Clustering & confluence

One window is a hint; many are a signal

A single cycle window says "a turn is somewhat more likely here." Useful, but modest. The method's strongest signals come from confluence — when several independent windows land on or near the same date. Each window is an independent estimate of when a turn is due; when independent estimates agree, the combined odds rise sharply.

Where confluence comes from

Confluence can stack from several sources:

• Multiple counts from one anchor. A Cycle 45 and a Fib window from the same pivot landing close together.
• Counts from different anchors. A Cycle 30 from a recent pivot overlapping a Cycle 90 from an older one.
• Time and price together. A window landing exactly where a major support or resistance level sits — the strongest setup of all.

Use the tool below to stack windows from two anchors. Drag them and watch the cluster meter: when projected dates bunch within a few days, the meter lights up. Those clusters are what the screener surfaces as its strongest windows.

Reading confluence without fooling yourself

Two cautions. First, near is not exact — windows rarely land on the identical day. A cluster spanning a few days is normal and still valuable; treat it as a zone, not a point. Second, more windows is not infinitely better. Three or four independent counts agreeing is meaningful; manufacturing "confluence" by counting every possible interval until something overlaps is just curve-fitting.

Confluence is where FIND delivers its best candidates. From here, Module 3 adds the Fibonacci and Date Echo tools that often participate in these clusters — and teaches you to read their statistics honestly.