Observed arrival · 2026-09-19
Anàlisi Castellera, Bayesian Models for Human Towers
An interactive Catalan-language analysis tool for comparing casteller groups, estimating the difficulty of human towers, and exploring contest outcomes.
- For
- Casteller groups, competition analysts, and Catalan tower enthusiasts
- Worth noticing
- The model estimates group trajectories, tower difficulty, and selection preferences using historical performance data and MCMC.
Field notes
The site organizes its analysis into contest rankings, longitudinal group comparisons, tower-difficulty estimates, and a focused group view. Its model combines outcome probabilities with a separate account of tower selection, estimating latent group trajectories and difficulty thresholds through MCMC. The homepage says the underlying data is updated through 13 September 2026 and emphasizes that displayed predictions retain uncertainty and variability.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue