Preparation happens before the market opens, not during it. The point is that by the time price starts moving, every level that matters is already drawn and already named. Nothing gets measured in the middle of a session.

Everything is done on the RTH chart — the regular trading hours chart, which shows only New York session data. The 24-hour electronic chart is deliberately not used for analysis, because overnight data changes which gaps look already filled and which levels look already taken.

The instrument is NQ. One market, one chart layout, one prep routine, run the same way every day.

Context is measured with a Fibonacci retracement, used for one thing: finding the 50% of a range. The previous week is measured low to high, and the previous day is measured low to high. The 50% of each is its equilibrium.

Above equilibrium is premium. Below equilibrium is discount. That is the whole vocabulary — the framework is not reading the Fibonacci levels as targets, it is using the midpoint to answer a location question.

Because two ranges are measured, price can be in premium of both, discount of both, or premium of one and discount of the other. The framework names each of those states and treats them differently.

Imbalances are marked in the same pass. A gap is an unfilled area left behind when price moved away quickly. Buy gaps and sell gaps are marked on the higher timeframes, and where two gaps from different timeframes overlap, the framework treats the overlap as the area of interest rather than either gap alone.

Bias in this framework comes from liquidity, and liquidity means resting orders sitting above old highs and below old lows. Highs are buy-side liquidity; lows are sell-side liquidity. Once one side has been taken, the framework's expectation is that attention turns to the other.

Where those levels are drawn is the part that makes the framework distinctive. Levels are taken from candle bodies, not wicks. The reasoning James teaches is that wicks are where price fluctuated and bodies are where the volume actually settled — so the highest body in a session is the level that matters, not the highest print.

Liquidity is scoped by session. The New York day is split into named intra-sessions, each of which produces its own high, its own low, and its own body-based extremes.

The framework's hardest constraint sits here: if no liquidity has been taken, there is no trade. A setup that looks correct in every other respect is not taken without it.

Execution is where the framework gets narrow. Having the right context and the right bias is not permission to enter — a named pattern has to appear, in a named sequence, at a permitted time.

The pattern is the One-For-Two signature, and it is worth being precise about what that name means. One For Two describes a three-candle shape, not a risk-to-reward ratio. A bullish signature is a low, then a lower low, then a higher low whose body closes back through the bodies of the first two candles. A bearish signature is the mirror of that. A candle that makes neither a new high nor a new low is skipped, which is why the pattern is sometimes four bars on the chart, and the framework caps it there.

The signature alone is not the entry. The sequence the framework requires is that a signature forms, that signature's own extreme gets run, and then a new signature in the same direction appears. That second signature is the trigger. Entering on the first one is the specific error the framework spends the most time warning about.

Timeframes are used as a hierarchy rather than as alternatives. The 15-minute is the bell weather chart, where the levels come from. The 5-minute is where the liquidity grab is confirmed. The 1-minute is where the entry is found — and the reason for dropping down is risk, because the same idea entered from a higher timeframe requires a far wider stop.

Time is a filter in its own right. The framework defines windows within every hour in which it will not take an entry, regardless of how good the setup looks, along with additional constraints around the lunch period and around scheduled news.

Risk in this framework is defined in advance and expressed in points, not as a feeling about how good a setup looks. The stop distance and the target are both fixed before entry, they are the same on every trade the model produces, and the stop is never widened once price is moving against it.

In the current version of the framework the trade is submitted with its stop and its target attached and then left alone. There is no moving to break-even, no adding to a loser, and no watching a position tick by tick looking for a reason to intervene.

Frequency is capped rather than left to discretion. The framework treats the session's first valid setup as the one to take and puts a limit on how many trades a day may produce — the constraint exists specifically so that a loss cannot be answered with another trade.

There is a distinction the framework insists on when a trade loses: whether it was a genuine loss, where the rules were followed and the market did something else, or an execution error, where a rule was broken. Only one of those is information about the model.

Nothing here is a claim about outcomes. A rule that caps risk limits how much a mistake costs. It does not make trades work.

Review is done on the chart, not from memory. The session is replayed bar by bar, the day's context is re-derived from scratch, and each decision is judged against the rules as they were written — not against how the trade turned out.

Every trade is graded on the genuine-loss-versus-execution-error distinction. So is every trade that was not taken, because a valid setup that was skipped is as much a rule failure as an invalid one that was entered.

When the model produces an unusual result, the first thing checked is the economic calendar. The framework treats a scheduled release as a common explanation for a day that did not behave, and it would rather find that than invent a new rule.

The framework is also reviewed against itself over time — which model the current market is paying attention to, and which is being ignored. That is tracked as an observation, not converted into a prediction.

The learning method the framework prescribes is forward testing: watching the market live, calling what the rules say, and not entering. James credits watching rather than trading as how he learned to read the market in the first place.

Backtesting is done the same way as review — day by day, marking each session's levels forward as it closes and rolling them into the next day, so the practice matches what live prep actually feels like.

The framework is deliberately kept small. The stated design principle is that it should not require micro-mechanics or sub-second judgement, because a rule you cannot execute under pressure is not a rule you have.

One more constraint belongs here: the framework does not try to predict the market beyond what its own model describes. Where the rules are silent, the answer is to stand down rather than to improvise.