A score between 0 and 100 maps to ten bands from A+ down to F. Thin data caps the band regardless of how clean the setup looks: too few bars of context, too few confirmed swings, or no readable higher timeframe each impose their own ceiling, and each names itself on the verdict.
The bands
Each band is a floor on the score. A grade describes how much of the evidence agrees — it is not a statement about suitability, position size, or whether you should take the trade.
- A+ — a score of 88 or above.
- A — a score of 81 or above.
- A- — a score of 75 or above.
- B+ — a score of 69 or above.
- B — a score of 62 or above.
- B- — a score of 56 or above.
- C+ — a score of 50 or above.
- C — a score of 42 or above.
- D — a score of 32 or above.
- F — anything below 32.
How a score is built
The weighted factor sum runs from −1 to +1 and is mapped onto the published score by a neutral point of 52 and a span of 44, then clamped so that neither a perfect nor a hopeless read reaches the extremes of the scale.
The clamp is deliberate. A score of 100 asserts an absence of doubt that no chart supports, and a score of 0 asserts the same in the other direction.
Data-quality caps
A cap is a statement about the evidence, not about the setup. Each one appears on the verdict with its reason.
- Too few bars of context available — the grade is capped in the B family.
- Fewer than four confirmed swings — capped at B−, because structure cannot be read confidently from less.
- No readable higher timeframe — capped in the B family; a read without context above it is missing an input the score assumes.
- Proximity to a scheduled earnings event — its own cap, applied because a known catalyst makes chart structure less predictive.
- Below the hard minimum number of bars, the read is refused outright rather than graded at all.
Strategy rules act as caps too
Your own rules cap the grade the same way data quality does, each with a plain-sentence reason. A rule that fires does not veto the read; it limits how high the read can be scored.
What a grade is not
A grade is not a probability, and the product does not claim it ranks outcomes. A walk-forward replay found that out of sample the score did not usefully separate winners from losers — that finding is published rather than buried, and it is the reason confidence exists as a separate number.
The claim the product does make is about the calibrated confidence, which is checked against resolved outcomes. Where those two numbers disagree, confidence is the one with evidence behind it.
Frequently asked
- Why is my chart capped at B no matter how clean the setup is?
- A cap names its reason on the verdict: too few bars of context, fewer than four confirmed swings, no readable higher timeframe, or earnings proximity. The cap is a statement about the available data, not a judgement of the setup itself.
- Does an A+ grade mean the trade is more likely to win?
- No. The grade measures how much of the evidence checklist agrees. A walk-forward replay found the score did not rank outcomes out of sample, which is published rather than hidden — the calibrated confidence is the number with a checked claim behind it.
Updated Sep 1, 2026