How this site's figures are produced
The rate on this site is one figure, read from one place, and the rest is arithmetic. This page sets out where the close comes from, how the karat rows are derived, how the world gap is measured, and what is refused when a source looks wrong.
Why the 24 karat tola close is the only market figure
The Karachi Saraf Jewellers Association fixes one price: one tola of 24 karat gold. That is the number the market observes, and it is republished by gold.pk with the association named as its source. Every other karat on this site is that close multiplied by the karat over twenty-four, the same arithmetic the sarafa itself publishes. The 22, 21 and 18 karat rows therefore carry no independent market information; they are a convenience for a buyer who wants to know what a piece should weigh at that purity.
Because the karat rows are derived, the ladder can never disagree with itself. If the 24 karat tola close rises, every lower karat rises in the same proportion. A copy site that prints a different 22 karat figure is either using a different close or a different tola weight; it is not observing a separate market. The only number a reader needs to verify is the 24 karat close.
Where the close is read and how a frozen source is caught
The close is read from gold.pk, which names the Karachi Saraf Jewellers Association as its source. A reading is refused if the tola close falls outside a plausible band, if the page's headline and its closing table disagree, if the gap to the world metal leaves a band of minus four to plus eight per cent, or if the latest close is more than three days old. When a reading is refused, the previous one stays on the page, so the site never shows a figure it cannot stand behind.
A frozen source is caught by the age check: if the latest close is more than three days old, the site does not treat it as today's rate. The association does not fix on Sundays, so a Monday morning check will often see a close from Saturday; that is not a fault. The refusal rules are deliberately conservative: a figure that fails any of them is not shown.
Where the world metal and the rupee come from
The world price of gold is read in US dollars per troy ounce from one source, and the rupee from another. The two are multiplied to give a world price in rupees for the same weight as the sarafa's tola. Neither figure is taken from the page that carries the sarafa rate, so a problem on one source cannot silently infect the other.
The gap shown on this site is the difference between the Karachi close and that world figure. It is measured in rupees per tola, and it is a spread that exists for local reasons: import duties and restrictions, the open-market dollar, and dealer margin. The gap is set in Karachi, not in London, and it moves when the rupee or the world price moves faster than the sarafa re-fixes.
What this site refuses to print
Copy sites print a day high and day low and a table of city rates. On the source used here, those columns are a fixed offset from the close on every row: they are generated, not observed, and they give a false impression of a trading range. This site does not print them. A day high and day low that are the same offset from the close on every row carry no information about what actually traded.
No forecast is made. The site records the closes it has seen and shows them as they were published. A reader who wants a prediction must look elsewhere; this site's job is to show the last published close, the karat ladder derived from it, and the gap to the world metal at today's rupee.