The five questions of the Saturday pass, asked in sequence.

The sequence

Five questions, each with its own instruments.

Are entries precise — answered by average MAE in R, adverse utilisation, MAE quality, and the stop efficiency score. Are setups creating enough opportunity — answered by average MFE, MFE quality, the MFE distribution, and the MFE/MAE ratio. Are exits converting that opportunity — answered by capture efficiency, giveback, and the efficient-exit quadrants. Are fees damaging EV — answered by fee R drag, fees as a share of balance, swap counts and swap fees. And which branch or timeframe is leaking edge — answered by the branch and timeframe analytics read against the trade log. Five questions, five instrument sets, one order.

Why entries come first

A capture figure means nothing until you know what there was to capture.

Capture efficiency is realised outcome divided by MFE, which makes it entirely dependent on the denominator. If setups are not generating opportunity — low MFE across the week — then capture will read impressively while the trades were never worth much, and an operator reading exits first will conclude their exit discipline is excellent. The reverse case is equally misleading: strong opportunity generation with poor capture reads as an exit problem, which it may be, or may be a stop that was never survivable enough for the runner to develop. Establishing entry quality and opportunity generation first is what makes the exit reading interpretable, because it fixes both ends of the ratio before the ratio is used.

• Every loser scores zero capture and every tiny-MFE trade scores wildly — properties of the formula, not the trading.
• High MAE with high MFE says the thesis was right and the entry was early; high MAE with low MFE indicts setup quality.
• The MFE/MAE ratio is the bridge between questions one and two, and worth reading before either verdict is fixed.

Why fees come fourth and not last

Friction is a real EV term, and it is invisible in R.

Fee R drag divides total fees by initial risk in dollars, converting friction into the same unit as everything else, and net outcome R subtracts it from the raw result. This placement matters: fees are asked about after the execution questions but before the location question, because friction can masquerade as an execution problem. A branch showing consistently disappointing net results with clean MAE, MFE, and capture is usually not executing badly — it is being eaten by commission, spread, or swap, and no amount of examining its exits will reveal that. Overflow is the branch where this appears most often, since supplemental flow carries volume and cost characteristics the core plan does not.

Why location comes last

The fifth question is where, and it only makes sense once you know what.

Branch and timeframe analytics answer where a problem lives, and that question is only tractable once the first four have established what the problem is. Opening the branch analytics first produces a familiar and unproductive result: several branches look somewhat worse than others on several metrics, with no basis for deciding which difference matters. Arriving at the branch view already knowing that the week's issue is giveback on high-MFE trades turns it into a targeted lookup — which branch is producing them — and that is a question with an answer. The manual's own warning applies here: the All Blended view can hide branch-specific problems and should be read only after the branches.

The key idea

The order is the analysis; the metrics are just the instruments.

All five questions can be answered from the same tab in any sequence, and the sequence is what determines whether the answers compose into a diagnosis or accumulate into a list of observations. Following the trade's own life cycle — entry, opportunity, exit, friction, then location — means each reading arrives with the context that makes it mean something.