k-Cut: Simple Approximate Sampling 15
We showed that even for small values of k, our procedure provides a sample
that is close to being chosen uniformly at random. We designed a simple mit-
igation procedure for RLAs that accounts for any remnant non-uniformity, by
adjusting the risk limit. Finally, we provided a recommendation of k = 6 cuts
to use in practice, for sample sizes up to 1,000 ballots, based on our empirical
data, with a 1% risk limit adjustment.
An earlier version of k-cut was used in pilot audits in Marion County, Indiana
to increase audit efficiency. This paper provides theoretical justification for this
technique, which is also scheduled to be used in Michigan in December 2018.
References
1. Baign`eres, T., Vaudenay, S.: The complexity of distinguishing distributions (2008),
results also in Baigne`eres’ PhD thesis.
2. Banuelos, J.H., Stark, P.B.: Limiting risk by turning manifest phantoms into evil
zombies. https://arxiv.org/abs/1207.3413 (2012)
3. Bretschneider, J., Flaherty, S., Goodman, S., Halvorson, M., John-
ston, R., Lindeman, M., Rivest, R., Smith, P., Stark, P.: Risk-
limiting post-election audits: Why and how? (Oct 2012), (ver. 1.1)
http://people.csail.mit.edu/rivest/pubs.html#RLAWG12
4. Goggin, S.N., Byrne, M.D., Gilbert, J.E.: Post-election auditing effects of procedure
and ballot type on manual counting accuracy, efficiency, and auditor satisfaction
and confidence. Election Law Journal (2012)
5. Johnson, K.: Election verification by statistical audit of voter-verified paper ballots.
http://ssrn.com/abstract=640943 (Oct 31 2004)
6. Lindeman, M., Halvorseon, M., Smith, P., Garland, L., Addona, V.,
McCrea, D.: Principle and best practices for post-election audits.
www.electionaudits.org/files/best%20practices%20final 0.pdf (2008)
7. Lindeman, M., Stark, P.B.: A gentle introduction to risk-limiting audits. IEEE
Security and Privacy 10, 42–49 (2012)
8. Lindeman, M., Stark, P.B., Yates, V.S.: BRAVO: Ballot-polling risk-limiting au-
dits to verify outcomes. In: Halderman, A., Pereira, O. (eds.) Proceedings 2012
EVT/WOTE Conference (2012)
9. Miller, S.J., Nigrini, M.J.: The modulo 1 Central Limit Theorem and Benford’s
law for products. https://arxiv.org/abs/math/0607686 (2007)
10. Rivest, R.L.: Reference implementation code for pseudo-random sampler.
http://people.csail.mit.edu/rivest/sampler.py (2011)
11. Rivest, R.L.: Bayesian tabulation audits: Explained and extended.
https://arxiv.org/abs/1801.00528 (January 1, 2018)
12. Rivest, R.L., Shen, E.: A Bayesian method for auditing elections. In: Halder-
man, J.A., Pereira, O. (eds.) Proceedings 2012 EVT/WOTE Conference (2012),
https://www.usenix.org/system/files/conference/evtwote12/rivest bayes rev 073112.pdf,
https://www.usenix.org/conference/evtwote12/workshop-
program/presentation/rivest
13. Stark, P.B.: Papers, talks, video, legislation, software, and other documents on vot-
ing and election auditing. https://www.stat.berkeley.edu/ stark/Vote/index.htm
14. Stark, P.B.: Tools for ballot-polling risk-limiting election audits.
https://www.stat.berkeley.edu/ stark/Vote/ballotPollTools.htm (2017)