Audit Selections and Results for COUNTY CORONER Vote For 1

Audit units are selected for audit weighted by the maximum possible miscount (up) they could contain. The audit units listed below were selected using the random seed value 286630844411402 combined with the "Sequence" numbers according to the "Sum of Square Roots" pseudorandom number generator, to generate the values in the "Random" column. Each value from the "Random" column was then looked up in the "Cumulative Probability Total" column of the Unofficial Vote Tabulation Report, and the last row for which the "Cumulative Probability Total" was still larger than the Random number was selected.

This table lists the original machine counts, followed by the hand counts (as they become available), for each audit unit. If there is a discrepancy, the "Margin reduction" column lists the net reduction in the margin of victory for the unofficial winner, in votes. It will be negative if the margin increased. If there is a discrepancy, the "Taint" column will have a non-zero number. The KM column (Kaplan-Markov factor) indicates by what fraction each hand count reduces the probability of an incorrect outcome. It is larger when the Taint is greater than zero.

The Cumulative KM column summarizes the progress of the auditing for this contest. It comes from multiplying together all the KM factors up to and including that audit unit. When it becomes less than or equal to the pre-specified P-value (0.5 for this audit), the audit has reached its risk-reduction target. The progress is also shown in the graph, as the remaining risk drops down towards the target.

The first half of this list will be audited first, and if there are not many discrepancies with the hand count, that is all that will be needed. Otherwise, additional units moving down the list will be audited, until the target is reached. For more details, see the references on the main page.

Sequence Random Batches Ballots Maximum relative miscount: up Daniel H. Pruett Emma R. Hall Over Under Daniel H. Pruett Emma R. Hall Over Under Margin reduction Taint KM Cumulative KM
000001 0.962128 p371_ed_430 286 0.032638 66 118 0 102 66 118 0 102 0 0.924321 0.924321
000002 0.977478 p379_mb_258 294 0.034280 58 119 0 117 59 119 0 116 1 (see note) 0.002817 0.926932 0.856783
000003 0.226795 p078_mb_182 296 0.030224 96 113 0 87 96 113 0 87 0 0.924321 0.791942
000004 0.599014 p238_mb_085 300 0.026844 130 108 0 62 130 108 0 62 0 0.924321 0.732009
000005 0.316512 p118_mb_240 291 0.027424 105 98 1 87 105 98 1 87 0 0.924321 0.676611
000006 0.114737 p029_mb_153 299 0.033217 88 133 0 78 88 133 0 78 0 0.924321 0.625406
000007 0.639247 p255_mb_472 297 0.035245 69 137 0 91 69 137 0 91 0 0.924321 0.578076
000008 0.971298 p377_mb_260 299 0.038625 44 145 1 109 44 145 1 109 0 0.924321 0.534327
000009 0.866513 p337_ed_381 316 0.038818 60 146 0 110 60 146 0 110 0 0.924321 0.493890
000010 0.877588 p341_ed_412 281 0.035728 50 139 0 92
000011 0.423271 p164_mb_044 296 0.031769 90 123 0 83
000012 0.829834 p325_ed_417 343 0.039204 68 131 0 144
000013 0.357363 p135_mb_058 292 0.027520 113 106 0 73
000014 0.217157 p074_mb_231 295 0.029355 103 112 0 80
000015 0.583652 p231_mb_144 592 0.059000 196 215 0 181
000016 0.602845 p240_mb_099 299 0.031576 105 133 0 61
000017 0.398388 p153_mb_103 299 0.030321 107 122 0 70
000018 0.076099 p013_ev_220 300 0.035149 68 132 0 100

Discrepancy Notes

ContestBatch p379_mb_258: Ballot 046-DS-19: A check mark was resolved as a Pruett vote by the audit team and an undervote by the resolution team

KM Factor Graph

Here is an interactive graph of the KM factor data. Hover the mouse over the points to see the exact values.

Page generated: Mon, 29 Nov 2010 23:38:07 -0700