AI BEHAVIOR SCORING

How we score discipline.

Six categories. 100 points. Pass at 90+. P&L is never an input. The rubric is fully published, fully deterministic, fully auditable.

PASS THRESHOLD

Cross 90, advance.

90 / 100
Pass threshold to advance

What gets scored.

Each trade you log is graded against six behavior dimensions. Weights are fixed. Math is shown. No subjective coach mood, no surprise rules.

30 pts

Rule Adherence

Did you follow your written plan? Entries within session windows, no revenge trades, no rule overrides. The largest single weight, because rules are everything.

20 pts

Stop-Loss Discipline

Stop placed before entry, never widened, honored on touch. Moving a stop against you zeroes this category for the trade.

15 pts

Journaling Completeness

Every required field filled β€” setup, thesis, screenshot, post-trade reflection. Empty fields cost points even if the trade was perfect.

15 pts

Position Sizing Consistency

Contract count matches your plan for the setup. Doubling up on a hunch is a full deduction even if it works.

10 pts

Plan Adherence

Did the trade match a setup written in your plan before the session? Off-plan trades score zero here regardless of outcome.

10 pts

Behavior Streaks

Consecutive sessions of clean execution. Rewards consistency. One reckless session breaks the streak and the bonus.

What is NOT scored

These are deliberately excluded. We are scoring the trader, not the market.

TIE-BREAKING

Deterministic, in this order.

  1. Higher Rule Adherence sub-score wins.
  2. Higher Stop-Loss Discipline sub-score wins.
  3. Lower count of journaling gaps across the contest period wins.
  4. Earlier qualifying timestamp (UTC) wins.
  5. Coin flip seeded with both user IDs β€” fully reproducible if anyone disputes.
DISPUTES

Seven days. Same algorithm.

Open a dispute within 7 days of any score you believe is wrong. We re-run the exact same deterministic scoring with your raw data, return the new output and the diff against the original. No human override, no negotiation β€” if the algorithm produces the same number, the score stands. If it produces a different number, we honor the corrected score.

The Free Leaderboard uses the identical rubric, identical math, identical thresholds. Practicing for free is practicing the real thing.