The indicator refuses to count waves — and that refusal is the product.

What a count is

A wave label is a claim about position within a completed structure.

To say a market is in wave three is to assert that waves one and two have finished, that the structure now unfolding is impulsive rather than corrective, and that a specific remaining path is implied. Each of those is a judgment, and each can be revised the moment the next swing prints. Two competent analysts routinely produce different counts from the same chart, and both counts are internally consistent. That is not a failure of skill — it is a property of a framework whose labels are assigned in hindsight and defended forward. A tool that automated the labelling would automate the disagreement along with it.

What survives

Direction and character — nothing else clears the filter.

The reduction keeps two bits of information per timeframe: which way the structure is pointing, and whether the current leg is trending or counter-trending. That produces the five-state scale — impulse up, corrective up, neutral, corrective down, impulse down. Everything the count would have added is dropped: the wave number, the degree, the projected target, the alternate scenario. What remains is a classification rather than an interpretation, and a classification has a property the count does not have: two timeframes can be compared without anyone deciding whether they are at the same degree.

Why the label blocks comparison

Wave three on the hourly and wave three on the five-minute are not the same object.

The obstacle to multi-timeframe wave analysis has never been computing the counts — it is that counts at different degrees do not commensurate. A five-minute impulse nests inside an hourly correction without contradiction; both labels are correct and they point opposite ways. Any attempt to sum or compare them has to first resolve the degree relationship, which is itself a judgment call and the single most contested part of applying the theory. Phase classification sidesteps the whole problem by refusing to assert degree. Two rows both reading impulse up are making the same claim at their own scale, and that claim can simply be counted.

The neutral state earns its place

A timeframe allowed to say nothing is what keeps the counts honest.

The scale is symmetric around a genuine zero, and the zero is not a rounding artefact. When a timeframe's structure is mixed, the row reports neutral rather than being forced onto the nearest side. This matters because the convergence figure is a count of agreement, and a scale without an abstention would inflate every count by whatever share of rows were merely ambiguous. Neutral rows contribute to no side. The practical effect is that convergence readings fall during transitions rather than flickering between opposing highs — the tool goes quiet exactly when the structure is genuinely undecided.

What this costs

The tool cannot tell you where a move ends, and it never claimed to.

Discarding the count discards the projection. There is no target, no invalidation level, no statement about how much of the structure remains — because all of those are derived from the wave label, and the label was never computed. This is the honest price of the reduction and it is worth stating plainly, because operators who arrive expecting a forecast will read one into the output. The convergence read grades the structural environment. Locating the actual price zone is a separate operation with separate machinery, and running the two together is where the tool starts producing signals with no location attached.

• No target, no invalidation, no remaining-path estimate — none of these exist in the output.
• A high convergence reading describes the environment, not the entry.
• The read is a quality filter on a setup the operator already found.

The key idea

The tool computes less than the theory so that its output can be counted.

There is a general pattern here worth carrying beyond this indicator. A framework rich enough to describe anything produces outputs too rich to aggregate, and aggregation is what turns observation into evidence. The convergence tool makes the trade deliberately: it accepts a much poorer description of each timeframe in exchange for a description that is identical in kind across all of them. Everything downstream — the thresholds, the recorded grade, the eventual question of whether convergence actually predicted anything — depends on that uniformity, and none of it would be available from a chart full of wave labels.