iiinova — where your digital self keeps evolving

Philip B. Stark

@philipbstark.bsky.social

40 followers 3 following 9 posts

Posts Replies Media

Philip B. Stark @philipbstark.bsky.social · 13/02/2025

Even compared to polio and smallpox? Please elaborate.

110

Reposted by Philip B. Stark

Marilyn Marks @marilynrmarks.bsky.social · 11/12/2024

If you write about, administer, or care about elections, this dialogue between @milesobrien3.bsky.social and Prof. Alex Halderman is a MUST listen for a fact-based look at the credibility of the November election. And more importantly, the dire need for election security improvements.

286

Philip B. Stark @philipbstark.bsky.social · 04/12/2024

5. Most importantly, an RLA **requires a trustworthy, organized record of the votes.** Many jurisdictions do not yet have such a record, for many reasons.

053

Philip B. Stark @philipbstark.bsky.social · 04/12/2024

4/x The sample size depends on the true margin, on whether the outcome is correct, on the sampling design (stratified, cluster, individual ballots), and on the type of data from the voting system the RLA relies on (just the outcome, margin, cast-vote records, batch subtotals, etc.)

130

Philip B. Stark @philipbstark.bsky.social · 04/12/2024

3/x. 3. Risk-limiting audits do not necessarily use a smaller sample size. They are designed to collect as much data as necessary to show that the outcome is right, or to lead to a full hand count if the outcome isn't right.

130

Philip B. Stark @philipbstark.bsky.social · 04/12/2024

2/x 2. Risk-limiting audits do not "tell the probability that a full audit would correct a wrong outcome." Risk-limiting audits have a known minimum probability of leading to a full handcount if the reported outcome is wrong.

141

Philip B. Stark @philipbstark.bsky.social · 04/12/2024

1/x This is seriously garbled. 1. No audit even attempts to ensure that every vote is accurately counted. The best audits, RLAs, only check whether the reported winner really won. ...

184

Philip B. Stark @philipbstark.bsky.social · 27/11/2024

Direct democracy is a great idea, but the National Popular Vote Interstate Compact is a terrible idea. It will undermine what election integrity the US has; it’s easily undermined; and it will not provide direct democracy. papers.ssrn.com/sol3/papers....

papers.ssrn.com

031

Philip B. Stark @philipbstark.bsky.social · 21/11/2024

If the votes that are counted pass through the internet, it’s internet voting. Internet voting is not trustworthy. arxiv.org/abs/2411.11796

arxiv.org

An Internet Voting System Fatally Flawed in Creative New Ways

The recently published "MERGE" protocol is designed to be used in the prototype CAC-vote system. The voting kiosk and protocol transmit votes over the internet and then transmit voter-verifiable paper...

076

Philip B. Stark @philipbstark.bsky.social · 18/11/2024

Hello, world!

100