Print Visualizations at Data Stories 2019

The great test of Hungarian weather-related popular wisdoms (A nagy népi időjóslás-teszt)

Author: Dénes Csala

- Data source: NOAA historical weather data (1952-2019)
- Description: We collected most Hungarian popular wisdoms of weather forecast and constructed a testable “accuracy” metric. Each saying has two parts: a condition (feltétel in Hungarian: e.g. “Ha Katalinkor kopog” – i.e. if the 25th of Novemberis colder than average) and a prophecy (jóslat in Hungarian: “Karácsonykor locsog” – in this case, between 24-27 of December, we will have warmer than average weather). Then, for each saying, we looked at the years that the condition was fulfilled and calculated the percentage of years when the prophecy did come true. Overall, we found that the Hungarian popular wisdoms of weather forecast re only slightly better than chance: 56%. Of course, the historically harbinger days of Spring or Autumn are the most accurate, but also the cold spell predictor days in January fare well, over 70%. The more reknown sayings (mostly connected to name-days) also perform better than average: Sándor, József, Bendek gets 65%, Medárd 66%, László 79%-ot, Szent Anna 75%, Péter 68%, Márton 62% and Katalin 67%. The data was formatted in python and the plotting tool used was matplotlib. Try the online version, too, at: https://szekelydata.csaladen.es/nepi/
- Message: We examined most Hungarian popular wisdoms of weather forecast and constructed a testable “accuracy” metric. Utilizing data from past 60 years we find that, on average, these predictions come true only 56% of the time.