Did P19.4-B election system flop? Eight incredible outcomes show it

Comelec broke its own rules on transparency last May 12th.

It bypassed its poll watchdogs, the media and political parties. It alone consolidated the votes, then fed the results to the public that was tired and wanted to move on.

A group of multifaith bishops, retired generals and civil society leaders is asking Comelec to explain what happened.

Without credible answers, the P19.4-billion automated election system – used only for one day – would have gone to waste.

Atty. Alex Lacson presented “eight grave concerns” about the balloting on May 12th. Last Wednesday’s forum was organized by Church Leaders Council for National Transformation, Alyansa ng Nagkaka-Isang Mamamayan and civil society.

(1) On Dec. 13, 2024, Comelec issued “ACM (automated counting machines) Operation Procedures.” It states: “10. Tap ‘Start Transmission’ button and then it will proceed with the transmission to the following servers:

“ • Municipal/City Board of Canvassers;

“ • Central Server;

“ • Citizen’s ARMM Server 1 (Parish Pastoral Council for Responsible Voting);

“ • Citizen’s ARMM Server 2 (National Citizens Movement for Free Elections);

“ • Dominant Majority Server;

“ • Dominant Minority Server;

“ • Media Server.”

But on Election Day, all precinct returns were not sent directly to PPCRV, Namfrel and media, but went through an intermediary server, Lacson said. Results were consolidated before being sent to PPCRV, Namfrel and media.

This was in violation of Comelec’s own Resolution 11098. The intermediary server was not among those listed to receive results from the ACMs.

(2) Comelec alone consolidated the results from all 93,287 precincts nationwide, Lacson noted. No participation of PPCRV, Namfrel and the media.

(3) From the 7 p.m. close of balloting, there was a long delay before Comelec released the first batch of the consolidated election results in the format agreed upon by the watchdogs and media. The delay lasted two hours, Lacson quoted Namfrel secretary general Eric Alviar.

“What happened during those two hours when PPCRV, Namfrel and media were blind as to the election results from precincts nationwide, when only Comelec had exclusive access and control to the results?” Lacson post.

PPCRV trustee-spokesperson Anna de Villa-Singson and Namfrel chairman Lito Averia were present at Wednesday’s forum. Both affirmed points 1 to 3. And both are still awaiting formal explanations from Comelec.

PNA photo

(4) Comelec merely supplied PPCRV, Namfrel and media with the consolidated election results, Lacson added.

Comelec was the only source of the complete election results from precincts. PPCRV, Namfrel and media had no other independent source of the complete election results aside from those supplied by Comelec, Lacson said.

Singson said later that PPCRV volunteers got printouts of 93,287 precinct results. As of Wednesday, PPCRV had tallied 40,000 and will complete it by next week. The sum will affirm or debunk’ Comelec’s count.

(5) A discrepancy of around five million votes was discovered by Rappler and other media teams, Lacson recounted.

Comelec said it was due to duplication in the transmission of some election returns. But Comelec was the only one who conducted the “cleaning” or “paglilinis” process of the vote discrepancy. It deleted around five million votes.

PPCRV, Namfrel and media were not present or involved in the clean-up process, Lacson said. Thereafter, Comelec merely supplied to PPCRV, Namfrel and media the cleaned or corrected data files.

The 2007 AES Law forbids any human intervention when balloting ends. Only machines commence precinct counts, transmit results to central servers and canvass the outcomes.

“Cleaning” or “paglilinis” are forms of human intervention.

(6) Countless voters complained that the voter receipts did not reflect the candidates they chose, Lacson said. Many ballots were also pre-shaded.

Comelec has yet to issue the official record of complaints.

(7) In the 2019 midterm elections, 1,170,331 official ballots had overvotes, Lacson recounted.

This 2025, Namfrel reported 17,028,780 overvotes. In elections 2010, 2013, 2016, 2019 and 2022, overvotes averaged 1.5-2.5 percent. This Election 2025, overvotes hit 29.89 percent.

Overvotes usually occur for 12 senatorial slots, eight for provincial board member and six or eight for municipal/city councilor.

But former information-communications technology chief Eliseo Rio noted that the 17,028,780 is exactly divisible by 12, for a 1,419,065 quotient. Meaning, the ACMs counted as 12 all overvotes for senator, whether 13, 14 or 15. As DICT secretary in 2019, he headed the Comelec advisory committee for that year’s election.

Former Comelec lawyer Melchor Magdamo said all overvotes in precincts, municipalities, cities and provinces were also divisible by 12.

In 2016, 2019 and 2022, Comelec defined overvotes by ballot, Singson recalled. Before that, Smartmatic-Dominion machines counted senatorial overvotes as the number of votes voided, or in increments of 12. She wondered why Miru reverted to the old definition this 2025.

(8) Comelec changed the program installed in the ACM and used an “updated” version 3.5 that did not go through the local source code review and independent certification by IT experts and organizations, Lacson asserted.

Comelec has claimed that versions 3.4 and 3.5 are the same.

If so, why didn’t it tell the public from the start, Lacson said.

In elections before 2010, ballots were counted openly, albeit slow. Precinct results were carted to municipal/city halls, but could be waylaid by goons. Votes were canvassed publicly, although sometimes padded or shaved (dagdag-bawas).

This Election 2025 was opaque. Voters never knew if machines counted or transmitted right. Comelec alone consolidated the precinct returns. Poll watchdogs and the public were fed only “final” results.

*      *      *

Follow me on Facebook: https://tinyurl.com/Jarius-Bondoc

Catch Sapol radio show, Saturdays, 8 to 10 a.m., dwIZ (882-AM)