By the Kunkafa team
How Kunkafa Estimates Price-Move Probabilities
One question, seventy answers. How agreement raises the probability, disagreement lowers it, and the record keeps the number honest.
Could gold go up 1% within 12 days? You want a number. And you want to know where the number came from before you trust it.
The finding. Seventy experts answer the same question on their own. Agreement raises the probability. Disagreement lowers it. Then every finished forecast is counted against what the market did, and the numbers are corrected.
Where Kunkafa stands. Kunkafa shows both directions, never makes them add up to 100, and publishes the count every morning.
What to do. Read the level, the direction and the duration together. Then look for the row that disagrees with you.
A forecast is four things
Every forecast in Kunkafa Predictions carries all four, and none means much without the other three.
- How far. The move as a change from today’s price, written as a percentage, so it reads the same on a share and on a currency.
- Which way. Up or down, each with its own value for Kunkafa’s confidence. The side you asked about comes first and the other sits below it. Either, both or neither could happen within the period, which is why the two values are never made to add up to 100.
- How long it has. The duration the move has to happen in. The same distance is a different proposition in an afternoon than in a year.
- Kunkafa’s confidence that it gets there. How likely the price is to reach that level within that duration, counted against how often similar forecasts actually did.
Put together, the homepage example reads: could gold go up 1% within 12 days? Kunkafa’s confidence: 84%. Below it, the other side: down 0.6% in the same period, 70%. The figures are illustrative. Nothing in them tells you what to do. They tell you what could happen, and how likely each path is.
One question, seventy answers
Kunkafa’s confidence is not one opinion. Seventy experts each look at the same market and answer the same question, how far and which way, from their own view of the price history. None sees what the others concluded.
They were trained on about 10 billion data points of price history across stocks, indices, currencies, commodities and crypto, and tested on about 2 billion they had never seen.
When the experts agree, the probability rises. When they disagree, it drops, and you see that too. A market that seventy views cannot settle is a market where both paths are genuinely open.
Then the check against outcomes
Agreement alone would only measure the experts. So every value for Kunkafa’s confidence was compared with how often the move actually arrived, and corrected where the two disagreed. That loop never stops. Every morning Kunkafa counts how past forecasts ended, on live markets, and publishes the result at Kunkafa forecast performance results, with the raw figures at /api/stats.json.
The same count sits beside every forecast in the app, over the last day, 7 days, 30 days and all time: forecasts evaluated, confident enough for the filter, and reached the level. Each completed result shows whether its level was reached within the duration. A near miss is not counted as reaching it.
Below 100 finished examples the record shows no rate and says why: not enough past examples to estimate performance reliably. A rate under 1% is written as under 1%.
How to read the ladder
Across, one row at a time. Each row says one complete thing: the price could move this far, within this long, with this value for Kunkafa’s confidence. Rows above the current price are the up paths. Rows below are the down paths. The price line beside the ladder carries the same levels drawn as a band, so the numbers and the picture are one forecast.
Three filters sit above it. “Only show forecasts Kunkafa is at least this confident about” keeps the less confident ones out of the list. “How far it has to go” narrows the distances. Both, Up or Down restricts the direction. All three narrow the list, and none changes the drawing. Removing a level from the picture would quietly claim the price cannot go there, and it can.
The duration control is a ceiling, not a choice between separate charts. Drag it from five minutes up to seven years, and you raise the longest duration you want to see. A market can lean up over the next hour and down over the next quarter without contradiction. Raising the ceiling shows both in one view.
Why most forecasts look dull
Whatever is known is already in the price, so most forecasts sit near 50/50, and a 52% is shown as 52%. A screen that leaned decisively every hour would be describing itself rather than the market. The dull majority is the honest answer, and the reason the few forecasts that lean are worth stopping on.
Why that is a feature of markets rather than a shortcoming is worked through in why no one can predict exact stock prices, and what to do with a number you cannot be sure of in understanding market uncertainty. Every term on the screen is defined in the Kunkafa market forecasting glossary.
Open a market you already have an opinion about in the Kunkafa market forecast app, and look for the row that disagrees with you. Note how far the other direction could run and how likely that is. Then see whether your opinion survives it.
Keep reading
Twelve guides that go with this one, by the question you came with.