Learn / Dashboard
What should I explore?
Which planetary conditions are active today — and what they did before
So what are you looking at?
Pick a day. You get every planetary condition active on it — where each planet sat, which events landed, what the panchang called that day. And beside each one, what this market has actually done on days like it before: which way it leaned, how often it gapped, how choppy it got, and how much history that reading rests on.
Judge by decides which of those things the significance column actually tests — direction by default, or gap probability, 1-day return, 3-day return, volatility. The rows stay the same either way; only how notable each one reads changes.
Most people would call this a financial astrology dashboard. That is what it is. The difference is that nothing on it is a forecast. It’s a record of what already happened, lined up against a date you choose.
Why should you trust these numbers?
Two records go in, and nothing else. Where the planets were on any given day, computed rather than interpreted. And what the market did on those days, from its own price history. Every figure here is counting over those two.
Nothing is filtered to look good. Conditions with no lean sit in the list beside the ones with a strong lean, and the note above the table says how many conditions were tested and how far the strongest of them deviated — for whichever factor Judge by is currently set to. Across the full history, none of them clears the bar that testing this many conditions demands, on any of the six.
You also don’t have to take any row on trust. The arrow at the end of most rows drops you into the Explorer, on that exact condition, looking at the days behind it. Every number on this screen is one you can go and rebuild yourself.
Is this where you should start?
Probably, if any of this sounds like you. You’re new to financial astrology and don’t know where to look first. You’re curious whether the sky moves markets at all, and want a real answer before picking a side. Or you already have a view on a market and want one more record to weigh it against.
Every other screen here asks you to bring a question — a market, an event, a rule to test. This is the one that doesn’t. Open it, read what’s active, and let your question come from there.
Once you have one, take the longer route, because that is the one that settles things. Sky shows you what a planetary event actually looks like. Explorer counts what markets did on every other day like it. Backtest turns that into a rule and tells you whether you could have traded it. This screen is the doorway; those three are the method. And if you have arrived here from Backtest instead, the job is the same — this is where the next thing worth testing comes from.
How do you read a day?
- 1
Scrub to any day. Step back and forward with the arrows, or jump straight to a cell in the strip — 15 trading days either side of today. Today stays outlined even once you’ve scrubbed away from it.
- 2
Read this day's own numbers first. Open, high, low, close and the previous close, then how the day actually moved: return, gap and volatility. Each one is checkable against the OHLC sitting right beside it.
- 3
Read the note before the rows. How many conditions were tested for this market, and how far the strongest of them deviated. It also says what a Daily/KP combination row below is being ranked against.
- 4
Judge by decides what the Z column tests. Direction is the default — how often the market closed up. Switch to gap probability, 1-day or 3-day return, or volatility and every row’s Z recomputes against that instead. Nothing else on the row moves — same bias, same n, same everything else — only the significance reading changes.
- 5
Narrow with Show and Filters. Show separates fast-moving conditions from slow-moving ones. Filters applies the same minimum-days, minimum-deviation and minimum-recurrence floors the Explorer uses. Both are off when you arrive, so nothing starts hidden — switching Judge by can shrink what’s visible here, since the deviation filter re-applies to the newly-chosen factor.
- 6
Read one row. Bias, Z (how far this row sits from what you would otherwise expect, counted in standard steps — bigger means more unusual), gap probability, volatility, 1-day and 3-day return, sample size, spans (how many separate episodes those days come from, since one long unbroken run is a weaker sample than the same number of days spread across many occasions) and a quality badge. Every column carries its own “?”. A blue bar marks a combined row, and the vs market column only fills in on those. A small tag beside the combination names any OTHER factor that earned the row its place on the page, when that differs from what Judge by is currently set to.
- 7
Click through. Planetary, KP and Daily rows open straight into the Explorer tab that owns them, with the condition already selected. Panchaang rows don’t: a panchaang bucket is a set of days, not one condition to re-open.
Where could these numbers mislead you?
What these numbers do and don't mean
Across every condition tested for this market, the strongest deviation from its own baseline doesn’t clear the bar that this many comparisons demand — on any of the six things Judge by can test. The note above the table says so in its own numbers, for the run and the factor you are looking at. Read every row as a historical base rate, not a proven edge.
The quality badge tells you how firm an estimate is, never whether it is real. Z reads against two different things depending on the row. On a single-axis row it compares against the market’s own baseline. On a 2-axis combination it compares against what those two axes predict together. The vs market column is the one that is always the market comparison. One exception: volatility never reads against the market baseline at all, even on a single-axis row — that comparison turns out to measure which years a slow-moving position passed through, not the position itself, so it is left blank rather than shown wrong.
Spans is blank on planetary-event rows on purpose. An event is a scattered occurrence rather than a condition that runs for days, so there is no episode to count.
The likeliest way to misread this
Picking whichever row happens to sort to the top. The table sorts by bias by default, not by how firm the estimate behind it is, so a big lean on a thin sample sorts above a smaller and sturdier one. Check the quality badge before treating a top row as a lead.
And reading a day’s realised move, in the panel pinned at the top, as a verdict on the rows below it. A row with a real lean is still wrong on plenty of individual days — that is arithmetic, not a broken pattern. One day is one draw, not a track record.
Which is the whole reason the rest of this site exists. A row here tells you a condition is active and what it has done before; it cannot tell you whether that will hold. Following the arrow into the Explorer, and from there into a backtest, is how you find out.
The standing caveat
Celestial Market Lens measures what markets did around historical planetary events. That is a description of the past. It is not a prediction, not a recommendation, and not investment, financial, legal or tax advice. Past results do not tell you what will happen next, and you are responsible for anything you do with money.