Checked again every night, across every pair we track
What moves together, and what just looks like it
When two lines rise together for years it is tempting to see a connection. Often there is none: almost everything that grows looks related to everything else that grows. So GLOVIZ runs one extra check on every pair, and sorts the results into real links and lucky ones.
How the test works
1Line them up
Every night we take each pair of series measured on the same rhythm, hourly with hourly, daily with daily, and line them up point by point.
2Ignore the long trend
Two things that both grow for years will always look related. So we also compare only the day-to-day changes: on the days one went up, did the other go up too?
3Keep what survives
Pairs whose daily changes agree across at least 30 shared days pass. Pairs that only match in the long lines land in the coincidence list below, on purpose.
See it for yourself
Each chart shows a pair that passed. One line per axis, so differently sized things can share a picture. Watch the wiggles, not the levels: they move together, or in mirror image.
Day-ahead power price DK1 and Day-ahead power price DK2Very strong, same direction · agreed across 400 shared days
Both series from the GLOVIZ database
Temperature Bogota and Humidity BogotaVery strong, opposite directions · agreed across 400 shared days
Both series from the GLOVIZ database
Passed: these move together day by day
PairDay-to-day agreementShared daysExtra evidence
Day-ahead power price DK1 × Day-ahead power price DK2
These pairs look impressively connected, until you check the daily movements and find nothing there. They just drifted in the same direction over the same years. We show them because recognising this pattern is the most useful statistics lesson there is.
Unemployment rate ES and Consumer price index NOLooks connected, is not · long-line match -0.962, day-to-day match only -0.098