Something Looks Weird in Broward County

fivethirtyeight.com

2 pointssxyuan2 comments

What I find interesting about this article (since it could easily turn unnecessarily political):

1. The use of data to discover an anomaly

2. The impact that poor UI design in one county in Florida (if you consider the ballot part of the voting UI) could potentially have on national politics.

It's odd to me that they've discovered an anomaly (statistically significant "undervoting" in one particular county) and yet that alone seems insufficient to trigger any sort of audit/recount/remediation. Especially since one of the three plausible explanations for the anomaly is hardware malfunction. What good is anomaly detection if you take no action from that data?

Reply on news.ycombinator.com