Converging on Each Other, Not the Jar

Guess-Spread Collapse Across a Visible Jellybean-Census Sign-In Sheet

A research poster from the School of Continuous Improvement compares 428 jar-guessing entries from eleven Union Market stalls, tracking within-stall guess spread and accuracy against the jar's counted total across a sealed-ballot and an open sign-in-sheet condition over an 11-week term. Guess spread on the open sheet narrows by 72% across the entrant sequence, but median absolute error against the counted total rises by 46% over the same span, while the sealed condition holds flat on both measures --- undercutting the Adaptive Metrics Lab's practice of reading the sheet's narrowing spread as evidence of improving collective estimation.

@misc{slop_hh9kvt,
  author       = {Mirela Hanke and Sten Okwuosa},
  title        = {Converging on Each Other, Not the Jar: Guess-Spread Collapse Across a Visible Jellybean-Census Sign-In Sheet},
  year         = {2026},
  publisher    = {Slop University},
  doi          = {10.5555/slop.hh9kvt},
  url          = {https://slop.university/outputs/slop-poster-a-jellybean-jar-guessing-hh9kvt/},
  version      = {1.0},
  note         = {Research poster},
}