What markets did on days like this

Nobody who tells you a conjunction crashes the market ever tells you how many times it did not. This is the count.

Explorer takes decades of market history and sorts it into groups. You choose what to sort by — which planetary event happened, where a planet sat, what a KP chart said, or what the panchang called that day — and every trading day that matches gets gathered up.

Then you see how each group behaved: how the days split between up and down, how many opened away from the previous close, what the average return was. Open a group and you can read the actual dates inside it. A tally of what already happened, never a forecast of what comes next.

Say a conjunction landed on a day the market dropped hard, and you want to know whether that means anything. This is the only thing that can tell you whether that is what usually happens, or whether you found one day out of four thousand.

Most of what comes back is unremarkable, and that is the point. A screen full of near-even splits has told you something true and saved you a strategy. We’d rather show you that than dress it up.

What you are hunting for is one group that behaves differently from its neighbours, on enough days to be worth taking seriously.

Explorer has four tabs, and they all work the same way — the form, then the bars, then the drill-down you just saw. The only thing that changes is what puts a day into a group. So once you can read one of them, you can read all of them. Two of the four also carry a Scanner, which gets its own page here.

• Planetary Events— Conjunctions, ingresses, retrograde turns — count what markets did on the day each one landed.
• Daily Planetary Position— Group every trading day by where one planet sat that day.
• Daily Position Scanner— Search every pair of planet positions at once, then see whether any of them hold up on years the search never saw.
• KP Position— Group days by a Krishnamurti Paddhati position — a planet's or a house's sub lord.
• KP Position Scanner— Search thousands of KP pairs at once, then see whether any of them hold up on years the search never saw.
• Panchang (Almanac)— Group days by what the almanac called them — tithi, paksha, yoga.

The Daily and KP tabs each carry a second view beside Breakdown, called Heatmap. Breakdown groups days by one position. Heatmap crosses that against a second one and asks the sharper question: do these two skew together, beyond what either explains on its own?

Every cell is one combination of Axis A and Axis B. The number printed on it is the plain win rate and day count for those days. The colour is something else entirely — it shows the residual (how far the cell sits from the independence expectation: what those days should have done if crossing the two axes added nothing beyond what each one already tells you alone). So a pairing that only looks interesting because one of its two axes is interesting comes out neutral here, which is the entire reason to cross them.

The colour scale is fixed rather than stretched to each grid’s own range, so a flat, unremarkable grid stays a flat, unremarkable colour instead of being auto-scaled into looking dramatic. Cells drawn dimmed and hatched sit below the trust floor: too few matched days, or too few separate stretches of them. They are never hidden, and the Min days and Min spans controls beside the legend let you move that floor yourself and watch which cells were only just clearing it.

Picking one pair by eye still says nothing about the thousands of pairs you did not check. That is what each tab’s Scanner is for — it searches every valid pair at once and tests whether the survivors hold up on years the search never saw.