The whole method, explained in plain English.
Every percentage on this site is worked out by a program, not picked by a person. This page explains how, in plain English. If you cannot check how a prediction was reached, it is not worth much.
We start with a player's date of birth, taken from public records. Wikipedia is the main source, and we store the exact page and revision each date came from. We never estimate a date of birth. If we cannot find a reliable one, that player is left out of the calculation rather than filled in with a guess.
From that date we work out where each planet was in the sky. The Sun, the Moon, Mercury, Venus, Mars, Jupiter and Saturn, plus the two points where the Moon's orbit crosses the Sun's path. Those two points are usually called the North Node and the South Node. They are not planets, but Indian astrology has always treated them as though they were, and they carry real weight in a reading.
The positions come from standard astronomical formulas, the same sort used by planetarium software. We then shift them onto the sidereal zodiac, which is the version of the zodiac tied to the actual constellations rather than to the seasons. That shift is currently about 24 degrees. Western astrology normally uses the seasonal version, so if you are used to Western star signs, the signs quoted here will often be one back from the one you expect.
Public records give a player's birth date but hardly ever the time of day. So every chart here is worked out for midday. The Moon moves about thirteen degrees in a day, which means its position could be out by up to about six degrees either way, and we cannot work out the rising sign at all. Any astrologer would call that a real limitation. We would rather say so than pretend to a precision we do not have.
We then work out the same positions for the day of the match. The score comes from comparing the two. Where the planets are now, measured against where they were when the player was born.
Eight rules run on every player. Each one either adds to or subtracts from the score, and each one writes its own explanation. Those explanations are what you see when you tap a player's name in a report.
Everyone starts on 50 and the rules move them up or down from there. Scores are capped at 5 and 95, because no chart makes anybody certain to succeed or certain to fail.
Team scores are weighted. A captain counts for more than a squad player, an all rounder counts for more than a specialist bowler, and anyone not in the eleven counts for very little. The difference between the two weighted figures is then converted into a percentage split, calibrated so that a ten point gap comes out at roughly 70 to 30. That calibration is what stops every match reading as a meaningless 50 to 50, or as an absurd 99 to 1.
It uses no cricket statistics at all. No batting averages, no bowling figures, no recent form, no pitch report, no head to head record, no team news, no conditions. It is birth charts against current planetary positions and nothing else. If you want analytical cricket forecasting, this is not it, and we would rather tell you that up front.
The program is repeatable. Feed it the same two dates and it gives the same answer every time. Nothing gets adjusted by hand for a particular match, and no published prediction gets quietly edited after a result comes in.