MU Management / R:R Lab — the 10:10 A+ short, managed 4 ways
A follow-on to the MU “VWAP Reject — Short Continuation” playbook, which flagged the management gap: “scalped the −2σ touches instead of holding a VWAP-trailed runner — the management gap, not the read.” This quantifies that gap on the REAL 2-min path.
LAB · UNVALIDATED · n=1
A management/discipline study on one trade path. The point is the discipline (size the A+, then ride the runner from VWAP), not a proven number — a single path can't validate an edge, and the “ladder optimum” below is an isolated peak, not a robust rule. Nothing here is promoted to the Pine / HUD.
+0.06R
Actual — what I did
A+ block covered in 3 min (+$22.80)
+1.54R
Playbook rule (held)
−2σ/−4σ partials + runner to VWAP-reclaim
+2.47R
Ladder optimum (isolated)
100% runner on the 10-min trail
+4.59R
MFE ceiling (diagnostic)
low 950.27 @ 15:26 — never an exit
THE SETUP — ONE A+ SHORT, RISK FIXED AT THE SIGNAL
The 10:10 signal (from the playbook)
R:R throughout = risk is |entry − stop| at the 10:10 bar (the shakeout-high + 0.25·ATR stop); reward is each variant's realized, in R. The trade never stopped (MAE only ) and bled from ~1046 to the 950s.
Your attached TV read = the anchor
The attached chart is your live J3P panel on MU 7/2 with the ladder framing drawn: T1 +2SD 80% off / T2 +4SD 20% runner → exit on BACKSIDE, stop below the shakeout, and the R:R panel. This lab takes exactly that framing and grades it on the real path from the 10:10 short.
The interactive chart below reproduces that read (entry / stop / −2σ / −4σ / VWAP + 2-day AVWAP) and marks where each management variant would have covered. (Static screenshot embed pending the PNG on disk; the live chart is the reproducible equivalent.)
THE REFERENCE READ — THE REAL PATH FROM 10:10
MU 2-min — the 10:10 A+ vwap_reject_short and where each variant covers
2 · PLAYBOOK RULE — “track from VWAP for the trend”
TRAP
THE LADDER SURFACE — SWEEP t1% / t2% / trail
Reuses ladder_blended_R from J3P_EXIT_SURFACE / J3P_LADDER_OOS: t1% at −2σ, t2% at −4σ, remainder (runner) trailed on the higher-TF higher-low. Blended R over THIS path. n=1 → each cell is one number, not an average; read the SHAPE, not the decimal.
HONEST
THE AUTOMATED SHORTS — EVERY J3P_PLAYS SHORT TRIGGER ON THIS TAPE
Play with the ladder:— recomputes every row live (JS mirror of ladder_blended_R)
Time
Play
Entry
Stop
−2σ / −4σ
100% @ T1
60/20/20 drop
YOURS 20/60/trail
Your mix ⟵ live
n=1 best
HONEST
PYTHON ↔ JS BLENDED-R RECONCILE
The client mirrors the Python ladder_blended_R exactly (as the exit-sweep pages require). It recomputes every surface cell from the recorded path and diffs against the Python values embedded here.
CHECKreconciling…
THE TAKEAWAY
The read was A+ and the risk was tiny (MAE , never stopped). The whole trade was the management: the A+ block was rung out in three minutes for +0.06R, while the same position held under the playbook rule was worth +1.5R and, on this path, up to +2.5R on a loose trail — with a +4.6R ceiling. He did recapture most of the move (+$90.81) — but as ~30 fragmented re-shorts (churn: more decisions, more commission, more screen-time), not one clean held runner.
Size the A+, then hold from VWAP. On a fresh + aligned short where VWAP never reclaims, the trend rail holds all session — the runner is free to ride.
“Cover on VWAP reclaim” means a CONFIRMED reclaim, not a single poke. The naive one-close-above-VWAP rule whipsawed out at 10:12 for −0.26R; the trailing higher-low (the ladder trail) is what actually rides it.
n=1 honesty: the ladder “optimum” is an isolated peak on this one path — a discipline illustration, not a validated number. The lesson is the process.
HOW THIS WAS BUILT
Bars/bands: compute_2min (verified OHLC4 VWAP/σ parity) + add_avwap2 (2-day AVWAP). The 10:10 signal from scan_symbol_day's vwap_reject_short.
Path + R at every rung: record_path from j3p_exit_sweep — causal (≤ current bar); MFE/MAE diagnostic-only, never an exit rung.
ACTUAL: reconstructed from the reconciled fills (DATA-TRUST GATE PASS, +$110.86 penny-reconciled); the A+ block realized reconciles to the fills.
Ladder: ladder_blended_R swept over t1%/t2%/trail; the JS mirror reconciles to it exactly.
Frame: LAB, not cockpit — n=1 discipline study; nothing promoted.