Pass marks and tolerances
Every grade in this app follows a published rule. The numbers below are read from the code that grades, so this page cannot drift from it. Pilot standards are written the same way: a steep turn passes when altitude stays within 100 feet of the entry altitude, not when it looks about right.
The seven-step chart reading
Each step is graded Right (1 point), Partly (0.5) or Not yet (0). The process score is the sum, out of 7. Charts carry their reference answers by construction, so the same answers always get the same grade. A typical candle range is the chart's own unit of distance.
| Step | Right | Partly | Not yet |
|---|---|---|---|
| Trend | The answer is one the chart accepts (a few charts accept two). | Not used: a trend call is right or not. | Any other answer, or none. |
| Tops and bottoms | Every confirmed top and bottom is marked and no mark sits inside a leg. A mark finds a turn of the same kind within 1 candle of it. Equal tops or bottoms within 0.2 typical candle ranges count as one turn, whichever is marked. Marks on a small swing (the highest high or lowest low of 2 candles on each side) or on the last 2 candles before the pattern, where no turn is confirmed yet, cost nothing. | At least one turn found, with a turn missed or a mark inside a leg. | No turn found, or no marks. |
| Support and resistance | Every level that matters has a line within 0.9 typical candle ranges of it, and there are no other lines. The coach prints this distance in price on every chart. | At least one level matched, with a level missed or an extra line. | No level matched, or no lines. |
| Price action | The answer is one the chart accepts. | Not used. | Any other answer, or none. |
| Chart pattern | The answer is one the chart accepts. | Not used. | Any other answer, or none. |
| Candle pattern | The planted pattern, or any other name whose rule also holds on the decision candle. | The twin of the right name: the same shape with the other trend before it (a hammer called a hanging man). | Any other name, or none. |
| Trade plan | The side follows the reading (long, short or stand aside), and a long or short places its lines well: the stop beyond the candle pattern, at least 1 typical candle range and at most 4 typical candle ranges from the entry, and a target of at least 1.5R. | The right side with lines that need work. | A side against the reading, or none. |
The standard plan the coach compares with puts the stop 0.15 typical candle ranges beyond the pattern (and never closer than 1 typical candle range) and the target at 2R. The reveal plays 16 candles of a driftless random walk, so the result in R is shown and never graded: a good plan can lose and a poor one can win.
Lessons
- Passed: at least 80% of a lesson's checks right, rounded up.
- Reviews come back after 1, 3, 7, 21, 60 days. A pass moves the next review further out; a miss brings it back the next day.
- Mastered: passed again at a spaced review, at least a day after the first pass.
- Opening a lesson does not count as learning it; answering a check marks it practised.
Checkrides
- Track A: one full simulator exam of 10 new charts. Pass with an average process score of at least 5.5 of 7 and a Brier score of 0.2 or less on the confidence given. The result in R is never counted.
- Tracks B to E: 12 questions from the track's drills, about 15 minutes, no feedback until the end. Pass with at least 10 of 12 right (80%, rounded up) and a Brier score of 0.2 or less.
- The Brier score is the average squared gap between the confidence given and the outcome: 0 is perfect, and always saying 50% scores 0.25.
Arena and worlds
- The biased coin: $25 to start, heads 60% of the time, up to 60 flips and a $250 cap. The review compares your flips with Kelly's 20% on heads.
- 80 in 8: 80 arithmetic questions in 8 minutes, the one place in the app where speed counts.
- Make a market: quotes are scored on how close they are centred on the expected value and how well the spread fits the uncertainty, not on the luck of the trades.
- Worlds: a research budget of 1500 past charts, then a committed estimate with a 80% interval and 50 forward trades, with costs of 0.05R a trade. The score out of 100 gives 50 for the edge call, 25 for an interval that catches the true edge and 25 for the forward play compared with the best possible.
- The patience drill: 30 charts, 3 with a planted edge. The score adds up the true expectancy of the trades taken, so standing aside costs nothing.
Charts here are simulated. Passing any of these shows practice measured in this app; it is not a professional qualification.