PropSurvival · Articles

MC calculation receipt — corpus-bound sample + variance

A depth page cites one address: corpus id + inputs + honest sample outputs. A visitor who pays re-mints under the same corpus id with their portfolio.

hash b2b24d3c · calc-receipt-mc-1 · topstep@40ecf3b35a8ccb90

Public face (free)

Corpustopstep · rules hash 40ecf3b35a8ccb90 · source rules/topstep
Account size$50,000
Sample mean (net R)0.03834R ± 0.010999 SE
Analytical E_net0.05R
Concordance |err|0.01166
Mixture variance1.21 (analytical Var of one net-R draw)
Paths · seed10000 · 704639342
Reference trader50% win · 1.2R / 1R · cost 0.05R

Evidence class: Model-derived Monte Carlo estimate bound to a disclosed rules corpus id. Not an empirical claim about any named market outcome or pass rate.

Citation

PropSurvival calc-receipt MC b2b24d3c · calc-receipt-mc-1 · topstep@40ecf3b35a8ccb90

Reproduce

Regenerate and prove byte-identity:

node tools/build-calc-receipt-mc.js --check

Formula: sampleMean = (1/N) Σ netR_i ; netR ∈ {W-c, -(L+c), -c}; SE = √(var/N)