Every piece in the Resources hub ships with the same four labels pinned to the top of the page: data source, sample window, reproduction steps, and the broker feed we re-ran it against. Nothing is paraphrased from a Reddit thread; nothing is paraphrased from a competitor's marketing post either. If we cannot reproduce the result on a Friday afternoon in Tallinn, we do not publish it on a Monday morning in Zug.
01 · Data
Tick-level, not screenshot
Every playbook and backtest cites its raw tick source — Interactive Brokers historical, Polygon paid tier, or our own NY4/LD4 captures. We publish the dataset fingerprint (symbol, exchange, adjustment basis) so a reader at a different broker can match it to the nearest equivalent feed before running the strategy.
02 · Window
Stated, not implied
Sample windows are written out in plain English and pinned in the page footer — start date, end date, total trades, and the regime the window spans (low-vol grind, gap-and-go, central-bank event). A 2021-only backtest on a mean-reversion strategy is a different artifact than the same strategy across 2018–2024; we do not let the chart do that work for us.
03 · Reproduction
Built inside StrategyLab
Every published backtest links to a runnable StrategyLab module — the same engine that processed 38,400 strategy variations in 2024. A reader with a free account opens the link, sees the exact parameter set, and can rerun the same test on their own book before committing capital or course material to it.
04 · Freshness
Quarterly decay check
Each playbook is re-tested against the most recent quarter the Tuesday after quarter-end. If a 2022 momentum playbook stops working in Q2 2024, the page header carries a "decay flag" with the date and the rolling Sharpe delta. We do not quietly retire the post; we annotate it.
