Corrections

O'Brien's response was originally recorded as "Yes" (For the removal) based on his initial survey email. Following a clarification email, his answer was changed to "No" (Against). We have updated the record to better reflect the candidate's position.

Gavin Morris has updated his answer on this question. He advises he had misread the initial question and has confirmed he supports the conservation of Link Road Forest from rezoning to housing. His response has been updated from "Against" to "For".

When a user skipped every question, the tool previously displayed candidates in the order they were entered into the database, which could be mistaken for a meaningful result. The tool now shows a message indicating no preference could be calculated.

When a user answered only one or two questions, the tool would still display a full ranked result. With so few data points, the alignment calculation has very little signal to work with, and candidates could appear in a consistent order that did not meaningfully reflect the user's values. The tool now requires a minimum of 6 questions to be answered before showing results, and displays a warning when fewer than 8 questions have been answered.

Alignment scores are calculated using a weighted Manhattan distance — measuring how far apart a user's answers are from each candidate's across all questions — and displayed rounded to the nearest whole number. Previously the tiebreaker only triggered when two candidates had identical raw scores, meaning candidates could show the same percentage without prompting a tiebreaker. The tool now triggers the tiebreaker whenever two or more candidates share the same rounded percentage, ensuring the displayed result always reflects a deliberate preference.