Data and method

Where the figures come from

Every value is transcribed from the Visa Bulletin published by U.S. Department of State, Bureau of Consular Affairs. Each issue on this site records the URL it was read from, and every table links back to it. Nothing is taken from a mirror, an aggregator, or from memory, and the build rejects any source that is not a travel.state.gov address.

What the build refuses to publish

• An issue missing any category or any chargeability area, on either chart.
• A gap in the monthly series, which would make a movement chart overstate the pace.
• A cell whose state and date disagree, such as a current cell carrying a date.
• Data whose newest issue is more than 75 days old.

Retrogressions are reported as warnings rather than errors. They are real events, but they are rare enough that flagging them catches a transcription slip before it is published as news.

How movement is calculated

Movement is the difference between two published cut-offs, in days, shown in months at one decimal place. Pace divides that by the number of months between the two issues. A pace below one means the cut-off advanced less than a month for each month that passed, so the queue lengthened in real terms.

Every headline pace figure is measured over the trailing twelve months, and says so. The bulletin runs on an annual quota that resets in October, so a shorter window samples one phase of that cycle and misreads it, while an endpoint measured across the whole archive hides the months a cut-off moved backwards. Those months are counted separately and reported alongside the figure.

A category that is current or unavailable has no cut-off, so it has no movement. Those months are shown as state changes instead of being silently treated as zero.

Projection, and how wrong it is

The history pages offer a projection: given the gap between a priority date and the current cut-off, and a pace expressed as months of cut-off per month of real time, it divides one by the other. The pace is chosen by the reader. The site suggests the category and country’s own figure for the last twelve months, but it does not claim that pace will hold, and the result is not a probability. There is no distribution behind it and no model, only division.

That method has been checked against this site’s own record. Standing at every month with a year of history behind it, projecting the cut-off 12 months forward at the pace prevailing then, and comparing against what was actually published, gives 505 projections across every category and country. The median miss is 6.3 months, 51% were out by more than six months, and the middle 80% landed between -18.5 and +10.9 months of the truth. That is over a one-year horizon; most real gaps run for years, where the error compounds.

Those figures are recomputed from the data on every build and shown beside every projection, because a projected date without its track record is the most harmful thing this site could publish.

What is still deliberately absent

The site never asserts a date on its own authority, never ranks a reader’s chances, and never suggests what anyone should do about what it shows. The factors that actually move a cut-off, demand, per-country limits, and administrative capacity, are not published in the bulletin and are not modelled here.