ForexBestRobots V229 — Alligator H4 first pullback research kit

Educational historical simulation, not a broker statement or investment advice.
Original HistData ASCII M1 EURUSD archives are preserved with source hashes.
EST is fixed UTC−05:00. Identical duplicate minutes only are removed; conflicting
duplicates abort. Missing intervals are not filled. January 2026 permits exits
only. See fixed-protocol.json and data-cleaning-and-boundary-addendum.json.

Requires Python 3, numpy and pandas. From this extracted folder run:
python prepare_data.py
python run_research.py
python verify_research.py
python prepare_viewer.py

The exported CSVs, JSON and independent check appear in assets/ and research/.
The input-validation report details source coverage. Minute timestamps locate
observations in minute intervals, not exact ticks. Gaps can hide level crossings.
1R is initial Bid-to-stop distance. Summed R is not account percentage return.
Transaction and per-UTC-midnight fees are allowances, not measured broker costs.
No rule optimisation or post-hoc substitution of later pullbacks was performed.

Original source pages are listed in source/source-manifest.json.
Indicator definition: MetaTrader 4 official Alligator help.
Data specification: HistData official detailed specification.
Both independent calculations use the same provider, not an independent feed.
