How a figure is verified before it is published on this site
Verification here is mechanical rather than editorial. A number does not appear because it is plausible, widely repeated or attributed to somebody trustworthy. It appears when four questions can be answered about it, in writing, on the page: who published it, what exactly it measures, for what period, and how a reader reaches the document. Anything that fails one of the four is left out, and several pages on this site end in a band or an absence as a result.
Question one: who published it
A named organisation, not a construction. "Reports suggest", "according to sources" and "widely believed" all fail at this step, as does an aggregator that does not say where its own number came from.
The publisher also has to be in a position to know. A tour publishing its own prize money is in that position. A newspaper reporting a private contract is relaying, and the page says so, which is a lower tier, not a disqualification. The register map sets out which bodies are in a position to know, class by class.
Question two: what exactly does it measure
This is where most published figures fail, and they fail by mislabelling rather than by invention. Prize money is not salary. Revenue is not net assets. A purse paid to a team is not driver income. A transfer fee is not a wage. An investment ask is not a valuation.
The rule applied here is that the label travels with the number, always, including in headings and in the plate at the top of each page. A figure without its label is a free-floating large number that can be presented as almost anything. Bracketed filings are the clearest case: what an asset range establishes is narrower than it looks.
| Presented as | Actually | Consequence |
|---|---|---|
| A player’s earnings | Prize money paid by one tour | Excludes most of career income |
| A driver’s income | Race purse paid to the entrant | Off by the driver’s private share |
| A company’s worth | Revenue for a period | Different measure entirely |
| A valuation | An investment ask on a broadcast | No transaction occurred |
| A wage | A transfer fee paid between clubs | The player received none of it |
Question three: for what period
A career total, a season, a single event, or a balance-sheet date. Without a period a figure cannot be compared with anything, and comparison is most of what a reader wants.
Periods also decide whether a figure is stable. A closed career total is permanent; an active competitor’s total is a running balance and will be wrong on a static page within weeks. Pages here print closed totals and anchor live ones on a fixed, dated payment.
Question four: can a reader reach it
The page has to say where the document is: the register and the section within it, the filing type and the financial year, the commission and the event date. A citation that cannot be followed is decoration.
This is also the step that keeps the site honest about its own limits. Some registers cannot be read automatically: one major tour blocks machine access to its player pages, and where that is the case the page says which body publishes the figure and how a human reader reaches it, rather than pretending to a level of access it does not have.
The two errors this test is designed to catch
Attribution: money paid to an organisation reported as income to a person. A transfer fee goes to a club, a race purse to an entrant, revenue to a company. Each of those appears in ordinary coverage as personal earnings, and each can be wrong by any factor.
Gross for net: a published figure treated as what somebody kept. Every earnings figure on this site is before tax, agent commission and professional costs, and in some sports before an enormous expense base.
What happens when two documents disagree
Both are printed, with the publisher of each named. There are cases on this site where a tour and a hall of fame record slightly different career statistics for the same player, and cases where two contemporaneous newspaper reports describe one contract at different values.
Choosing between them (usually by taking the rounder number, or the one that appears more often) manufactures a consensus that does not exist. Showing the disagreement leaves the reader better informed and the record intact. An estimate fails this test before it starts: there is no document for a second one to disagree with.
Questions readers send
What are the four verification questions?
Who published it; what exactly it measures; for what period; and how a reader can reach the document. A figure that fails any one of them is not printed.
Why does “not published” count as a result?
Because it is usually a fact about the world rather than a gap in research. For most people, in most sports, in most eras, nobody published what they were paid, and saying so leaves a reader better informed than an invented figure would.
What if a register cannot be read automatically?
The page names the body that publishes the figure and how a human reader reaches it, rather than claiming access it does not have.
Does a report in a newspaper clear the test?
Not on its own. A contemporaneous report is evidence that a figure was announced, and where that is all that exists the page says so in those words. What it cannot do is stand in the place of the document: an outlet that prints a number without naming where it came from has moved the question rather than answered it, however reputable the outlet is.
The documents this page rests on
- 01 Tours, leagues, commissions and public authorities, That every printed figure has a publisher, a defined measure, a period and a route Each page carries its own register block listing exactly those four things