A session framework describes how a trading day unfolds: where the opening range forms, whether price accepts or rejects it, and what that implies for the rest of the day. Traders have used such frameworks for decades. The difficulty is that they are usually taught through examples, and examples cannot show how often something happens.
DR-LENS is NOVAQ's answer. It takes the published defining-range method of TheMas7er and turns it into something that can be counted. It annotates about 1.1 million five-minute bars per instrument, across about 4,100 sessions of NQ and ES since 2010, with more than 400 columns covering session ranges, breaks, acceptance and rejection, multi-timeframe context and session statistics.
What it makes possible
A structural question can be answered with a sample and a confidence interval. The same engine serves three audiences. Researchers use it to test ideas across 16 years without new data work. The live platform shows session statistics during trading. AI agents receive those statistics as part of their market context, so a brief can cite a measured frequency in place of an impression.
Together with joint work by Vašek (BTE), this produced a detailed picture of how a session behaves.
From bars to counted session structure
The core structure is real
The method's main session frequencies replicate closely across independent implementations. A widely quoted frequency reproduced at 80.7% on ES across 2,148 sessions, within one point of the published figure. That agreement is valuable in itself, because it means two separate codebases read the market the same way.
Volatility carries, direction does not
Width passes from one session to the next, and a wide earlier session is followed by a wide opening hour far more often than chance. Candle colours behave differently: only 1 of 21 pairs differed from independence.
Position is one vote
This was the clearest result. Once price has shown where it stands within the session, earlier sessions, composite bias, regime labels and the order of extremes added no measurable information about direction, in leave-one-year-out tests. That simplifies analysis considerably. Many apparently separate signals are one signal seen from different sides, and a system that counts them separately is counting the same evidence several times.
Night and day differ
In an exploratory study against a random-path control, overnight weakness tended to reverse and daytime strength tended to continue, by a few percentage points. The overnight effect is too small to survive costs, though it is a useful description of how the two halves of the futures day differ.
Hidden labels
One author's split appeared to add about 20 points of edge. At each decision time, though, the probability stayed at its unconditional level, because the label is known only after the session ends. The same discipline that exposes this kind of effect also protects NOVAQ's own research from look-ahead.
Why it matters
Knowing which session facts carry information, and which only look as if they do, makes research cheaper and strategies simpler. It also gives traders and AI agents a shared vocabulary backed by counted history.
Limits
The night and day study was not pre-registered, and the asymmetry fits the rising market of recent years. High hit rates have not become profit by themselves: in the joint register, none of 44 strategies passed. The value of this work lies in understanding structure and in removing false signals.