Evaluating a rounded bottom as a testable payoff structure

An edge is a nonzero expected gain

A trading game or pattern has an edge when average expected gain per play is not zero. A useful setup is one that can be taken from the positive side of that imbalance.

When outcome probabilities and payoffs are known, an edge can be computed directly as net profit divided by the sum of notional bets, as in a 15% coin-game example and an approximately 3.33% dice-game example. The notional-bet is the unit stake used to scale profit and loss so an edge can be expressed as expected return per unit risked.

When probabilities and payoffs are unknown, as with chart patterns, an edge is estimated from many historical cases or from simulation that totals net profit and divides by the sum of notional bets. That ratio is the historical-edge: the average net profit divided by the sum of notional bets across a large set of objectively defined historical pattern cases.

Define the cup before you sample it

A usable pattern definition must be specific enough to collect comparable cases. Computer-definable cup or rounded-bottom rules are preferred because they support objective sampling.

In this archive sense, a rounded-bottom is a cup or rounding-bottom price structure whose bullish implication is treated as a testable hypothesis rather than a visual story. Editorial note: without that specificity, later profit and loss figures mix unlike events and cannot be read as one payoff structure.

Bound failure, then let winners expand

Evaluating a rounded bottom requires a definite failure threshold, typically a stop-loss scaled to the pattern's dimensions, so losses are measured uniformly when the bullish implication fails. A stop-loss is a predefined invalidation level, scaled to the pattern, that marks failure and caps the loss when the setup does not do what it was supposed to do.

A trailing-stop is an exit that follows price after entry so winning outcomes can expand while remaining tied to the pattern's dimensions. The two exits turn the pattern into a payoff structure: bounded losses when the bullish implication fails, and larger wins only if price continues after entry.

What the historical sample measured

A 2017 historical test of cup or rounded-bottom bullish implications over 1982 to 2014 produced a mean profit factor of 0.166, interpreted as about $0.16 average profit per $1 notional bet.

In that same test, only 39% of trades were profitable, consistent with many small losses under an initial stop-loss and fewer but larger wins under a trailing stop.

The test's profit-and-loss measurement used a trailing stop that had to pass through the closing price, booked the exit at the exit bar's low, and included gaps beyond the trailing stop.

Editorial reading of theoretical-versus-real-edge

The theoretical-versus-real-edge is the gap between an edge estimated from historical pattern outcomes and the edge a trader actually captures in live execution.

Editorial interpretation: the archive shows a rounded-bottom can be written as a falsifiable payoff structure and then scored as a historical-edge. The close-through trailing-stop rule, the exit at the bar low, and the inclusion of gaps are part of that score. Whether a mean profit factor of 0.166 is large enough to survive those same frictions in real time is a live-execution question, not a result the archive settled.