•
Gould et al. (2013) survey limit order books: mechanics, stylised facts, and modelling strands. Plain-language map for after-work Python learners. Not investment advice.
Takeaway. Before you code order-book features, you need a shared map: how LOBs work, what "stylised facts" look like, and which model families people actually use.
Limit Order Books (Martin D. Gould and co-authors, survey ~2013, arXiv:1012.0349) is the open-PDF tour many after-work learners start with—definitions, empirical regularities, and pointers into the literature.
Fit for learners: Open-PDF survey of LOB mechanics, stylised facts, and models—map the vocabulary before coding book features.
This Hub note is a plain-language intro. Not investment advice. Pair later with Cont–de Larrard Markovian LOB and Avellaneda–Stoikov.
Blog posts and broker APIs throw around "L1/L2," "queue priority," and "spread" without a coherent picture. Jumping straight into feature engineering produces brittle notebooks and wrong mental models of latency, cancellations, and depth.
Read selectively—treat it as a syllabus, not a single-night cover-to-cover grind.
Almost every HFT/execution notebook eventually needs columns like best bid/ask, depth imbalance, or event counts. This PDF tells you what those names mean in research language.
Tiny evening exercise: sketch a dict-of-lists LOB (price → size), process 20 fake events (add/cancel/trade), and print top-of-book after each—pure Python, no exchange socket.
Landing: arXiv:1012.0349
Not investment advice. This Hub note is educational only. The paper is research—not a licence to trade, quote, or execute with real capital.
Single course CTA: SoloQuant Academy's Introduction to Python Programming—get comfortable with functions, files, and plots first; then these microstructure PDFs become much less scary.