◆ PARALLAX

Phil, park that unless you can state a testable prediction for this log set.

A claim we can use would be something like frequency, time window, exposed group, unexposed group, and expected difference. If it cannot enter the table, it does not drive the thread.

Denominator update: I removed the three nights where the receiver was down or the tape ran out. That leaves 41 observed windows, 9 flagged as coupling candidates.

If we include only nights with confirmed local logging, it drops to 34 windows, 8 candidates. The ratio is still not proof, but it is less fragile than the first pass.

A usable prediction has to be stated before the window opens.

Example: between 0200 and 0430 MST, if the heater pattern is present and geomagnetic noise is below our cutoff, then the 7.3 MHz scatter trace should widen by at least 20 percent against the prior two-night median. Anything written after the fact goes in the notes, not the score column.

What are we using for controls? I want matched nights with similar solar flux, same receiver chain, same antenna orientation, and no known heater schedule.

Also one sham window chosen by someone who does not see the logs until after timestamps are locked.

You are all ignoring the delivery vector. If the same contractors handle water treatment, ionospheric heating, and municipal emergency radios, then the coupling is not just RF.

The nervous system is the receiver.

Phil, park the water theory. This thread is only for measurable RF and sky-condition claims.

If you have timestamped receiver data, post the counts and method. If not, do not dilute the denominator Mags just cleaned up.

Seconding K7RADIO. The strongest version of this case is narrow: predicted windows, locked logs, comparable controls.

Broad contractor webs make the claim impossible to falsify, which makes it useless for our purposes here.

Safe logging note: keep raw receiver times, local clock-drift checks, weather notes, and edits in separate sheets. Initial entries should be ink or write-once media, then mirrored.

Do not put names, addresses, or station locations in the public digest. Grid square is enough unless the operator volunteers more.