15 out of 36

A total of 15 out of 36 falls below eighteen, the figure most matching tables print as a minimum. It falls short of it by 3 points. If the number arrived with a verdict attached, the verdict was somebody’s reading of it. The number by itself is a position, and the position can be stated exactly.

Scored over every pairing that can exist — 11,664 of them, the whole population rather than a sample — 489 land on exactly this total, or 4.2% of all pairings. 80.3% reach this score or better. The mean across the population is 20.1.

How far below the line this sits

Eighteen is a convention. It is widely printed and widely applied, and it is not the middle of the range. The mean above is higher than it, so the pass mark already sits below what two people chosen at random produce. A total in this band is a short way under a threshold that was set low to begin with.

That is a statement about the threshold, not about two people. Traditions that treat eighteen as a bar conclude that a total here does not clear it. Others treat the total as a summary to be opened up rather than a line to be crossed, and go to the eight kootas to see which points were lost. Both conclusions come from the same figure. Which of them was applied to your chart is worth knowing, and it is a fair thing to ask.

Which tests give up the points at this total

A score out of 36 is not one measurement. It is eight separate tests with very unequal weights, and several of them pay all or nothing. Totals in this part of the range are reached in a recognisable way. Among the pairings landing on exactly this score:

• Bhakoot scores zero for 84% of them. It measures the distance between the two moon signs, counted in both directions, and pays nothing at all for certain distances.
• Vashya scores zero for 81% of them. It sorts the twelve signs into five classes, and asks whether one partner sits in a class the other is said to hold sway over.
• Nadi scores zero for 58.7% of them. It sorts each birth star into one of three constitutional channels, and pays nothing when both fall in the same one.

Meanwhile 61.1% of pairings at this total take full marks on Varna. That is the ordinary shape here: the small tests are mostly satisfied, and one or two of the large all-or-nothing blocks are not. It matters which, because the large blocks are where a different convention or a corrected chart changes the answer outright rather than by a fraction.

The same total can be assembled from one large block lost or from several small ones, and those are not the same chart.

Where a total in this range can move

Two things behind the number are worth checking. Neither is a remedy.

Which tables produced it. Several kootas are scored by more than one documented convention — the direction the tara count is taken in, which graha maitri table is used, which varna table a source prints. Each is traceable to a named tradition. The same two charts scored under one set of tables and then under another can come out a point or two apart, and neither calculation is an error. Close to a threshold, that difference is the whole question.

What birth time the chart was cast from. The moon moves through a nakshatra pada quickly. A birth time rounded to the hour, or remembered rather than written down, can place the moon in the pada next door — and the pada is what the entire Ashtakoota calculation is built from.

Neither of these is a promise. Checking them may move the total and may leave it exactly where it is. The point is that both are answerable rather than matters of opinion.

What to do with the number

Ask which tradition the tables came from. Check both birth times against whatever they were originally recorded on, rather than against memory. Then read the eight kootas separately instead of the total, because the breakdown says where the points went and the total cannot. Seeing the same chart scored side by side under several documented traditions shows how much of a figure is the chart and how much is the convention.