Stage 1 → Stage 2 Transition

What this screens for. Weinstein Stage 1→2: a stock leaving a long accumulation base as its 30-week MA rolls from falling → flat → up. The longer the base, the bigger the expected move. Deliberately earlier than a confirmed-leader screen — it buys the turn, not the trend.
The meter on each card spans the base: left edge is the base low, the vertical tick is the pivot (base resistance), the dot is the current price. A dot past the tick has cleared the base.
Mandatory gates — EQ series only · ≥₹20cr/week turnover · price > 30wk MA · 30wk MA slope ≥ 0 · base ≥30 weeks · depth ≤40% · within 10% of pivot.
Score = Base structure 30 + 30wk-MA transition 25 + Trigger 20 + Volume 15 + RS 10 (the five bars).
RS is only 10 points and is scored as improving, not already-high — a stock emerging from a long base has by definition been underperforming, so demanding strong absolute RS would reject every genuine candidate. That is the deliberate opposite of the confirmed-leader screen, where RS must be 25/25.
Two stops on every card. The base-low stop is Weinstein-correct for a position trade but implies 25–40% risk. The probe line is the tradeable one: entry at/above the pivot, stop below the pivot cluster and never closer than 1.5 daily ATR, R:R measured to the P&F count. R:R < 2 is the framework's own viability floor and is flagged UNVIABLE. Breakout systems run ~45–50% win rates — the payoff ratio is the edge, so the R:R gate is not optional.
RS pct is the 26-week return percentile vs all ~700 scanned names (Minervini gates leaders at ≥70; here it is context, not a gate — see the RS-asymmetry note above).
Check the earnings calendar before any entry (house rule) — a fresh probe into a results date is a coin flip; the screen does not read earnings dates.
P&F 2%×3 (the position-trade box) is now on every card: the column (X demand / O supply), whether a BUY is live or what price would trigger one, and the vertical count target. P&F has no time axis — it filters the noise a base breakout is most vulnerable to, and the count is an objective price derived from column width. That is why a count IS shown here while a measured move off the base is not: the count is computed, the measured move is a guess. Both come from the same ta_score.pnf_read(), so the number here and in a single-stock run can never disagree.