Edge Scanner
Every market event, tested on every NSE index and checked for luck the way quant researchers do. See which patterns survive, which are noise, and which are true in the latest week.
About the Edge Scanner
Most trading patterns you hear about were found by looking at enough charts until something seemed to work. Quant researchers know that if you test enough ideas, some will look brilliant by pure luck. This scanner tests 21 market events on every NSE index and checks each one the way a quant desk would.
Each pattern gets a luck chance (how often a result that good appears by chance), a check that it worked in both halves of the data, and a correction for the number of patterns tested. Only patterns that survive all three are marked as passing. The page also shows which of them are true in the latest week, as a description of the data, not a signal.
A second section measures each index's character with a variance-ratio test: whether its moves tend to keep going (trend) or snap back (bounce back), over the full history and the last three years.
Common questions
What is data mining bias?
When many patterns are tested on the same data, some look good by chance. About 1 in 20 random patterns passes a 5% significance test. The Edge Scanner counts every test it runs and corrects for it, so a lucky pattern is not mistaken for a real one.
What is a variance ratio test?
It compares how much an index moves over several weeks with how much it would move if every week were independent. A ratio above 1 means moves tend to continue (trend), below 1 means they tend to reverse (mean reversion).
Can I trade the patterns that pass?
They describe what happened in NSE index data since 2016. Passing the checks makes luck a less likely explanation, but markets change and past patterns can stop working. It is education, not a trading signal.