How to Read Your Expectancy & SQN Numbers (And Why They Matter More Than Win Rate)

Two Metrics for Reviewing a Trading Sample

Two statistics answer different questions about the same recorded sample:

• Expectancy — The sample's average net result per unit of recorded protective-stop risk
• SQN (System Quality Number) — A score using the sample mean, standard deviation and count

Pro's Expectancy Engine summarizes recorded R-multiples. Its metrics describe the available sample, not a guaranteed trading future.

What Is Expectancy?

Expectancy is the average amount you expect to make (or lose) per dollar risked, expressed in R-multiples.

The Formula

Expectancy = (Win% x Avg Win R) - (Loss% x Avg Loss R)

Where:

• Win% = Percentage of trades that are winners
• Avg Win R = Average R-multiple of winning trades
• Loss% = Fraction of the sample with negative R; breakeven trades belong to neither wins nor losses
• Avg Loss R = Absolute magnitude of the mean negative R-multiple; fills, gaps and costs can move it away from the planned 1R loss

Example

Let's say over 100 trades:

• You won 40% of the time (Win% = 0.40)
• Your average winner was 2.5R (Avg Win R = 2.5)
• Your average loser was 1.0R (Avg Loss R = 1.0)

Expectancy = (0.40 x 2.5) - (0.60 x 1.0) = 1.00 - 0.60 = 0.40R

In this illustrative sample, the average result is 0.40R. At $100 of planned risk per trade, that historical sample averages $40 per trade; future results can differ.

What the numbers mean:

• Positive expectancy (above 0) — The measured sample has a positive average result.
• Zero expectancy — The sample's mean R-multiple is zero.
• Negative expectancy (below 0) — The measured sample has a negative average result; investigate costs, data quality and the trading process.
• Larger positive values — Check whether a few outliers, missing costs or a small sample explain the result. There is no universal profitable-trader range.

The Dollar Projection

The Expectancy Engine doesn't just show you the R-multiple — it projects the dollar value. If your expectancy is 0.35R and you risk $200 per trade, your expected profit per trade is:

0.35 x $200 = $70 per trade

For 200 hypothetical trades with that unchanged average, the arithmetic is $14,000 before any excluded costs. This is a projection from an assumption, not a return estimate. Now you can evaluate whether your system is worth your time — not based on feelings, but on math.

What Is SQN (System Quality Number)?

Expectancy measures the sample mean. SQN scales that mean by variability and sample size; neither proves that the process will remain profitable.

The Formula

SQN = (Mean R / StdDev R) x sqrt(n)

Where:

• Mean R = Average R-multiple across all trades (same as expectancy)
• StdDev R = Standard deviation of your R-multiples (how much individual trades vary)
• n = Number of recorded R-multiples used by the app; the current implementation does not cap this count at 100

Why standard deviation matters

Two traders can both have 0.40R expectancy:

Trader A: Wins are consistently 1.5-2.5R. Losses are consistently 0.8-1.2R. Low variance. Smooth equity curve.

Trader B: Occasional 10R winners mixed with lots of -1R losers. Same average, but wild swings. Emotional rollercoaster.

For equal trade counts and the same positive mean R, the sample with lower standard deviation has the higher SQN. Compare all three inputs before attributing a score difference to consistency.

The App's SQN Review Scale:

• 1.6 - 1.9 — Below average on the displayed review scale; this label does not establish that the system works.
• 2.0 - 2.4 — Average. Room for improvement.
• 2.5 - 2.9 — Good in the displayed review scale; not proof of a future edge.
• 3.0 - 5.0 — Excellent in the displayed review scale; the score is sample-dependent.
• 5.0 - 6.9 — Superb in the displayed review scale; check the sample and data quality.
• 7.0+ — Very high on the displayed scale. Check outliers, count and data quality.

The Expectancy Engine shows a visual scale so you can see exactly where you fall.

The R-Distribution Histogram

Below your expectancy and SQN scores, you'll see a histogram of your R-multiple distribution across 7 buckets:

• < -2R — Catastrophic losses (these should be near zero with good stops)
• -2R to -1R — Large losses
• -1R to 0 — Small losses (expected)
• 0 to 1R — Small wins
• 1R to 2R — Moderate wins
• 2R to 3R — Large wins
• > 3R — Home runs

What to look for:

Healthy distribution:

• Tall bar at -1R to 0 (lots of small, controlled losses)
• Decent bars at 1R-2R and 2R-3R (good winners)
• Small or empty bar at < -2R (you cut losses fast)

Unhealthy distribution:

• Tall bar at < -2R (you're holding losers too long or not using stops)
• Tall bar at 0 to 1R (you're cutting winners too short)
• Empty bars above 2R (you never let winners run)

The histogram makes the abstract concept of R-multiples visual and actionable.

Win/Loss Streaks

The Expectancy Engine tracks:

• Current Win Streak / Current Loss Streak
• Maximum Win Streak / Maximum Loss Streak

Why streaks matter

The length of a losing sequence depends on the number of trades, their probabilities and their dependence. An observed win rate alone cannot determine the next streak.

A past six-loss streak does not prove that the next drawdown will recover. Compare the current sample, costs and conditions before attributing a change solely to psychology.

Inspect individual exits and costs separately. A short winning streak does not establish that winners were cut early.

How to Improve Your Expectancy

Compare the components before changing the strategy:

• Win and loss frequency — Compare defined setups and conditions using complete records.
• Average winning R — Compare planned and actual exits, including costs and trades that reversed.
• Average losing R — Review protective stops, fills, gaps and fees.

Changing an exit or filter can affect more than one component. Trade frequency changes exposure and total outcomes; it does not by itself improve expectancy per trade.

The Minimum Sample Size

A critical warning: small samples can be dominated by a few outliers. Thirty trades is not a universal reliability threshold; sample size, dependence, variability and changing conditions all matter.

The engine computes from the data available. Keep the count and distribution visible, and treat a historical estimate as provisional rather than a proof of reliability.

This is why logging every trade matters. You need the data volume. Skip trades and your statistics lie to you.

Keep the sample complete and inspect the trades behind an aggregate before acting on the number.

For a worked payoff comparison, read win rate versus expectancy. Use the equity curve and drawdown guide to review the path those averages can hide.

The app includes only closed trades with a finite R value based on a recorded protective stop in these R metrics. Trades without that risk basis can still have P&L and streak outcomes, but do not become invented R values. SQN requires at least five such values and nonzero variation; that display threshold does not make a small sample reliable.