No poll fraud, says Brillantes
Comelec chief dismisses ‘60-30-10’ patternBy Jerome Aning, Philip C. Tubeza, Tarra Quismundo |Philippine Daily Inquirer
Commission on Elections (Comelec) Chairman Sixto Brillantes Jr. on Thursday rebutted the so-called “60-30-10” theory, saying he was willing to have all the ballot boxes in the country opened to show that there was no fraud.
Brillantes said he was confident that the review by political parties and citizen watchdogs of the source code of the precinct count optical scan (PCOS) machines would show that these were not programmed to cheat in the elections.
“Definitely, there was no fraud. I’m sure of that because the pattern that they’re saying, (if true), would mean that it was programmed. Definitely, it was not programmed,” Brillantes said.
“I will stake my reputation on that: No one programmed it,” he added.
Information technology (IT) experts have been linking with election watchdogs throughout the country to validate the so-called “60-30-10” pattern in which senatorial candidates from Team PNoy consistently obtained 60 percent of the votes cast in various precincts, those from the United Nationalist Alliance 30 percent and the rest of the candidates 10 percent.
“If there’s a pattern, it could probably be a coincidence but it cannot have happened everywhere,” Brillantes said.
“We are having the source code reviewed now. We invite the parties to review it. No pattern will be discerned,” he said.
“In fact, the final evidence would be if we open all the ballot boxes in the Philippines. I’m willing to make a bet that 98 (or) 99 percent of the ballots and the PCOS would have the same results,” he added.
No predetermined pattern
The Parish Pastoral Council for Responsible Voting (PPCRV) also saw “no real pattern” that would suggest that election results had been predetermined, as doubts piled up over the number of votes in favor of the administration coalition.
Ana de Villa Singson, PPCRV media and communication director, said an initial analysis of election results had showed no pattern in the numbers, contrary to what some IT experts found.
“There doesn’t seem to be a predetermined pattern that would show that the results are predictable and formulaic,” Singson told the Inquirer.
“The insinuation (in the 60-30-10 pattern) is that the results are predetermined, but it doesn’t appear to be that way,” she added.
The PPCRV has been receiving election results both electronically and manually as the Comelec’s accredited citizen’s arm for the parallel count and audit, conducted in the interest of transparency.
Initiating its own analysis, the PPCRV on Thursday assembled a team of three experts in mathematics, statistics and computer technology to study the figures and issue a publicly accessible opinion on the 60-30-10 theory.
“We’ll be coming out with a consolidated opinion tomorrow (Friday) on the 60-30-10 theory. I told them to come up with something that would be easy for the public to understand,” Singson said.
New data analyses released by several IT and statistics experts seem to debunk the 60-30-10 theory.
The IT experts and statisticians generated new charts and graphs using different methods to show that the pattern may not be tenable.
Peter Julian Cayton, University of the Philippines statistics professor, computed the change in the accumulated votes for Team PNoy, UNA and the rest of the candidates, and calculated the proportion of the total change that went to each slate in the 16 canvass reports of the Comelec.
The computation of change in the accumulated votes per slate per interval was just the difference from the result of the slate from the last reported interval with respect to a current reported interval.
From this, Cayton computed the proportion of how much of the total change went to specific slates.
“Here, the 60-30-10 hypothesis becomes a little shaky (pun intended),” Cayton said in a public status update on his Facebook account, as he contradicted the findings of Ateneo de Manila University math professor Felix Muga II and blogger Conrad Miguel Gozalo of Radar Sweep.
Cayton said additions to the accumulated votes per slate varied during the intervals. He found that Team PNoy got increment votes ranging from 54.7567 percent to 63.9845 percent, while UNA got between 26.7176 percent and 39.4894 percent.
The candidates from the minor political parties, grouped together, obtained 5.7539 percent to 10.2232 percent of total increments in accumulated votes.
Insufficient fraud evidence
“What does this mean? As more and more votes got tallied, the additional votes coming to each slate are added with no systematic pattern, which leads me to conclude there is no sufficient evidence for election fraud. I’m not saying there is no fraud, but based on the evidence or statistics collected, fraud cannot be concluded unless there is stronger evidence for it,” he said.
Cayton said he was not shocked by the initial reports of the 60-30-10 voting pattern.
“Why? Because in statistics, especially in dealing with proportions, as more and more ballots get counted, the proportion of voters favoring a candidate will converge at a specific decimal or fraction of total votes. This is known as ‘statistical regularity.’ To be more specific, it’s called the Law of Large Numbers,” he added.
Muga, a fellow of the Center for People Empowerment in Governance, earlier said that while the certificates of canvas (COCs) were randomly received by the Comelec, there should have been variations in the votes cast as candidates would get big number of votes in places considered their strongholds.
Rick Bahague of the Computer Professionals Union also found the 60-30-10 pattern in his analysis of results from the Visayas.
However, while Muga was the first to call the voting pattern “60-30-10,” it was former Comelec IT director Ernie del Rosario who first referred to the consistent 9-3 pattern (nine for Team PNoy and three UNA candidates who were later proclaimed winners) in the COCs.
Calling the pattern “9-3 formula,” Del Rosario said there was “a predesigned results reporting mechanism” that would fit the results of the preelection survey showing a 9-3 victory in favor of the administration.
Del Rosario’s claims were quoted by Rene Azurin, a convener of the poll watchdog Automated Election System Watch and a columnist, who later voiced out his suspicion about the possible internal tampering of results.
Blogger M Beduya of Synthesist (http://synthesistblog.com), however, contradicted Azurin and Del Rosario’s allegations.
“By observation, despite the correlation, causation can be seen to run the opposite way—i.e. the winner ratios dictated the final voter proportions,” he said.
Using the 16 canvassing reports, Beduya generated a graph of the votes obtained by the individual candidates per interval. The graph showed candidates getting or losing votes ranging from three percent to eight percent per canvassing.
From Beduya’s accompanying chart of votes per candidate, Team PNoy candidate Grace Poe consistently topped the canvass. Juan Miguel Zubiri also gained votes when the 15th canvassing report, which contains results from his bailiwick Lanao del Norte, came in.
On the other hand, reelectionist Gregorio Honasan placed 13th in two intervals but rose to 12th in the rest of the canvassing period. Although his votes began to shrink in the 15th and 16th canvassing reports, the 13th placer, Richard Gordon, was unable to gain more votes to surpass Honasan’s.