Broward County finally finished counting ballots at about 11:30 p.m. Thursday, said Broward elections spokeswoman Evelyn Perez-Verdia. Palm Beach and Duval were still tabulating their absentees as of Thursday afternoon.

Miami-Dade staffers on Thursday also reviewed about 2,870 provisional ballots. They accepted only 1,000 of them, rejecting the others for various reasons, deputy elections supervisor Christina White said.

All the voting results were sent to the state. The county canvassing board will certify the election on Nov. 16.

Townsley said her elections staff was prepared for the presidential race turnout and lengthy ballot, which included numerous county and state amendment questions. She said she deployed 200-plus more scanning machines and 400 more poll workers for this election compared with 2008, and made trouble-shooting decisions Tuesday to shift resources where needed.

Asked why there were waits up to six hours at various precincts in the Brickell area of Miami, as well as in West Kendall, Country Walk, Goulds and Homestead, Townsley ducked the question without providing details.

“That is precisely the reason we will be conducting an after-action report to determine what actually went wrong,” she said. “We will learn from those lessons.”

The Election Day ballot, which many officials blamed for the voting delays, was one for the record books, with 11 state constitutional proposals, 10 county charter changes, assorted municipal questions, congressional races, judicial contests and the presidential race.

All but a tiny percentage of voters made a choice for president, but to varying degrees ignored other races and questions further down the 10-page ballot, according to statistics released by the elections department.

Many voters skipped the state amendment questions, in percentages ranging from 13 to nearly 21 percent.

County charter questions also drew less attention. About 19 percent of voters ignored a question to impose term limits on county commissioners, a measure overwhelmingly approved by 77 percent of the people who did vote on it. About a quarter of voters ignored a charter change that made it tougher to build outside the county’s urban development line, which was approved by 68 percent of people who did vote on it.

A whopping 37 percent of voters ignored the single county judge race on the ballot.

Christopher Mann, an assistant professor of political science at the University of Miami, said he saw no unusual trends in what elections analysts call “down-ballot roll-off.’’ The term reflects the phenomenon of questions farther down on ballots typically getting fewer votes. High-interest topics like gay marriage or marijuana legalization, however, would likely prove exceptions.

Mann said the roll-off from a presidential race can typically hit 25 to 30 percent, with the least attention paid to races like county judgeships, where there is no political party listed and most voters know little about candidates.

“A lot of voters don’t really understand what they’re voting on. They don’t know the judges, they don’t have any cues like political party affiliations,’’ Mann said. “They don’t feel like they have enough information, so they skip it.”

Typically, the longer the ballot, the higher the roll-off, Mann said. Excessive roll-off can be an indicator of a problem ballot. But after examining the data from Miami-Dade, Mann said he did not see any major red flags.

“Overall, the length of the ballot is a challenge for voting participation, but the roll-off in Miami-Dade is in the normal range,’’ he said.

Miami Herald staff writer Charles Rabin contributed to this report.