@InProceedings{RS12b,
author = { Ronald L. Rivest and Emily Shen },
title = { A Bayesian Method for Auditing Elections },
url = { https://www.usenix.org/system/files/conference/evtwote12/rivest_bayes_rev_073112.pdf },
booktitle = { Proceedings 2012 EVT/WOTE Conference },
editor = { J. Alex Halderman and Olivier Pereira },
date = { 2012-08-07 },
eventtitle = { EVT/WOTE'12 },
eventdate = { 2012-08-06/2012-08-07 },
venue = { Seattle, WA },
OPTyear = { 2012 },
OPTmonth = { August 6-7, },
urla = { EVT/WOTE'12 },
urlb = { video },
urlc = { interactive implementation },
abstract = {
We propose an approach to post-election auditing based on Bayesian
principles, and give experimental evidence for its efficiency and
effectiveness. We call such an audit a ``Bayes audit''. It aims to
control the probability of miscertification (certifying a wrong
election outcome). The miscertification probability is computed using
a Bayesian model based on information gathered by the audit so far.
\par
A Bayes audit is a single-ballot audit method applicable to any voting
system (e.g.\ plurality, approval, IRV, Borda, Schulze, etc.)\ as long
as the number of ballot types is not too large. The method requires
only the ability to randomly sample single ballots and the ability to
compute the election outcome for a profile of ballots. A Bayes audit
does not require the computation of a ``margin of victory'' in order
to get started.
\par
Bayes audits are applicable both to ballot-polling audits, which work just
from the paper ballots, and to comparison audits, which work by
comparing the paper ballots to their electronic representations. The
procedure is quite simple and can be described on a single page.
\par
The Bayes audit uses an efficient method (which may be based on the
use of gamma variates or on P\'{o}lya's Urn) for simulating a Bayesian
posterior distribution on the tally of a profile of ballots.
\par
A Bayes audit is very similar to single-ballot risk-limiting audits.
However, since Bayes audits are based on different principles, the
precise relationship between risk-limiting audits and Bayes audits
remains open.
\par
We provide some initial experimental results indicating that Bayes
audits are quite efficient, requiring few ballots to be examined, and
that the miscertification rate is indeed kept small, even for very close
elections.
}
}