The record view

The working, shown

The homepage draws its numbers; this page is where they show their working. Nothing was cut to make the board visual. It moved here, and it is the same build, from the same data, dated 25 Jul 2026.

How sure we are

The confidence ladder, and the fourth number

the full method →

A 27 read off a tail record and a 27 read off a press release are different claims about the world. Four tiers keep them apart, every answer names its tier, and the forward-looking number lives behind a fence of its own.

ATail-verified

The aircraft on your flight is resolved to a registration, and that registration to an install record. A number for one flight, with the sample size attached.

United

BType-derived

The tails are verified the same way and nobody publishes which aircraft is scheduled onto which flight, so there is no history to count. A number for an aircraft type. Not for a departure.

Alaska

CFleet-share

No per-tail verification exists, so the input is what the airline itself said publicly about how many aircraft are equipped. A number for an airline. Nothing narrower.

airBaltic, JSX, ZIPAIR, Air France, WestJet, Hawaiian, Qatar Airways, jetBlue, SAS, American, Delta, Virgin Atlantic, Emirates, Air Canada, Southwest, British Airways

DAnnounced only

A signed deal with nothing in the air. Today’s odds are zero and the projected figure carries the forward view, fenced, in grey. A promise with a year on it. Never a measurement.

jetBlue, American, Delta, Southwest

38begins 2028FIRM

Where a carrier has signed and published an aircraft count, a projected score appears in that grey chip and nowhere else. It never sorts a table. It carries its horizon date and a confidence word everywhere it goes, and when a horizon passes with nothing installed it flips to SLIPPED while keeping the date it missed. That flip is computed from the build date, so it does not wait for anyone to notice. The arithmetic and all five rules →

Three limits hold everywhere. Tail swaps happen up to pushback, which is why the T-48h re-check is in every playbook. This site tracks aircraft, never seats. And nobody has load-tested a full cabin, so every crowding claim, from anyone, is inference — I think that is the largest open question in the subject.

Measured · the load-test gap holds across Jang et al., ACM IMC ’25, Oct 2025 · Ullah et al., arXiv:2508.09839, Aug 2025 · Ookla Speedtest Intelligence, 28 Apr 2026

The loop

What the record cannot see, you can

Fleet records stop at the cabin door. Whether the login worked, what a video call survived, which badge was wrong: the person in the seat is the only instrument on board, and this site runs no analytics to guess with. Five channels carry observations back, and none of them asks for your email.

The report form: At the foot of every airline page, posting to the same intake the extension uses. A flight, a date, what the WiFi did. Plain feedback about the site or the extension rides the same channel.
The extension itself: The panel carries a report action, and the design calls for one accuracy prompt after a few real uses. Never on install day, never twice.
A store review: Chrome ranks listings partly on reviews, so a review is the one act that puts the extension in front of the next traveller. If it earned one, two sentences there carry more weight than anything I could write about it myself.
GitHub issues: Bugs, wrong badges with a screenshot, feature requests. Slower than the form, better for anything with a repro.
Reddit: Field reports keep surfacing in airline and Starlink threads. I read them, and a confirmed one enters the record wearing its REPORTED label like everything else.

The intake · POST /api/report · no channel attaches an account · what the server keeps is on the privacy page