Platform — Branch Logic
Four branches. Four jobs. Zero identity drift.
Layer 01 — Why branches exist
Total return hides which engine is working.
Without branch decomposition, a trader may wrongly believe the whole system is broken when only one branch is deteriorating — or believe the system is healthy while one strong branch hides weakness everywhere else. Branch-level EV, hit rates, and quotas make the diagnosis sharp.
- Normal — the stable compounding engine. Static, controlled, the baseline.
- Trend Partial — the compromise: bank partial value early, keep runner participation.
- Trend No-Partial — the fat-tail branch: full exposure for 3R–5R+ continuation, paid for in variance.
- Overflow — controlled supplemental flow. Never a permission slip to overtrade.
Layer 02 — The iron rule
Branch identity is chosen before entry.
The single most common branch failure is drift: a Normal trade that “becomes” a trend trade after entry because it started running. Branch must be selected pre-entry from setup, ATR regime, and doctrine. The variant layer can tune management inside a branch — it never chooses the branch.
How MARS uses this
MARS never trusts a single blended figure. Weekly branch probabilities produce a separate EV per branch, weighted by the live profile (45 / 33.75 / 11.25 / 10 by default) into the blend. Reviews read the decomposition first: which branch produced the edge, which one leaked it, and whether the weights still match the evidence.
How it benefits you
You stop repairing the wrong thing. When results dip, the decomposition shows whether the whole system weakened or one branch is dragging - so you fix a branch instead of overhauling a working system, and you spot a single strong branch masking weakness everywhere else.
Branch-level EV decomposition with blend weights. The blended number alone would hide exactly this picture.
Interactive — drive the formula
Instrument 02 / 03 — Branch Expectancy
The live Normal-branch formula, in your hands.
Not an illustration — the exact expectancy equation the Weekly Scorecard runs on the Normal branch. Two checkpoint probabilities in, one governed verdict out.
Branch EV / trade
+0.33R
GREEN
Breakeven P(1R)
48.1%
your margin: +15.9pp at this conversion rate
EV = p₁ · (0.5 + p₂) − (1 − p₁) · Runner that stalls after 1R is modeled at breakeven.
The checkpoint funnel — per 100 trades
Entered
100
Survived to +1R · half banked at +0.5R
64
Runner converted to 2R
37
Stopped before 1R · full −1R
36
Where the expectancy comes from — three legs, exact sum
Loss leg
-0.36R
stops before the checkpoint
Partial leg
+0.32R
banked at 1R — the branch's floor
Runner leg
+0.37R
2R conversions — the branch's engine
The three legs sum to +0.33R exactly. Watch what the loss leg costs even in a green profile — that is the tax the two checkpoints exist to govern.
This is one branch of four. The Scorecard runs Normal, Trend Partial, Trend No-Partial, and Overflow simultaneously, blends them by the active weight profile, and grades every week. The full four-branch Scorecard ships in MARS →
Open the full Foundry Lab sandbox ↗Computed locally in your browser. Nothing is uploaded.
Reference
MARS is a closed loop, not a pipeline. Trades produce evidence, evidence produces weekly grades, grades move gate state, gate state feeds the throttle, and the throttle governs what the next execution is allowed to be. No stage is optional and no stage runs on memory or mood - each one reads the outputs of the stage before it.
The MARS operating loop: execute, capture, score, gate, throttle, deploy - every cycle feeding evidence into the next.
Layer 03 — The aggregation trap
One blended number would hide exactly what you need to see.
Aggregate expectancy is a diplomatic average: it lets one strong branch cover for a decaying one until the strong branch has an ordinary cold streak and the whole book appears to collapse at once. Separating Normal, Trend Partial, Trend No-Partial, and Overflow into distinct evidence streams means each management philosophy is judged on its own record — and repaired, weighted, or retired on its own merits.
Layer 04 — Blend governance
Weights follow evidence, and evidence gets audited.
The default 45 / 33.75 / 11.25 / 10 profile is a starting posture, not a commitment. Variant attribution tracks what each branch contributed against what its weight predicted, and the weights migrate toward demonstrated quality on review cadence — never mid-week, never on feel. The blend is therefore a living allocation with a paper trail, which is what separates weighting from wishing.
The governing idea
Inside this module
5 pages go deeper than this one.
Go deeper
Operator briefs on this territory.
Deep dive — 01
The four branches: one edge, four ways of monetizing it.
Not four edges — four exit architectures. Identity is declared at entry and never drifts.
Read the full brief →
Deep dive — 02
How four branch EVs become one system EV — and why weights are the profile.
Branch EV → Trend Blend → All-Blended, weighted by the active profile. The weights are the strategy's mix, stated as numbers.
Read the full brief →
Deep dive — 03
The variant axes: time and exposure, and the aggression they trade.
Time = when protection starts. Exposure = how much size stays uncapped. Five locked variants per branch, zero improvisation.
Read the full brief →
Every module ships in the complete MARS package.
One price. Eleven workbooks, three TradingView indicators, and the full manual library — $497.