How this works
Everything here is deliberately narrow: a fixed set of pre-tested templates, a fixed historical dataset, and a strict split between the period used to build a template and the period whose results you actually see.
The template library
Every query maps onto one of five fixed analysis templates — there is no general-purpose backtesting engine here, and nothing is generated on the fly. Each template is a precise, documented definition (what counts as a "touch," what counts as a "bounce" vs. a "break," exactly how a continuation target is set) written and unit-tested before it's ever queryable, then checked against real historical examples by hand before going live.
Opening Range Breakout
The opening range is the high/low of the first N minutes from 09:30 ET. A breakout is the first bar whose close moves beyond that range; the outcome is continuation (reaches one range-width further), failure (reverts through the opposite side), or unresolved (session ends first).
Gap Behavior
A gap is the percent difference between today's open and the prior session's close. A qualifying gap either fills (price trades back to the prior close during the session) or doesn't.
Support/Resistance Bounce
Levels are either the highest high / lowest low of the prior N trading days, or the nearest round number to current price. A touch is classified as a bounce (reverses away) or a break (continues through), decided within a fixed time window and point threshold.
Prior Day High/Low Breakout
Same breakout/continuation/failure logic as the opening range template, but the level is the prior trading day's high/low rather than an opening range.
VWAP Reversion/Breakout
Levels are the session VWAP plus/minus a volume-weighted standard deviation band, recomputed fresh every session. Touches are classified as reversion back toward VWAP or continuation through the band.
In-sample vs. out-of-sample
The full history is split into two fixed periods. The in-sample period is what gets used internally while building and testing a template — it never appears in any result you see. The out-of-sample period is the only period ever reported: every query you run, regardless of what date range you ask for, is answered only from data in that out-of-sample window, and the exact window is always shown alongside the result.
This exists to keep a basic honesty guarantee: a template's logic isn't shaped by looking at the same data its results are reported from.
How credit pricing works
The historical data behind every query is pulled once, in bulk, and stored locally — a query never triggers a live data-provider call. That means the real marginal cost of running a query is just compute (a few years of 1-minute bars processed once) and hosting, which is small and roughly the same across every template. Credit packs are priced with a margin above that cost, with a modest discount on larger packs — see pricing for the current packs.
This tool provides historical statistical information only. It is not investment advice, and past performance does not indicate future results. Trading involves risk of loss.