The demonstrations that continue to draw hundreds of thousands of protesters into the streets of Tehran a week after the disputed presidential election are stoked by a conviction that the vote was shamelessly rigged.
Supporters of Mir Hossein Mousavi say they will settle for nothing less than a new election to rectify the injustice of a result that gave nearly 64 percent of the vote to incumbent Mahmoud Ahmadinejad.
They point to several unusual factors: the tabulation of more than 39 million handwritten ballots occurred far too quickly; hundreds of instances of irregularities were reported at polling places around the country; Ahmadinejad did well where previous election results showed he had done poorly; and 10 million ballots were missing personal identification numbers, rendering them untraceable.
"I am convinced they just pulled it out of their hats," said Farideh Farhi, an expert of Iranian elections at the University of Hawaii who was quoted in the Christian Science Monitor.
The partial recount the government has promised to begin today is unlikely to address all of the grievances, but a robust debate about the election results persists as political scientists, journalists and average citizens attempt to extract truth from a process that the Iranian government keeps deliberately opaque to international monitors.
Still, opinions are not uniform about whether something fishy happened in an election that had to be extended two hours because 11 million new voters showed up.
Writing in the online journal Slate, University of Wisconsin mathematician Jordan Ellenberg applied some statistical analysis to what critics identified very early as a suspicious consistency in Ahmadinejad's results. Hours after polls closed June 12, the blog Tehran Bureau pointed out that in the six waves of tallied votes, Ahmadinejad's total ranged from 62 to 70 percent.
Ellenberg concedes that consistency might look odd to American voters "who may be more used to seeing wide swings in the vote totals" because "our fine-grained media start reporting results when just a few percent of the votes are in."
But he finds nothing statistically inexplicable about the results. The batches of the votes were sufficiently large, he said, that they naturally absorbed the extremes of popularity that one would expect in a diverse electorate and delivered averages that were more or less uniform but not identically so.
Ellenberg doesn't pronounce on whether the election was fixed. "The official numbers may or may not be authentic," he wrote, "but they're messy enough to be true."
Mark Blumenthal in Pollster.com stakes his claim to fraud on the number crunching of Walter Mebane, the eminent political scientist and statistics professor at the University of Michigan, who zeroed in on "the large number of outliers" or tallies that diverge from predictable patterns.
Mebane looked at how Ahmadinejad fared in 320 towns in the 2005 election, expecting to find he did "best in towns where his support in 2005 was highest and tending to do worst where turnout surged the most."
But Ahmadinejad's results were much better than Mebane anticipated, enough that he sees "moderately strong support" for fraud.
Just look at the head-scratching results in Mousavi's home province of East Azerbaijan, say his supporters, where Ahmadinejad took 57 percent of the vote, an astonishing improvement over his 2005 total of 10 percent.
Still, polling done before the election indicated that Ahmadinejad — as unpopular as he has become for his handling of the economy — would nevertheless win by more than a 2-1 ratio. Phone interviews conducted May 11 and 20 across Iran by Terror Free Tomorrow: The Center for Public Opinion and the New America Foundation, showed Ahmadinejad with 34 percent of the vote and Mousavi with 14 percent.
Those results, say Mousavi's backers, don't reflect the dramatic surge in support for their candidate in the final weeks of the contest. Women and younger voters, unlikely to support the incumbent, were drawn into this race in droves.
"Analysts expected a closer race, if not a reverse of that result," said Farhi, the University of Hawaii professor.
Secret memos and "smoking gun" documents are playing their role in fueling outrage.
Before the election, Newsweek quoted "secret Iranian government polls" that showed Mousavi would win 16 million to 18 million votes and the incumbent a third of that.
A widely circulated letter, purportedly written by the minister of the interior to Supreme Leader Ayatollah Ali Khamenei, appears to prove the results were phonied at Khamenei's orders. The letter, the authenticity of which cannot be verified, also provides "real" vote totals that more or less jibe with the secret polls Newsweek cited — 19 million for Mousavi and just under 6 million for Ahmadinejad.
Robert Fisk, writing in Britain's Independent, wondered aloud whether the letter is a fake.
"However incredible Mr. Ahmadinejad's 63 percent of the vote may have been, could he really — as a man who has immense support among the poor of Iran — have picked up only 5½ million votes?"
What will happen if the protesters dressed in black and green for Mousavi don't get answers to these questions may well determine Iran's political future.
Bill Duryea can be reached at email@example.com.