Article library
The article library
Free essays, no paywall and no signup. Every piece takes on a single way that machine-learning evaluation breaks, and walks through it end to end.
- Essay 01 · October 2026 Why backtests flatter A backtest is an experiment on the one history we have. Five habits — from lookahead to frictionless fiction — bend that experiment toward the answer you were hoping for. Read →
- Essay 02 · October 2026 Data leakage, catalogued Leakage is any path by which information from outside the training sample — usually the future, often the label — enters model construction. Six named routes, each closable. Read →
- Essay 03 · October 2026 Point-in-time data The dataset you load today is not the dataset a trader saw then. Delistings, restatements, reconstituted indices and recycled tickers all rewrite the past — point-in-time discipline is how the rewrite is undone. Read →
Questions people ask
Is any of this investment advice?
No — and it is never personalised. The essays are editorial research about how machine learning gets evaluated in investment. They discuss failure modes in general terms and do not recommend any security, fund, broker, or parameter configuration. Decisions in markets stay with you.
Do you sell anything here?
No. There is no subscription, no consulting tier, no managed product, and no advertising on the site. The library is the entire offer, and it is free. Should anything paid ever appear on this website, these pages and every legal notice will say so before it goes live.
Can I suggest a topic or report an error?
Yes, please. The inquiry form on the contact page goes straight to the editorial desk, and corrections are treated as a normal part of publishing: each essay carries its first-published label, and substantive corrections are acknowledged in the text rather than quietly overwritten.
Do you track readers?
No analytics, no advertising pixels, and no third-party trackers run on this website. The only item kept in your browser is your own cookie notice choice, described in the cookie notice.