Judging a Filter by the Trades It Removed

Every filter pays for itself in the same currency: the trades it prevents. Some of those were losses and some were the best results of the quarter, and the net figure is the only price worth quoting. The intervals orb trading review consultoriainnova uses for this are longer than a single quarter, since a strategy for the opening range does not produce enough excluded trades in twelve weeks to say anything, and the improvement in the remaining sample is not evidence on its own.

The Remaining Sample Always Looks Better

Removing any subset of trades that contains more losers than winners raises the average of what is left. That happens with a genuinely useful filter and with a random one applied to a small sample. Reading the improvement in the survivors as proof the filter works is the single most common error in this kind of review, and it survives because the number moves in the right direction.

Keep the Exclusion Log

Each session, the entries that met every other condition and were blocked by the filter get recorded with their trigger, stop loss and what they would have returned at the planned exit. That list is the actual subject of the evaluation. Without it the filter cannot be assessed at all, only observed alongside a result it may have had nothing to do with.

Read the Pattern, Then the Total

Sort the excluded trades by outcome. A filter doing real work removes a recognisable group: the narrow ranges, the low relative volume mornings, the sessions where price was already extended before the opening bell. A filter doing nothing removes an assortment that looks like the rest of the sample with fewer entries. The composition tells you more than the sum.

Count the Best Trades It Blocked

Most systems earn a large share of their result from a small number of outsized sessions. A filter that happens to exclude two of those in a year can look sound on every average and still be expensive. Checking specifically whether the exclusion log contains any of the period's largest winners is a short, direct test that averages will never show.

Set the Bar Before Looking

Decide in advance what the excluded set has to look like for the filter to stay, in the same way review criteria are fixed before a quarter starts. A total that is negative by some stated margin, or a clear concentration in one kind of session. Choosing the bar after reading the log turns the whole exercise back into a justification.