Monday, May 20, 2013

Mining the Election Data for Evidence of Good or Evil (Part 1 of 4)

My good friend, RENE AZURIN, writes a stinging condemnation of the recent May 13 elections in Business World last week "Uncritical Media, Illegitimate Elections".

I heartily agree with his criticisms of much of the main stream media
The problem is that many media practitioners today are not intellectually prepared to deal with complex issues and are often too lazy to put in the work needed to prepare to deal with such issues. This is evident in the quality of our talk shows where the host is content to lob superficial "gimme" questions at a guest instead of engaging him/her in serious discussions that will illuminate difficult subjects or unmask hidden agendas.
Let me just add what most people already know that that self-same media needed no more intellectual preparation--corporately or individually--to receive the BILLIONS in election campaign spending, than to know NOT to bite the generous hands that feed it.

So let me turn to the second part of the article dealing with the LEGITIMACY of the recent polls, which utilized an Automated Election System based on portable Precinct Count Optical Scanners.  I shall focus mainly on the technical critique  and leave Comelec's allegedly irregular deviation from various elections laws and procedures for a later piece.

Perhaps because of those irregularities, long-running suspicions and accusations have once more arisen that the Smartmatic PCOS system is hackable from within, and vulnerable to all sorts of digital legerdemain that could be used for wholesale, computerized cheating. Indeed, since the election data has become publicly available and online, many technically adept and even not-so-adept critics say there are obvious signs of such "hocus PCOS" in the election data itself.

   Rene quotes two critics of the system and their observations, as follows:

As first pointed out by political activist Ado Paglinawan, the way the 2013 election results came in was "highly suspicious." He correctly observed that, "from the smallest count to the biggest count, there is consistency in the space between the first 15 senatorial candidates…. The progression through the night is mathematically predictable, and is a statistical improbability." The nationwide trend observed by Mr. Paglinawan in the senatorial tally indicated to him that the count "was following a pre-programmed formula based on earlier pre-paid surveys, rather than the actual vote." It was clear to him, he wrote, "that an earlier decision of ranking had been predetermined and the proportion of votes had been pre-designated from a national perspective, with a total disregard for provincial and regional nuances…. From 10% of the vote to 60%, the tally has been running a consistent vote share. As the votes from different provinces came in, the voting pattern was identical for the senatorial positions, something contrary to historical experience in Philippine politics."

Former Comelec IT director Ernie del Rosario adds: "The progressive tallies follow some sort of deterministic linear equation devoid of the influence of any probabilistic parameter or variable. This can only mean one thing -- it is a pre-designed results reporting mechanism that fits the 9-3 survey instead of a tally of the actual votes. I will call it the 9-3 Formula. Notice that the rankings of the candidates in the entire tally (1st to 33rd place) from the time the first report was published to subsequent ones are practically unchanged. What happened to the individual candidates’ known bailiwicks that should have caused some ranking movements in the tallied results? Smoothened by the 9-3 linear formula?" Mr. del Rosario then wryly remarks, "Magdadaya rin lang ang mga ito, medyo sana lagyan nila ng konting pag-iisip [These guys who planned to cheat should have maybe put a little more thought into it]."
 Throughout the week since the election, I've been engaged with many others on Twitter and Facebook discussing and debating these very points.  So I decided to look at all the data myself and decide for myself about these accusations, without any prejudices or pre-conceived notions. Luckily the data both gentlemen quoted by Rene above is contained in one compact package maintained by the superb team at Rapplerdotcom  headed by Ms. Maria Ressa. The Official Tally of Votes of the 2013 Elections

A table at the above location contains the individual votes of the 33 Senatorial candidates who vied for the 12 available seats in each of 16 Canvassing Stages conducted by the Comelec last week and which resulted in the proclamation by last Friday night (in 3 separate, widely-criticized partial proclamations) of the Magic 12 winners.

It is this data that both Mr. Ado Paglinawan and Ernie del Rosario are referring to.  In what follows I shall endeavour to present the actual data in a form that will allow readers of Philippine Commentary to participate with me in a thoroughgoing examination and critique of the accusations.

In this post I shall do not much more than present the information before sharing my own conclusions about them.

First to this statement from Mr. Paglinawan quoted above:

"from the smallest count to the biggest count, there is consistency in the space between the first 15 senatorial candidates…. 
In the graph below, which I prepared using a spreadsheet program and the data in the Official Tally, readers will see a series of 33 curves each with 16 data points. The X-axis of this graph represents the 16 tallies performed by the National Board of Canvassers on some 304 local and overseas "Certificates of Canvass" containing some 294 million votes for the 33 candidates cast by some more than 39 million voters who turned out for the election.   The Y-axis shows the PERCENTAGE of the TOTAL VOTES ALREADY CANVASSED at each stage with which each of the 33 candidates were credited.

Please examine this graphic carefully and consider the statement above of Mr. Paglinawan in the clear light of this simple plot.  The 'smallest count' he speaks of is Canvas #1 and the 'biggest count' is Canvas #16 at which point some 294 million votes had been counted.




The tabulated data is prettier  found at the Rappler website, but is also here as a simple listing:

ALCANTARA, Samson SJS 1,227,521 1,219,652 1,206,022 1,174,068 1,090,613 1,048,079 959,870 780,065 605,021 417,657 382,929 260,495 215,956 186,692 85,331 51,351

.
ANGARA, Sonny LDP 15,858,995 15,770,821 15,626,960 15,286,774 14,714,561 14,079,635 12,759,795 10,287,276 7,919,081 5,593,601 5,232,977 3,790,897 3,244,883 2,803,181 1,250,443 698,784

.
AQUINO, Bam LP 15,388,992 15,280,039 15,123,026 14,761,981 14,086,288 13,466,762 12,263,781 9,874,431 7,488,847 5,255,177 4,927,748 3,535,501 3,046,469 2,532,438 1,181,655 649,222

.
BELGICA, Greco DPP 1,118,829 1,113,725 1,104,566 1,083,284 1,030,445 986,686 896,573 745,918 592,252 419,104 396,543 286,000 253,021 225,203 87,520 54,430

.
BINAY, Nancy UNA 16,645,515 16,554,119 16,399,739 15,997,420 15,358,420 14,725,283 13,525,314 10,802,550 8,329,945 5,787,423 5,315,184 3,825,405 3,181,537 2,724,274 1,341,777 750,816

.
CASINO, Teddy Makabayan 4,254,245 4,242,259 4,203,937 4,093,259 3,930,504 3,748,036 3,339,166 2,806,706 2,151,295 1,549,394 1,449,793 1,063,605 921,212 812,330 337,732 182,651

.
CAYETANO, Alan NP 17,408,543 17,318,674 17,153,733 16,773,997 16,168,156 15,453,195 14,022,191 11,145,518 8,597,603 5,989,369 5,597,679 4,049,550 3,423,951 2,954,461 1,321,587 764,740

.
COJUANGCO, Tingting UNA 3,091,642 3,080,639 3,057,518 2,991,871 2,847,173 2,755,337 2,559,892 2,102,359 1,650,075 1,210,914 1,156,801 922,370 753,688 552,396 270,442 149,116

.
DAVID, Lito Kapatiran 1,026,096 1,020,879 1,012,215 983,718 927,379 892,476 814,936 651,555 487,883 337,831 315,428 229,612 200,547 173,925 78,781 47,584

.
DELOS REYES, JC Kapatiran 1,226,470 1,221,045 1,210,298 1,178,654 1,108,529 1,063,738 962,319 785,224 577,663 411,082 388,969 286,015 252,545 222,073 96,466 55,679

.
EJERCITO, JV UNA 13,552,991 13,476,446 13,360,579 12,999,725 12,566,069 12,026,870 10,950,878 8,877,587 7,044,012 4,939,197 4,602,150 3,319,079 2,846,576 2,485,370 1,131,921 620,151

.
ENRILE, Jack NPC 11,419,246 11,346,552 11,253,382 10,994,555 10,575,933 10,160,318 9,354,186 7,374,497 5,796,044 4,125,323 3,686,140 2,704,522 2,231,785 1,917,484 912,764 527,218

.
ESCUDERO, Chiz independent 17,332,952 17,246,756 17,071,568 16,718,197 16,148,955 15,438,853 14,011,873 11,326,358 8,652,216 5,944,573 5,581,201 4,067,421 3,476,228 3,023,792 1,367,597 768,547

.
FALCONE, Baldomero DPP 659,073 654,576 646,870 631,217 592,909 569,099 523,901 419,062 319,184 220,572 203,432 142,560 121,084 104,033 47,167 29,669

.
GORDON, Dick UNA 12,364,091 12,324,291 12,252,317 12,048,097 11,673,359 11,261,192 10,172,274 8,194,864 6,235,930 4,479,287 4,254,472 3,177,270 2,798,395 2,441,159 1,081,080 584,822

.
HAGEDORN, Ed independent 8,323,835 8,305,240 8,249,271 8,133,173 7,901,113 7,582,899 6,561,396 5,387,475 4,203,297 3,032,770 2,935,767 2,178,915 1,964,644 1,780,551 728,735 448,544

.
HONASAN, Gringo UNA 13,070,031 13,019,299 12,909,206 12,569,266 12,153,688 11,669,991 10,634,845 8,543,963 6,730,774 4,667,276 4,335,339 3,191,420 2,701,796 2,339,377 1,028,376 600,326

.
HONTIVEROS, Risa Akbayan 10,840,047 10,786,438 10,693,676 10,462,646 10,115,604 9,666,484 8,711,159 7,055,561 5,344,173 3,723,857 3,526,134 2,525,968 2,171,727 1,889,947 812,230 463,594

.
LEGARDA, Loren NPC 18,482,961 18,387,850 18,197,504 17,791,849 17,160,208 16,468,744 14,922,018 11,896,817 9,175,488 6,405,625 5,967,495 4,343,113 3,626,984 3,103,432 1,420,068 835,114

.
LLASOS, Marwil Kapatiran 695,260 692,371 687,252 667,132 629,789 601,039 548,277 458,725 349,850 236,675 223,233 160,766 145,303 130,177 51,043 30,541

.
MACEDA, Ernie UNA 3,388,936 3,376,159 3,353,442 3,286,111 3,170,748 3,052,521 2,770,465 2,269,605 1,776,549 1,295,902 1,216,797 931,610 767,289 686,409 311,907 167,272

.
MADRIGAL, Jamby LP 6,727,877 6,686,088 6,611,189 6,474,155 6,173,959 5,908,546 5,385,103 4,293,929 3,228,413 2,238,534 2,082,226 1,470,667 1,251,959 1,076,984 488,405 287,740

.
MAGSAYSAY, Mitos UNA 5,569,077 5,534,697 5,492,306 5,354,545 5,115,384 4,887,849 4,490,187 3,541,173 2,677,457 1,874,627 1,755,815 1,277,692 1,101,619 967,269 484,364 266,221

.
MAGSAYSAY, Ramon LP 11,252,335 11,184,979 11,079,149 10,835,974 10,377,948 9,929,494 9,012,758 7,207,461 5,492,410 3,913,064 3,682,585 2,657,890 2,302,425 2,015,378 927,154 528,165

.
MONTANO, Mon independent 1,029,439 1,022,853 1,012,475 985,743 902,787 867,213 792,362 608,672 466,688 333,514 309,177 210,762 177,611 154,701 68,548 44,650

.
PENSON, Dick independent 1,030,107 1,024,739 1,015,097 993,195 947,431 906,543 821,812 669,039 520,659 367,963 345,938 258,867 224,329 199,030 83,849 52,263

.
PIMENTEL, Koko PDP-Laban 14,584,612 14,492,401 14,365,080 14,000,348 13,466,306 12,860,609 11,700,820 9,530,039 7,336,084 5,186,529 4,828,881 3,510,976 3,011,023 2,626,465 1,193,232 658,102

.
POE, Grace independent 20,147,423 20,029,370 19,828,262 19,376,744 18,697,429 17,895,281 16,260,239 12,986,253 10,049,795 6,978,111 6,504,294 4,729,889 4,011,600 3,440,764 1,520,880 850,911

.
SENERES, Christian DPP 698,440 694,770 689,671 663,322 625,781 597,128 539,761 442,083 331,665 235,004 222,477 161,070 141,095 125,288 53,555 30,757

.
TRILLANES, Sonny NP 13,995,603 13,911,783 13,785,975 13,430,245 12,958,307 12,371,458 11,221,926 9,115,190 7,098,940 4,968,187 4,666,117 3,325,169 2,829,457 2,440,836 1,076,984 631,426

.
VILLANUEVA, Eddie Bangon Pilipinas 6,868,774 6,844,731 6,796,369 6,681,232 6,446,252 6,210,511 5,589,056 4,597,592 3,549,486 2,405,127 2,254,514 1,660,673 1,473,986 1,278,655 557,457 321,880

.
VILLAR, Cynthia NP 13,696,120 13,625,351 13,494,702 13,206,695 12,725,809 12,221,970 11,143,934 8,897,239 6,815,833 4,833,722 4,504,611 3,249,937 2,722,646 2,327,995 1,066,122 613,661

.
ZUBIRI, Migz UNA 11,707,146 11,617,336 11,544,658 11,191,359 10,506,446 10,046,457 9,184,995 7,526,167 5,697,367 4,164,541 3,838,705 2,697,780 2,314,780 2,023,140 998,939 512,574

COMMENT LINES ARE OPEN and I shall reserve my own for that venue. I shall continue an examination of the other allegations quoted by Rene Azurin later tonight or tomorrow.  THANKS for slogging through this if you have gotten this far.

2 comments:

DeanJorge Bocobo said...

I took great pains to produce a graphic that people can zoom in on to see the real detail in the data. It's pretty clear that on an individual basis, the percentage of the total votes garnered by each candidate at each stage of the canvas--whether they lost or not--was NOT CONSTANT. I shall present my analysis of each of the other allegations in the next three parts of this study.

Synthesist said...

Waiting with bated breath for 2-3-4.