This lesson is what makes the A+ series trustworthy — and what makes you a harder trader to fool. The A+ Chop Filter does something almost no indicator does: it keeps score of its own verdicts as if it traded them, with a real stop, and prints the expectancy. That self-measurement is powerful and dangerous in equal measure, and learning to read it correctly is the single best defence against believing a number that means nothing. The rule you'll leave with: read the ± before you read the number.

This is the same discipline behind why backtesting isn't prediction — a result without an honest error bar is a story, not evidence.


The self-measurement engine, in plain terms

For each bucket of verdicts (PASS, BLOCK, MATURED, READY), the filter simulates entering at the flip with a real stop (1 ATR) and target (2 ATR), checks the stop first, and reports E(R) — expectancy in R per trade — plus how far into the move the entry already was. In other words, it grades its own homework against a realistic trade, not against a cherry-picked exit.

Why build that? Because of a trap the series fell into and then caught. An earlier version scored PASS at 76% and BLOCK at 25% — a 50-point gap that read as proof the filter worked. Rebuilt independently with a real stop, the same flips separated by one point: 31% versus 32%.

🚨 DANGER
The gap was manufactured. Blocks fire in chop; in chop the cloud flips back fast, so blocked trades died before they could reach target. The exit condition was correlated with the thing being tested. An indicator that measures itself keeps agreeing with itself — which is why every number is rebuilt offline from independent bars whose job is to refute it.

Expectancy needs an interval

Here's the statistical fact that changes how you read every number on the pane: expectancy per trade has a huge per-sample spread — a standard deviation of about 1.4R. That means a small sample tells you almost nothing. Set an E(R) and a sample size and watch the interval:

At n=50 the 95% interval is roughly ±0.39R — so +0.32R and 0.00R are the same measurement. If the interval spans zero, the bucket has demonstrated nothing. Real thresholds need something like n=400 or more, and at ~1.6 qualifying samples per day, that's hundreds of sessions. This is the same reason a handful of trades can't validate any edge — the law of large numbers is not optional.

💡 TIP
Every E(R) on the pane prints its interval. Your reading habit: look at the ± first. If the band crosses zero, treat the headline number as "unknown," not as small-but-positive. A number without a clearing interval is a rumour.

The READY lesson: a fact is not an edge

The most important honesty story in the series is READY. One session, READY looked great: +0.50R over 48 trades. Twenty sessions later it was +0.02R with an interval from −0.06 to +0.11 — flat at every delay. It did not replicate. The root error was subtle and common: the confidence intervals had been computed on win rates, never on E(R).

The fix wasn't to delete READY — it's a true fact ("this flip's side has held five bars"). The fix was to keep it as an observation, rewrite every claim to print an interval, and never present it as a signal again. That's why, in Lesson 7, READY is explicitly not something you act on. A fact about the chart is not an edge until it survives being tested on data it has never seen.


What the pane is, and isn't

Hold these three readings and you'll never be misled by the dashboard's own numbers:

  • The per-gate and per-bucket percentages are in-sample, current-chart-only — a smoke detector for gross breakage, not a scale you can tune from.
  • Everything is measured on the chart in front of you, which is a description of that chart, not a backtest on unseen data.
  • The number that matters most is often how late the entry already was — winning by better judgement is not the same as winning by getting in later.

Carry this forward and the A+ tools become what they're meant to be: instruments you read critically, not oracles you obey. That habit — read the interval, distrust in-sample, respect the sample size — is worth more than any single setup.

Related reading

Next: Lesson 10 — risk, journaling, and the options reality, the capstone that keeps you in the game.