Skip to main content
← Following

Getting The Data

The claim that foreign-registered aircraft shadowed Charlie Kirk, Erika Kirk and TPUSA rests almost entirely on third-hand readings of flight-tracking websites. Somebody photographed an aircraft-history table. Somebody typed the rows into a spreadsheet. Somebody tallied the spreadsheet. Somebody read the tally on television. Every one of those steps is a place a row could be misread — and the published audits say a great many of them were.

This section replaces that chain with data we pull ourselves, from named sources, with the raw response saved and timestamped. It exists so that a reader who does not believe a word of this site can repeat the pulls and get the same answer.

The four passes​

1 — Public open source Free networks, no account, no key. Repeatable by anyone. Always runs first.

3 — Government records Registries, facility data, FOIA and state public-records law. Records, not readings.

2 — Commercial APIs The deep history the free networks do not hold. Costs money; terms restrict republication.

4 — Website capture Driving a real browser at the tracking sites. How we document that data is gone.

A pass is never a substitute for another. The same claim checked four ways and agreeing is the strongest thing this investigation can produce. Checked four ways and disagreeing is the second strongest — and it gets published too.

What the first pass found​

478 pulls against the free historical archive on 24 August 2026. 178 traces retrieved. Three results worth stating up front.

The September 10 departure has a primary answer. SU-BTT was on the ground at Provo Municipal at 13:07 UTC — 07:07 Mountain — lifted off at 13:14 UTC, and was on the ground at Wilmington New Castle, Delaware at 16:50 UTC. That is a departure about five hours before the assassination and a landing about ninety minutes before it, from a 1,229-point track. It matches the "7:08 AM MT" figure that has circulated and is inconsistent with the competing accounts. It says where an airframe was. It says nothing about who was aboard or who tasked it, and no public record establishes tasking in either direction.

An independent check agrees with the audit. Running every exactly-claimed date through primary traces, blind: pairings this repo's audit called accurate have a trace 75% of the time, ones called partial 88%, and ones called inaccurate only 11%.

And the sharpest result splits the claim in two. Individual flight legs — "this aircraft was at this field on this date" — verify at 98%. Overlap pairings — "and therefore it was following them" — verify at 38%. The aircraft movements are largely real. The inference that they constitute shadowing is where the spreadsheet falls apart. That finding flatters neither side of this argument, which is roughly the point.

What none of it can do​

The counts remain trackers' tallies, not records. 73, 72, 70+, 68, 29, 23 — amateur readings of public data by independent researchers, never reconciled with each other. Nothing in this section turns one of them into a verified total, and where they conflict this site shows the conflict rather than averaging it.

ADS-B has holes, and they cut both ways. A missing trace means a volunteer network heard nothing. It does not establish that an aircraft did not fly and it does not establish that a transponder was switched off — a parked aircraft, a coverage gap over the Atlantic, and a wrong claimed date all look identical. Every gap we hit is written down in a MISSING_DATA.md file beside the relevant aircraft and overlap, 90 of them, saying exactly that.

One of the two shut doors has since been forced. A later pass — run on the assumption that the data had been erased, and looking for backups — found that 2022 is partly reachable after all, that days the primary archive refuses are published in full elsewhere, and that the removed FlightRadar24 history for N102DZ survives in the Internet Archive. That work is Flight Data Recovery, and it corrects several statements made in this section.

The other door is still shut, and no backup opens it. Erika Kirk's side cannot be checked at all. Every recovery in that pass is about aircraft, and an aircraft track never establishes who was aboard. This repo holds a single sourced location for her before 10 September 2025, against a window somebody counted 73 overlaps across. See Erika Kirk's flight logs reported erased.

What this section will not do​

It will not merge the threads. The 18-month following pattern, the September 10 day-of timeline, the N1098L thread and the N888KG Wendover departure are four different claims about four different sets of aircraft, and blending them is how the whole angle gets dismissed at once.

It will not name who tasked an aircraft, because no public record establishes it. It will not name a living person as an accused. And it will not publish a page without the innocent explanation beside the suspicious one — including the innocent explanation for our own results, which is that an aircraft simply was not there.