failure or a reversion trigger. It is neither.
Practical Example
A day trader runs a $20,000 account with a strict 1% risk rule — $200 maximum risk per trade — scalping SPY on a 5-minute chart. The system has a documented 45% win rate and 1.5:1 reward-to-risk ratio.
After four consecutive stop-outs (-$800 total), the trader convinces themselves the system is “due.” They size up to $600 risk on trade five. That trade also stops out.
Result:
- Expected single-day max loss at $200/trade: -$200
- Actual loss on day (4 normal + 1 oversized): -$1,400
- Drawdown: 7% in one session vs. an expected 1%
When this trader reviews their journal, trades taken after three or more consecutive losses show a win rate of 38% — below the 45% baseline. The streak provided no predictive edge. The increased size multiplied the damage by 3x on the worst trade of the day.
A 45% win-rate system with a 1.5:1 reward-to-risk ratio has a positive expected value of +0.175R per trade. Doubling size after a losing streak doesn’t change that expected value — it doubles the dollar risk on a trade that is still more likely to lose than win. The fallacy doesn’t just feel wrong; the math confirms it is.
Common Mistakes
-
Increasing position size after losses. The instinct to “get it back” by sizing up is the most costly expression of this fallacy. Larger size during a losing streak — when conditions may actually be unfavorable — accelerates drawdowns.
-
Fading strong trends based on streak length. Counting how many sessions a stock has moved in one direction and trading against it purely on that basis is the fallacy applied to price action. Barber and Odean (2000) found that retail traders who overtrade — driven by overtrading behaviors including streak-chasing — underperform buy-and-hold by 6.5% annually.
-
Reducing size after wins. The reverse fallacy — cutting size after a winning streak because “the luck must run out” — is equally irrational and causes traders to undersize during high-confidence setups.
-
Confusing streaks with system failure. A run of six losses on a 40% win rate system is a ~4.7% probability event — statistically expected several times per year. Abandoning or dramatically altering a system after a normal streak destroys expectancy before the edge has time to play out.
How JournalPlus Tracks Gambler’s Fallacy
JournalPlus lets you tag trades by streak context and filter your trade history to isolate entries taken after two, three, or four consecutive losses. Comparing the win rate and average R-multiple of those post-streak trades against your overall baseline makes the fallacy visible in your own data — not just in theory. The streak performance report is available on all plans.
