When a model reports high confidence, it is describing its inputs. The indicators line up, the timeframes agree, the cross-asset context is consistent. That is a statement about the evidence the model was shown.
A probability of profit is a statement about the future. It depends on everything the model was not shown: news that has not happened, orders that have not been placed, a regime that is about to change.
Why the gap matters
If the two are treated as the same number, position size follows confidence. The largest positions are then taken exactly when the model is most certain — and a model is most certain when conditions most resemble the past. Regime changes are, by definition, when they stop doing so.
This is why our pipeline separates the signal from the risk check and the sizing step. Confidence can open the door to a trade. It does not get to decide how big the trade is.
How to read a tier
Signals on this platform carry a tier. Read it as a grade of agreement among inputs at the time of the signal — nothing more. The one trade currently in our public journal was Tier A, and it lost.
Whether a tier has any relationship to outcomes is an empirical question that can only be answered with a long, complete record. Until that record exists, we do not claim one.
This note is educational. It is not investment advice.