PHASE 1.20 — FINDINGS SUMMARY

The confidence tiers were based on visible evidence: runs of three, four, five or six qualifying cards.

Both intuition branches performed worse than the textbook control on average:

1. After low-card runs, standing because high cards felt due reduced the average ending bankroll. The mildest confidence tier caused the largest loss because it activated most often.

2. After high-card runs, hitting 15 or 16 against dealer 2–6 also reduced the average ending bankroll. Approximately 61% of those extra hits busted the player across the tiers.

3. Raising the threshold reduced the number of altered decisions and brought results closer to the textbook control. It did not create an advantage.

4. The independent holdout repeated the same direction of result.

5. Deterministic replay passed.

These findings do not criticise the player. They show why an apparent sequence can create strong confidence even when the complete population of outcomes does not support changing textbook play.
