Porovnanie s overvolby.sk (2022)

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [same person elsewhere: č.7 also regional_council 2022-OSK-C-207: 79 of 579 ballots; č.8 also regional_council 2022-OSK-C-207: 7 of 579 ballots; adversarial review 2022-OSO-C-506745-1|506745-3: confirmed_error (high; changes the elected; docs/reviews/2022-OSO-C-506745-1__506745-3.md)]

legal allocation of the published votes adds Erika Suhajdová to the flagged set; competing reading: Štefan Ádám (108 votes) is marked withdrawn here but is an active mayoral candidate with 0 mayoral votes (2022-OSO-M-514764), so the withdrawal mark may belong to the mayoral list; then the legal allocation elects Štefan Ádám, Štefan Agócs, Gabriel Habodász, Pavel Ráki, Tibor Singlár

our deterministic check flags it (tier A_arithmetic): the elected flags contradict tab02a seats filled; under the published withdrawal marks the legal allocation (§ 174(4)) gives the last seat to Erika Suhajdová; competing reading: Štefan Ádám (108 votes) is marked withdrawn here but is an active mayoral candidate with 0 mayoral votes (2022-OSO-M-514764), so the withdrawal mark may belong to the mayoral list; then the legal allocation elects Štefan Ádám, Štefan Agócs, Gabriel Habodász, Pavel Ráki, Tibor Singlár; anomaly confirmed, their correction contradicted: the more probable reading seats the 'withdrawn' candidate, not theirs [adversarial review 2022-OSO-C-514764-1|: confirmed_error (high; seat effect open; docs/reviews/2022-OSO-C-514764-1__.md)]

recorded candidate mix is compatible with a partial tally (composition p=0.87); missing ballots simulated: P(change)=0.173 (flat prior 0.06); 0.83 unchanged; 0.15 enters Richard Kočiš / exits Ingrid Faťolová; 0.02 tie; P(Richard Kočiš in, Ingrid Faťolová out)=0.15. [If instead the vector is not a count: P(change)=0.796]

we flag the record (tier C_ambiguous); utilisation 0.310 is below every other precinct of the contest (0.742–0.801 (n=5)); their deficit 560 is inside our 95 % interval for missing marks (485–644); turnout forms: OSO 306 valid council of 327, OSO_CITY 300 valid council of 327, OSK 311 valid council of 327

OSO−OSK participation gap 518 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 708 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 607 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 602 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 565 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (errorsearch:turnout_digit,joint:)

OSO−OSK participation gap 674 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 925 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 603 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (novel_turnout:other)

OSO−OSK participation gap 847 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 754 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 594 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 334 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 548 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 397 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 299 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 465 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 239 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 276 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 484 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 556 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 453 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 488 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 371 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 482 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 320 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 537 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 409 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 305 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 117 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 505 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 252 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 260 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 298 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 112 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 294 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 259 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 289 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 458 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 159 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 160 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 234 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 271 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 180 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 239 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (novel_turnout:other)

OSO−OSK participation gap 152 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 200 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 72 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 213 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 124 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 176 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 36 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 65 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 97 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 120 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 40 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 95 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 142 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 155 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 95 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 67 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 39 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 52 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 42 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 105 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 26 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 144 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 137 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 92 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:)

OSO−OSK participation gap 88 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 105 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [same person elsewhere: č.19 also regional_council 2022-OSK-C-101: 10 of 484 ballots; č.26 also president 2022-OSK-P-1: 6 of 497 ballots; č.26 also regional_council 2022-OSK-C-101: 9 of 484 ballots; č.29 also mayor_city 2022-OSO-M-582000: 11 of 493 ballots; č.17 also regional_council 2022-OSK-C-101: 16 of 484 ballots; č.30 also regional_council 2022-OSK-C-101: 4 of 484 ballots; č.32 also regional_council 2022-OSK-C-101: 28 of 484 ballots; č.37 also regional_council 2022-OSK-C-101: 8 of 484 ballots; č.18 also regional_council 2022-OSK-C-101: 36 of 484 ballots; č.36 also regional_council 2022-OSK-C-101: 5 of 484 ballots; č.21 also regional_council 2022-OSK-C-101: 2 of 484 ballots; č.31 also regional_council 2022-OSK-C-101: 58 of 484 ballots; č.40 also mayor 2022-OSO-M-528595: 79 of 491 ballots; č.40 also regional_council 2022-OSK-C-101: 70 of 484 ballots; č.39 also council 2022-OSO-C-528595-4: 126 of 480 ballots; č.39 also regional_council 2022-OSK-C-101: 68 of 484 ballots; č.27 also mayor 2022-OSO-M-528595: 4 of 491 ballots; č.27 also regional_council 2022-OSK-C-101: 5 of 484 ballots; č.35 also regional_council 2022-OSK-C-101: 1 of 484 ballots; adversarial review 2022-OSO-C-582000-1|528595-13: confirmed_error (high; no seat change; docs/reviews/2022-OSO-C-582000-1__528595-13.md)]

we flag the same cells (tier A/B), but our competing correction is ≥10× more likely and 29 touched cell(s) remain atypical after their correction [our correction (crossballot:missing_block,crossyear:,errorsearch:missing_block,joint:cross_contest_copy,novel_dup:duplicate_cross_contest,slates:missing_block) is 10^98.3× more likely; same person elsewhere: č.42 also regional_council 2022-OSK-C-102: 54 of 492 ballots; č.36 also regional_council 2022-OSK-C-102: 65 of 492 ballots; č.37 also regional_council 2022-OSK-C-102: 44 of 492 ballots; č.50 also mayor 2022-OSO-M-529320: 12 of 495 ballots; č.50 also regional_council 2022-OSK-C-102: 69 of 492 ballots; č.19 also regional_council 2022-OSK-C-102: 42 of 492 ballots; č.17 also mayor 2022-OSO-M-529320: 237 of 495 ballots; č.17 also regional_council 2022-OSK-C-102: 183 of 492 ballots; č.41 also regional_council 2022-OSK-C-102: 47 of 492 ballots; č.29 also regional_council 2022-OSK-C-102: 15 of 492 ballots; č.35 also regional_council 2022-OSK-C-102: 100 of 492 ballots; č.46 also regional_council 2022-OSK-C-102: 52 of 492 ballots; č.44 also mayor 2022-OSO-M-529320: 210 of 495 ballots; č.44 also council 2022-OSO-C-529320-9: 278 of 485 ballots; č.51 also regional_council 2022-OSK-C-102: 227 of 492 ballots; č.27 also regional_council 2022-OSK-C-102: 44 of 492 ballots; č.28 also council 2022-OSO-C-529320-9: 251 of 485 ballots; č.28 also regional_council 2022-OSK-C-102: 254 of 492 ballots; adversarial review 2022-OSO-C-582000-3|529320-69: confirmed_error (high; no seat change; docs/reviews/2022-OSO-C-582000-3__529320-69.md)]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [same person elsewhere: č.42 also regional_council 2022-OSK-C-102: 63 of 277 ballots; č.49 also council 2022-OSO-C-529320-5: 87 of 272 ballots; č.36 also regional_council 2022-OSK-C-102: 17 of 277 ballots; č.37 also council 2022-OSO-C-529320-5: 117 of 272 ballots; č.37 also regional_council 2022-OSK-C-102: 83 of 277 ballots; č.53 also regional_council 2022-OSK-C-102: 9 of 277 ballots; č.50 also mayor 2022-OSO-M-529320: 5 of 287 ballots; č.50 also regional_council 2022-OSK-C-102: 27 of 277 ballots; č.45 also regional_council 2022-OSK-C-102: 5 of 277 ballots; č.34 also council 2022-OSO-C-529320-5: 29 of 272 ballots; č.31 also regional_council 2022-OSK-C-102: 9 of 277 ballots; č.41 also regional_council 2022-OSK-C-102: 37 of 277 ballots; č.35 also regional_council 2022-OSK-C-102: 111 of 277 ballots; č.55 also regional_council 2022-OSK-C-102: 12 of 277 ballots; č.33 also regional_council 2022-OSK-C-102: 3 of 277 ballots; č.48 also regional_council 2022-OSK-C-102: 19 of 277 ballots; č.30 also regional_council 2022-OSK-C-102: 4 of 277 ballots; č.47 also regional_council 2022-OSK-C-102: 6 of 277 ballots; č.46 also regional_council 2022-OSK-C-102: 18 of 277 ballots; č.54 also regional_council 2022-OSK-C-102: 8 of 277 ballots; č.44 also mayor 2022-OSO-M-529320: 23 of 287 ballots; č.51 also regional_council 2022-OSK-C-102: 87 of 277 ballots; č.39 also regional_council 2022-OSK-C-102: 4 of 277 ballots; adversarial review 2022-OSO-C-582000-3|529320-36: confirmed_error (high; no seat change; docs/reviews/2022-OSO-C-582000-3__529320-36.md)]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [same person elsewhere: č.7 also president 2022-OSK-P-1: 2 of 359 ballots; adversarial review 2022-OSO-M-582000|529346-9: confirmed_error (high; no seat change; docs/reviews/2022-OSO-M-582000__529346-9.md)]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [same person elsewhere: č.7 also president 2022-OSK-P-1: 5 of 352 ballots; adversarial review 2022-OSO-M-582000|529346-33: confirmed_error (high; no seat change; docs/reviews/2022-OSO-M-582000__529346-33.md)]

recorded candidate mix is incompatible with a partial tally of this precinct (composition p=0.0003), so the vector is treated as uninformative and redrawn from our predictive: P(change)=0.016; 0.98 unchanged; 0.01 enters Matej Kučera / exits Peter Herceg; 0.00 enters Lucia Štasselová / exits Peter Herceg. [Rejected partial-tally model: P(change)=0.42, flat prior 0.96]

we flag the record independently (a review-confirmed error); utilisation 0.333 is below every other precinct of the contest (0.853–0.923 (n=6)); their deficit 458 is inside our 95 % interval for missing marks (405–617); turnout forms: OSO 298 valid council of 334, OSO_CITY 303 valid council of 334, OSK 315 valid council of 334 [adversarial review 2022-OSO-C-529320-5|529320-35: confirmed_error (high; no seat change; docs/reviews/2022-OSO-C-529320-5__529320-35.md)]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [same person elsewhere: č.7 also president 2022-OSK-P-1: 4 of 273 ballots; adversarial review 2022-OSO-M-582000|529460-52: confirmed_error (high; no seat change; docs/reviews/2022-OSO-M-582000__529460-52.md)]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [same person elsewhere: č.7 also president 2022-OSK-P-1: 3 of 517 ballots; adversarial review 2022-OSO-M-582000|529320-42: confirmed_error (high; no seat change; docs/reviews/2022-OSO-M-582000__529320-42.md)]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [same person elsewhere: č.7 also president 2022-OSK-P-1: 3 of 346 ballots; adversarial review 2022-OSO-M-582000|529389-24: confirmed_error (high; no seat change; docs/reviews/2022-OSO-M-582000__529389-24.md)]

we flag the same cells (tier A/B), but our competing correction is ≥10× more likely [our correction (crossballot:row_shift,crossyear:swap,errorsearch:row_shift,withdrawn:row_shift) is 10^3.4× more likely; same person elsewhere: č.7 also president 2022-OSK-P-1: 5 of 311 ballots; adversarial review 2022-OSO-M-582000|529460-5: confirmed_error (high; no seat change; docs/reviews/2022-OSO-M-582000__529460-5.md)]

we flag the same cells (tier A/B), but our competing correction is ≥10× more likely [our correction (crossballot:row_shift,crossyear:swap,errorsearch:row_shift,withdrawn:row_shift) is 10^1.5× more likely; same person elsewhere: č.7 also president 2022-OSK-P-1: 0 of 386 ballots; adversarial review 2022-OSO-M-582000|529311-2: confirmed_error (high; no seat change; docs/reviews/2022-OSO-M-582000__529311-2.md)]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [same person elsewhere: č.7 also president 2022-OSK-P-1: 3 of 388 ballots; adversarial review 2022-OSO-M-582000|529389-10: confirmed_error (high; no seat change; docs/reviews/2022-OSO-M-582000__529389-10.md)]

we flag the same cells (tier A/B), but our competing correction is ≥10× more likely [our correction (crossballot:row_shift,errorsearch:row_shift,withdrawn:row_shift) is 10^1.1× more likely; same person elsewhere: č.7 also president 2022-OSK-P-1: 3 of 336 ballots; adversarial review 2022-OSO-M-582000|529389-12: confirmed_error (high; no seat change; docs/reviews/2022-OSO-M-582000__529389-12.md)]

we flag the same cells (tier A/B), but our competing correction is ≥10× more likely and 1 touched cell(s) remain atypical after their correction [our correction (crossballot:row_shift,crossyear:,errorsearch:row_shift,withdrawn:row_shift) is 10^1.5× more likely; same person elsewhere: č.7 also president 2022-OSK-P-1: 5 of 330 ballots; adversarial review 2022-OSO-M-582000|529346-1: confirmed_error (high; no seat change; docs/reviews/2022-OSO-M-582000__529346-1.md)]

we flag the same cells (tier A/B), but our competing correction is ≥10× more likely and 1 touched cell(s) remain atypical after their correction [our correction (crossballot:row_shift,errorsearch:row_shift,withdrawn:row_shift) is 10^1.1× more likely; same person elsewhere: č.7 also president 2022-OSK-P-1: 4 of 302 ballots; adversarial review 2022-OSO-M-582000|529389-26: confirmed_error (high; no seat change; docs/reviews/2022-OSO-M-582000__529389-26.md)]

we flag the same cells (tier A/B), but 1 touched cell(s) remain atypical after their correction [same person elsewhere: č.7 also president 2022-OSK-P-1: 5 of 296 ballots; adversarial review 2022-OSO-M-582000|529389-27: confirmed_error (high; no seat change; docs/reviews/2022-OSO-M-582000__529389-27.md)]

we flag the same cells (tier A/B), but our competing correction is ≥10× more likely [our correction (crossballot:row_shift,errorsearch:row_shift,slates:row_shift) is 10^2.1× more likely; same person elsewhere: č.15 also president 2022-OSK-P-6: 17 of 312 ballots; č.8 also regional_council 2022-OSK-C-601: 25 of 291 ballots; č.14 also regional_council 2022-OSK-C-601: 4 of 291 ballots; č.10 also regional_council 2022-OSK-C-601: 66 of 291 ballots; č.18 also regional_council 2022-OSK-C-601: 88 of 291 ballots; adversarial review 2022-OSO-C-508438-3|508438-18: confirmed_error (high; no seat change; docs/reviews/2022-OSO-C-508438-3__508438-18.md)]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [adversarial review 2022-OSK-P-6|514462-25: probable_error (medium; no seat change; docs/reviews/2022-OSK-P-6__514462-25.md)]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [adversarial review 2022-OSK-P-6|511471-2: probable_error (high; no seat change; docs/reviews/2022-OSK-P-6__511471-2.md)]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [adversarial review 2022-OSK-P-6|508497-13: probable_error (high; no seat change; docs/reviews/2022-OSK-P-6__508497-13.md)]

their corrected values are less likely than the published ones under our model (ΔlogL=-26.2 nats) [same person elsewhere: č.30 also president 2022-OSK-P-7: 18 of 295 ballots]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [adversarial review 2022-OSK-P-6|526142-11: probable_error (high; no seat change; docs/reviews/2022-OSK-P-6__526142-11.md)]

recorded candidate mix is compatible with a partial tally (composition p=0.79); missing ballots simulated: P(change)=0.000 (flat prior 0.00); 1.00 unchanged. [If instead the vector is not a count: P(change)=0.000]

we flag the record independently (a review-confirmed error); utilisation 0.500 is below every other precinct of the contest (0.862–0.941 (n=17)); their deficit 296 is inside our 95 % interval for missing marks (271–380); turnout forms: OSO 448 valid council of 523, OSK 381 valid council of 514 [adversarial review 2022-OSK-C-117|508179-6: confirmed_error (high; no seat change; docs/reviews/2022-OSK-C-117__508179-6.md)]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [a claimed candidate lives in this municipality; same person elsewhere: č.6 also regional_council 2022-OSK-C-601: 36 of 226 ballots; č.8 also mayor 2022-OSO-M-508438: 0 of 236 ballots; č.8 also regional_council 2022-OSK-C-601: 3 of 226 ballots; adversarial review 2022-OSK-P-6|508438-16: probable_error (high; no seat change; docs/reviews/2022-OSK-P-6__508438-16.md)]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [same person elsewhere: č.17 also regional_council 2022-OSK-C-107: 34 of 528 ballots; č.16 also regional_council 2022-OSK-C-107: 22 of 528 ballots; č.18 also regional_council 2022-OSK-C-107: 28 of 528 ballots; adversarial review 2022-OSO-C-582000-11|529397-28: confirmed_error (high; no seat change; docs/reviews/2022-OSO-C-582000-11__529397-28.md)]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [adversarial review 2022-OSK-P-6|517143-1: probable_error (high; no seat change; docs/reviews/2022-OSK-P-6__517143-1.md)]

their corrected values are less likely than the published ones under our model (ΔlogL=-11.0 nats)

their correction fits better than the published value (ΔlogL=+6.6 nats) but not enough to overcome the prior of a clerical error (posterior 0.0074; errorsearch P(error)=0.0063); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+8.7 nats) but not enough to overcome the prior of a clerical error (posterior 0.37; errorsearch P(error)=0.33); no detector of ours flags it

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [same person elsewhere: č.3 also regional_council 2022-OSK-C-111: 44 of 301 ballots; adversarial review 2022-OSO-C-582000-16|529460-47: probable_error (high; no seat change; docs/reviews/2022-OSO-C-582000-16__529460-47.md)]

we flag the same cells independently (tier A/B or a review-confirmed error) and their correction fits our model [same person elsewhere: č.8 also council_city 2022-OSO-C-582000-11: 112 of 532 ballots; č.8 also regional_council 2022-OSK-C-107: 219 of 528 ballots; adversarial review 2022-OSO-C-529397-4|529397-28: probable_error (high; no seat change; docs/reviews/2022-OSO-C-529397-4__529397-28.md)]

we flag it independently (tier A_arithmetic: checks:X_CITY_BOROUGH_ENVELOPES,checks:X_CITY_BOROUGH_PARTICIPATED); the same commission's city and borough records disagree: participated city 460 vs borough 484 (other); envelopes city 460 vs borough 484 (other) [adversarial review precinct|529320-43: confirmed_error (high; no seat change; docs/reviews/precinct__529320-43.md)]

we flag it independently (tier A_arithmetic: checks:X_CITY_BOROUGH_REGISTERED,novel_turnout:turnout_digit); the same commission's city and borough records disagree: registered city 884 vs borough 844 (one_digit) [adversarial review precinct|529389-18: confirmed_error (high; no seat change; docs/reviews/precinct__529389-18.md)]

we flag it independently (tier A_arithmetic: checks:X_CITY_BOROUGH_ENVELOPES,checks:X_CITY_BOROUGH_PARTICIPATED,checks:X_CITY_BOROUGH_REGISTERED,joint:turnout_digit,novel_turnout:turnout_digit); the same commission's city and borough records disagree: registered city 784 vs borough 874 (transposition); participated city 236 vs borough 234 (one_digit); envelopes city 236 vs borough 234 (one_digit) [adversarial review precinct|599816-3: confirmed_error (high; no seat change; docs/reviews/precinct__599816-3.md)]

their correction fits better than the published value (ΔlogL=+7.4 nats) but not enough to overcome the prior of a clerical error (posterior 0.016; errorsearch P(error)=0.0036); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+6.9 nats) but not enough to overcome the prior of a clerical error (posterior 0.0094; errorsearch P(error)=0.002); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+0.6 nats) but not enough to overcome the prior of a clerical error (posterior 1.9e-05; errorsearch P(error)=0.0093); no detector of ours flags it

error-free explanation: 3 unclaimed candidates of the same residence (Sereď) on this ballot poll 3.7× their expectation; with rates scaled to these peers the published value is more likely than their correction (ΔlogL=-9.3 nats)

we flag it independently (tier A_arithmetic: checks:X_CITY_BOROUGH_ENVELOPES,checks:X_CITY_BOROUGH_PARTICIPATED,checks:X_CITY_BOROUGH_REGISTERED,joint:turnout_digit,novel_turnout:turnout_digit); OSO 874 vs OSK 784 registered (transposition); city record agrees with OSK [adversarial review precinct|599816-3: confirmed_error (high; no seat change; docs/reviews/precinct__599816-3.md)]

we flag it independently (tier B_strong: joint:turnout_digit,novel_turnout:turnout_digit); OSO 988 vs OSK 688 registered (one_digit) [adversarial review precinct|500933-2: confirmed_error (high; no seat change; docs/reviews/precinct__500933-2.md)]

we flag it (tier C: joint:turnout_digit,novel_turnout:turnout_digit); OSO 825 vs OSK 925 registered (one_digit)

we flag it independently (tier B_strong: joint:turnout_digit,novel_turnout:copy_from_OSK); OSO 293 vs OSK 737 registered (other) [adversarial review precinct|514462-17: confirmed_error (high; no seat change; docs/reviews/precinct__514462-17.md)]

we flag it independently (tier B_strong: joint:turnout_digit,novel_turnout:turnout_digit); OSO 1570 vs OSK 1270 registered (one_digit) [adversarial review precinct|508497-1: confirmed_error (high; no seat change; docs/reviews/precinct__508497-1.md)]

we flag it independently (tier B_strong: joint:turnout_digit,novel_turnout:turnout_digit); OSO 453 vs OSK 543 registered (transposition) [adversarial review precinct|523119-1: confirmed_error (high; no seat change; docs/reviews/precinct__523119-1.md)]

we flag it (tier C: novel_turnout:copy_from_OSK); OSO 1069 vs OSK 1008 registered (other)

we flag it independently (tier A_arithmetic: checks:X_CITY_BOROUGH_REGISTERED,novel_turnout:turnout_digit); OSO 844 vs OSK 884 registered (one_digit); city record agrees with OSK [adversarial review precinct|529389-18: confirmed_error (high; no seat change; docs/reviews/precinct__529389-18.md)]

we flag it (tier C: novel_turnout:turnout_digit); OSO 1438 vs OSK 1483 registered (transposition)

we flag it independently (tier B_strong: joint:turnout_digit,novel_turnout:turnout_digit); OSO 682 vs OSK 861 registered (other) [adversarial review precinct|523381-5: confirmed_error (high; no seat change; docs/reviews/precinct__523381-5.md)]

we flag it (tier C: novel_turnout:turnout_digit); OSO 822 vs OSK 882 registered (one_digit)

their correction fits better than the published value (ΔlogL=+6.6 nats) but not enough to overcome the prior of a clerical error (posterior 0.0074; errorsearch P(error)=0.0045); no detector of ours flags it

our adversarial review of this record concludes the published values stand (probable_genuine) [adversarial review 2022-OSK-C-608|514721-1: probable_genuine (medium; no seat change; docs/reviews/2022-OSK-C-608__514721-1.md)]

their corrected values are less likely than the published ones under our model (ΔlogL=-8.2 nats)

we flag it independently (tier A_arithmetic: checks:X_CITY_BOROUGH_REGISTERED,novel_turnout:turnout_digit); the same commission's city and borough records disagree: registered city 984 vs borough 974 (one_digit) [adversarial review precinct|529397-15: confirmed_error (high; no seat change; docs/reviews/precinct__529397-15.md)]

their correction fits better than the published value (ΔlogL=+8.2 nats) but not enough to overcome the prior of a clerical error (posterior 0.035; errorsearch P(error)=0.0072); no detector of ours flags it

their corrected values are less likely than the published ones under our model (ΔlogL=-10.9 nats)

their correction fits better than the published value (ΔlogL=+7.9 nats) but not enough to overcome the prior of a clerical error (posterior 0.026; errorsearch P(error)=0.037); no detector of ours flags it

no detector of ours flags it; OSO 278 vs OSK 268 registered (one_digit)

we flag it independently (tier A_arithmetic: checks:X_CITY_BOROUGH_ENVELOPES,novel_turnout:turnout_digit); the same commission's city and borough records disagree: envelopes city 532 vs borough 522 (one_digit) [adversarial review precinct|529320-42: confirmed_error (high; no seat change; docs/reviews/precinct__529320-42.md)]

their correction fits better than the published value (ΔlogL=+6.6 nats) but not enough to overcome the prior of a clerical error (posterior 0.007; errorsearch P(error)=0.0071); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+7.1 nats) but not enough to overcome the prior of a clerical error (posterior 0.012; errorsearch P(error)=0.018); no detector of ours flags it

we flag it independently (tier A_arithmetic: checks:X_CITY_BOROUGH_ENVELOPES,checks:X_CITY_BOROUGH_PARTICIPATED,checks:X_CITY_BOROUGH_REGISTERED,novel_turnout:turnout_digit); the same commission's city and borough records disagree: registered city 1276 vs borough 1279 (one_digit); participated city 638 vs borough 644 (other); envelopes city 628 vs borough 638 (one_digit) [adversarial review precinct|598216-1: confirmed_error (high; no seat change; docs/reviews/precinct__598216-1.md)]

their corrected values are less likely than the published ones under our model (ΔlogL=-3.9 nats) [a claimed candidate lives in this municipality]

their correction fits better than the published value (ΔlogL=+8.4 nats) but not enough to overcome the prior of a clerical error (posterior 0.043; errorsearch P(error)=0.0072); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+8.2 nats) but not enough to overcome the prior of a clerical error (posterior 0.26; errorsearch P(error)=0.24); no detector of ours flags it

their corrected values are less likely than the published ones under our model (ΔlogL=-0.5 nats)

their correction fits better than the published value (ΔlogL=+6.7 nats) but not enough to overcome the prior of a clerical error (posterior 0.0078; errorsearch P(error)=0.0036); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+3.9 nats) but not enough to overcome the prior of a clerical error (posterior 0.00048; errorsearch P(error)=0.0018); no detector of ours flags it

we flag the same cells (tier A/B), but our competing correction is ≥10× more likely and 1 touched cell(s) remain atypical after their correction [our correction (crossyear:,errorsearch:missing_block,joint:missing_block,novel_dup:missing_block,novel_util:missing_block,slates:missing_block) is 10^4.1× more likely; adversarial review 2022-OSK-C-301|505072-1: confirmed_error (high; no seat change; docs/reviews/2022-OSK-C-301__505072-1.md)]

our adversarial review of this record concludes the published values stand (probable_genuine) [adversarial review 2022-OSK-C-608|514721-1: probable_genuine (medium; no seat change; docs/reviews/2022-OSK-C-608__514721-1.md)]

error-free explanation: 4 unclaimed candidates of the same party list (145) on this ballot poll 3.7× their expectation; with rates scaled to these peers the published value is more likely than their correction (ΔlogL=-4.3 nats) [we flag this record, but on other cells (crossballot:digit_ins)]

their correction fits better than the published value (ΔlogL=+3.0 nats) but not enough to overcome the prior of a clerical error (posterior 0.04; errorsearch P(error)=0.0041); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+3.1 nats) but not enough to overcome the prior of a clerical error (posterior 0.00022; errorsearch P(error)=0.0046); no detector of ours flags it

their corrected values are less likely than the published ones under our model (ΔlogL=-4.1 nats)

their correction fits better than the published value (ΔlogL=+6.8 nats) but not enough to overcome the prior of a clerical error (posterior 0.15; errorsearch P(error)=0.00067); no detector of ours flags it

their corrected values are less likely than the published ones under our model (ΔlogL=-6.6 nats)

their correction fits better than the published value (ΔlogL=+3.3 nats) but not enough to overcome the prior of a clerical error (posterior 0.00026; errorsearch P(error)=0.0014); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+4.9 nats) but not enough to overcome the prior of a clerical error (posterior 0.0013; errorsearch P(error)=2.3e-05); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+6.2 nats) but not enough to overcome the prior of a clerical error (posterior 0.0049; errorsearch P(error)=0.015); no detector of ours flags it

error-free explanation: 18 unclaimed candidates of the same residence (Brezno) on this ballot poll 0.4× their expectation; with rates scaled to these peers the published value is more likely than their correction (ΔlogL=-7.9 nats)

their correction fits better than the published value (ΔlogL=+4.0 nats) but not enough to overcome the prior of a clerical error (posterior 0.00054; errorsearch P(error)=0.0027); no detector of ours flags it

we flag the same cells (tier A/B), but our competing correction is ≥10× more likely and 1 touched cell(s) remain atypical after their correction [our correction (crossyear:,errorsearch:missing_block,joint:missing_block,novel_dup:missing_block,novel_util:missing_block,slates:missing_block) is 10^4.3× more likely; adversarial review 2022-OSK-C-301|505072-1: confirmed_error (high; no seat change; docs/reviews/2022-OSK-C-301__505072-1.md)]

error-free explanation: 4 unclaimed candidates of the same residence (Kráľovský Chlmec) on this ballot poll 2.9× their expectation; with rates scaled to these peers the published value is more likely than their correction (ΔlogL=-6.9 nats)

their correction fits better than the published value (ΔlogL=+3.1 nats) but not enough to overcome the prior of a clerical error (posterior 0.00022; errorsearch P(error)=0.0059); no detector of ours flags it

their corrected values are less likely than the published ones under our model (ΔlogL=-3.8 nats) [a claimed candidate lives in this municipality; same person elsewhere: č.16 also council 2022-OSO-C-520004-2: 5 of 262 ballots]

their corrected values are less likely than the published ones under our model (ΔlogL=-8.7 nats)

error-free explanation: 2 unclaimed candidates of the same party list (106) on this ballot poll 4.2× their expectation; with rates scaled to these peers the published value is more likely than their correction (ΔlogL=-5.0 nats)

we flag the same cells (tier A/B), but our competing correction is ≥10× more likely and 1 touched cell(s) remain atypical after their correction [our correction (crossyear:,errorsearch:missing_block,joint:missing_block,novel_dup:missing_block,novel_util:missing_block,slates:missing_block) is 10^4.4× more likely; adversarial review 2022-OSK-C-301|505072-1: confirmed_error (high; no seat change; docs/reviews/2022-OSK-C-301__505072-1.md)]

our adversarial review of this record concludes the published values stand (probable_genuine) [adversarial review 2022-OSK-C-608|514721-1: probable_genuine (medium; no seat change; docs/reviews/2022-OSK-C-608__514721-1.md)]

error-free explanation: 2 unclaimed candidates of the same residence (Kráľovský Chlmec) on this ballot poll 2.7× their expectation; with rates scaled to these peers the published value is more likely than their correction (ΔlogL=-4.9 nats)

their correction fits better than the published value (ΔlogL=+7.8 nats) but not enough to overcome the prior of a clerical error (posterior 0.024; errorsearch P(error)=0.0069); no detector of ours flags it

their corrected values are less likely than the published ones under our model (ΔlogL=-4.9 nats)

their correction fits better than the published value (ΔlogL=+0.5 nats) but not enough to overcome the prior of a clerical error (posterior 1.6e-05; errorsearch P(error)=0.0019); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+2.0 nats) but not enough to overcome the prior of a clerical error (posterior 7.7e-05; errorsearch P(error)=0.0093); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+4.7 nats) but not enough to overcome the prior of a clerical error (posterior 0.0011; errorsearch P(error)=0.0049); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+5.5 nats) but not enough to overcome the prior of a clerical error (posterior 0.0025; errorsearch P(error)=0.047); no detector of ours flags it

error-free explanation: 3 unclaimed candidates of the same party list (145) on this ballot poll 3.6× their expectation; with rates scaled to these peers the published value is more likely than their correction (ΔlogL=-3.9 nats)

error-free explanation: 3 unclaimed candidates of the same party list (145) on this ballot poll 3.6× their expectation; with rates scaled to these peers the published value is more likely than their correction (ΔlogL=-5.4 nats)

their correction fits better than the published value (ΔlogL=+5.5 nats) but not enough to overcome the prior of a clerical error (posterior 0.0023; errorsearch P(error)=0.015); no detector of ours flags it

their corrected values are less likely than the published ones under our model (ΔlogL=-6.8 nats)

their correction fits better than the published value (ΔlogL=+2.6 nats) but not enough to overcome the prior of a clerical error (posterior 0.00014; errorsearch P(error)=0.0024); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+4.6 nats) but not enough to overcome the prior of a clerical error (posterior 0.00097; errorsearch P(error)=0.015); no detector of ours flags it

error-free explanation: 3 unclaimed candidates of the same party list (145) on this ballot poll 3.6× their expectation; with rates scaled to these peers the published value is more likely than their correction (ΔlogL=-5.5 nats)

their correction fits better than the published value (ΔlogL=+0.4 nats) but not enough to overcome the prior of a clerical error (posterior 1.5e-05; errorsearch P(error)=0.045); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+4.8 nats) but not enough to overcome the prior of a clerical error (posterior 0.0012; errorsearch P(error)=0.00079); no detector of ours flags it

their correction fits better than the published value (ΔlogL=+4.8 nats) but not enough to overcome the prior of a clerical error (posterior 0.0013; errorsearch P(error)=0.046); no detector of ours flags it

their corrected values are less likely than the published ones under our model (ΔlogL=-4.5 nats)

their corrected values are less likely than the published ones under our model (ΔlogL=-4.8 nats)

their corrected values are less likely than the published ones under our model (ΔlogL=-3.6 nats) [a claimed candidate lives in this municipality; same person elsewhere: č.16 also council 2022-OSO-C-520004-2: 5 of 262 ballots]

their correction fits better than the published value (ΔlogL=+2.9 nats) but not enough to overcome the prior of a clerical error (posterior 0.00018; errorsearch P(error)=0.0045); no detector of ours flags it

their corrected values are less likely than the published ones under our model (ΔlogL=-4.8 nats)

their correction fits better than the published value (ΔlogL=+5.5 nats) but not enough to overcome the prior of a clerical error (posterior 0.0025; errorsearch P(error)=0.0035); no detector of ours flags it

error-free explanation: 2 unclaimed candidates of the same residence (Kráľovský Chlmec) on this ballot poll 2.7× their expectation; with rates scaled to these peers the published value is more likely than their correction (ΔlogL=-6.1 nats)

their corrected values are less likely than the published ones under our model (ΔlogL=-3.1 nats)

OSO−OSK participation gap 34 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour

OSO−OSK participation gap 0 is in the published records, but a voter may lawfully take only the municipal ballots; no vote-level error or correction is claimed, so the data cannot separate an error from voter behaviour; our turnout detectors flag this precinct (joint:turnout_digit,novel_turnout:copy_from_OSK) [adversarial review precinct|514462-17: confirmed_error (high; no seat change; docs/reviews/precinct__514462-17.md)]