Philip Stark

Tag Archive

Colorado: Arapahoe County pioneering use of new vote verification system | The Denver Post

Arapahoe County is piloting a vote-checking system this week that promises to raise the level of confidence in the accuracy of election results in Colorado. Elections officials gathered Wednesday at the county’s clerk and recorder office in Littleton to put the system — dubbed the risk-limiting audit — through the paces. The goal is to work out the bugs and have it ready for statewide rollout by election day 2017, as required by the state legislature. “The way we do audits doesn’t present a good enough picture,” Arapahoe County Clerk Matt Crane said Wednesday. “Our citizens deserve to know that we have a fair, transparent and accurate voting process.” The way a post-election audit of ballots is done currently requires a canvass team to pull at least 500 randomly selected paper ballots and compare the results to the tally recorded by the tabulation machines used in the election. Under the risk-limiting audit, random numbers generated by a software program will identify certain ballots to be pulled for inspection. The sample size is statistically determined based on the total number of ballots cast, the margin of the contest and the audit results as they unfold. 

Full Article: Arapahoe County pioneering use of new vote verification system - The Denver Post.

Editorials: Saving throw: securing democracy with stats, spreadsheets, and 10-sided dice | Ars Technica

Armed with a set of 10-sided dice (we’ll get to those in a moment), an online Web tool, and a stack of hundreds of ballots, University of California-Berkeley statistics professor Philip Stark spent last Friday unleashing both science and technology upon a recent California election. He wanted to answer a very simple question—had the vote counting produced the proper result?—and he had developed a stats-based system to find out. On June 2, 6,573 citizens went to the polls in Napa County and cast primary ballots for supervisor of the 2nd District in one of California’s most famous wine-producing regions, on the northern edge of the San Francisco Bay Area. The three candidates—Juliana Inman, Mark van Gorder, and Mark Luce—would all have liked to come in first, but they really didn’t want to be third. That’s because only the two top vote-getters in the primary would proceed to the runoff election in November; number three was out. Napa County officials announced the official results a few days later: Luce, the incumbent, took in 2,806 votes, van Gorder got 1,911 votes, and Inman received 1,856 votes—a difference between second and third place of just 55 votes. Given the close result, even a small number of counting errors could have swung the election. Vote counting can go wrong in any number of ways, and even the auditing processes designed to ensure the integrity of close races can be a mess (did someone say “hanging, dimpled, or pregnant chads”?). Measuring human intent at the ballot box can be tricky. To take just one example, in California, many ballots are cast by completing an arrow, which is then optically read. While voters are instructed to fully complete the thickness of the arrow, in practice some only draw a line. The vote tabulation system used by counties sometimes do not always count those as votes. So Napa County invited Philip Stark to look more closely at their results. Stark has been on a four-year mission to encourage more elections officials to use statistical tools to ensure that the announced victor is indeed correct. He first described his method back in 2008, in a paper called “Conservative statistical post-election audits,” but he generally uses a catchier name for the process: “risk-limiting auditing.”

Full Article: Saving throw: securing democracy with stats, spreadsheets, and 10-sided dice | Ars Technica.

Florida: New post-election audit method could improve detection of vote-counting flaws – and fix them | Palm Beach Post

In California, 13 million people voted in the 2008 presidential race. But double-checking the result could have been as simple as looking at the right 96 ballots. Post-election audits in Florida are done with hand recounts of a sliver of ballots, taken from a few random precincts. They help identify widespread problems, as one did recently in Wellington, when it caught the fact that results for three races had been accidentally switched by Palm Beach County’s vote-counting software. But they tell you nothing about what happened in precincts that weren’t checked. In theory, a huge problem could go undetected. A new method of audit – developed by Philip Stark, statistics department vice chairman at the University of California, Berkeley – gets around that. Stark’s method works like an opinion poll, by looking at a random sample of ballots from across the race. The key word is random: The ballots have to be picked with precision, from a master list of every ballot cast. Once picked, only that ballot will do. The number of ballots reviewed depends on the margin of victory. Tighter races need more ballots. In California in 2008, Barack Obama won with 61 percent of the vote, so 96 ballots would do. If Obama had won with 52 percent, the state would have needed to check about 3,900 ballots, Stark said.

Full Article: New post-election audit method could improve detection of vote-counting flaws – and fix them.

Florida: New post-election audit method could improve detection of vote-counting flaws – and fix them | Palm Beach Post

In California, 13 million people voted in the 2008 presidential race. But double-checking the result could have been as simple as looking at the right 96 ballots. Post-election audits in Florida are done with hand recounts of a sliver of ballots, taken from a few random precincts. They help identify widespread problems, as one did recently in Wellington, when it caught the fact that results for three races had been accidentally switched by Palm Beach County’s vote-counting software. But they tell you nothing about what happened in precincts that weren’t checked. In theory, a huge problem could go undetected. A new method of audit – developed by Philip Stark, statistics department vice chairman at the University of California, Berkeley – gets around that. Stark’s method works like an opinion poll, by looking at a random sample of ballots from across the race. The key word is random: The ballots have to be picked with precision, from a master list of every ballot cast. Once picked, only that ballot will do. The number of ballots reviewed depends on the margin of victory. Tighter races need more ballots. In California in 2008, Barack Obama won with 61 percent of the vote, so 96 ballots would do. If Obama had won with 52 percent, the state would have needed to check about 3,900 ballots, Stark said.

Full Article: New post-election audit method could improve detection of vote-counting flaws – and fix them.