Using PB to detect fraud on mexican elections?

mike74 · Post by **mike74** » Tue Sep 12, 2006 12:30 am

ricardo wrote: Did you count how many ballots did we get on each case (not only how many cases with discrepancy, but the total of ballots on each discrepancy case)?

Yes. I got 3,127,090 for cases where there were less ballots than voters and 966,559 for cases where there were more.

I suspect that part of the problem is probably that the numbers of people that could (or should) have voted are somewhat inaccurate. The numbers of ballots provided to the voting locations may be inaccurate, also.

Dare · Post by **Dare** » Tue Sep 12, 2006 12:59 am

Ricardo, a couple of questions:

First, a discrepancy between the field headers and your fields as shown in first post:

Code: Select all

;1 ESTADO            State
;2 NOM_ESTADO        Name of the State
;3 DISTRITO          District
;4 SECCION           Section
;5 ID_CASILLA        ID of voting-place
;6 TIPO_CASILLA      Kinf of voting place
;7 EXT_CONTIGUA      exterior
;8 NUM_ACTA_IMPRESO             number of act
;9 NUM_BOLETAS_RECIBIDAS        *Number of voting tickets receiben on that voting-place
;10 NUM_BOLETAS_SOBRANTES       *Number of voting tickets that was Not used at the End of the day
;11 TOTAL_CIUDADANOS_VOTARON    *Total of citizens that vote at that place
;12 NUM_BOLETAS_DEPOSITADAS     *Number of voting tickets found in the ballot box
;13 PAN                         Party 1
;14 PAN_VCAP                    Party 2
;15 ALIANZA_POR_MEXICO          Party 3
;16 APM_VCAP                    Party 4
;17 POR_EL_BIEN_DE_TODOS        Party 5
;18 PBT_VCAP                    Party 6
;19 NUEVA_ALIANZA               Party 7
;20 NA_VCAP                     Party 8
;21 ALTERNATIVA_SOCIALDEMOCRATA Party 9
;22 ASDC_VCAP                   ------------------------------------ ???????
;23 NUM_VOTOS_CAN_NREG          *Number of votes canceled
;24 NUM_VOTOS_NULOS             *Number of nule votes
;25 TOTAL_VOTOS                 *Total of votes
;26 CASILLA                     Voting-place
;27 LISTA_NOMINAL               *Number of people allowed To vote at that voting place
;28 HORA_RECEPCION_CEDAT        Time of reception
;29 HORA_CAPTURA_CEDAT          time of capture
;30 HORA_REGISTRO               time of capture

Or my error? (Field 22). The tenth party?

Second, what exactly is the difference between a null vote and a cancelled vote?

Third, the actual meaning of these values:
TOTAL_CIUDADANOS_VOTARON - is this people who pitched up at the voting booth (got a ballot paper)?
NUM_BOLETAS_DEPOSITADAS - is this the number of ballot papers put in the box?
TOTAL_VOTOS - so what is this?
LISTA_NOMINAL - how is this significant?

Reason for the questions is it can make a significant difference to any sums done.

ricardo · Post by **ricardo** » Tue Sep 12, 2006 1:12 am

Dare wrote: Second, what exactly is the difference between a null vote and a cancelled vote?

Reason for the questions is it can make a significant difference to any sums done.

1.- Not sure the 22th

2.- Not sure, i think the nule vote is some people that auto nule his vote. Canceled is one that was canceled by authoritie because they don't understand what the pople vote.

Third, the actual meaning of these values:
TOTAL_CIUDADANOS_VOTARON - is this people who pitched up at the voting booth (got a ballot paper)?

Yes

NUM_BOLETAS_DEPOSITADAS - is this the number of ballot papers put in the box?

Yes

TOTAL_VOTOS - so what is this?

The numer of votes we get as result of aritmetical operation counting votes per each party. The result of all votes makes the TOTAL_VOTOS

LISTA_NOMINAL - how is this significant?

That the number of people on the list of each voting place.

In theory the number of received ballots should be the LISTA_NOMINAL plus 10 more ballots (in case need to replace one).

As you point, all this columns are important, because (for some reason that nobody understand, some claims fraud, other just don't say nothing) there are many (but MANY) differences.

Some can have some reasonable explanation, but some are hard to explain. per example: more ballots that citizens that get a ballot on that place!

My intention is try to make asome reserach made from people (like us) that has no other interest than some 'academic' one. Im very glad that all fo you are participating on this that can be also usefull. Its a real world exercise that is getting more interesting...

ricardo · Post by **ricardo** » Tue Sep 12, 2006 1:20 am

mike74 wrote:
ricardo wrote: Did you count how many ballots did we get on each case (not only how many cases with discrepancy, but the total of ballots on each discrepancy case)?
Yes. I got 3,127,090 for cases where there were less ballots than voters and 966,559 for cases where there were more.

I suspect that part of the problem is probably that the numbers of people that could (or should) have voted are somewhat inaccurate. The numbers of ballots provided to the voting locations may be inaccurate, also.

The same results that i get:

More ballots than difference (received - left): 966559 ballots
Less ballots than difference (received - left): 3127090 ballots

---

That number CAN'T be (by law) inaccurate. We are using the OFFICIAL database from the official Electoral Institute of Mexico.

The difference betwen the guy who win to the 2nd place is less than 250 k votes. So having big amounts of missed/added ballots puts the mexican election on a hard moment, where some party (the one that gets the 2nd place) claim that goverment must recount each vote and each counting place because they said that ATM this election can't be validated.

The problem is that the tribunal declare that election was okay and gice some party the winning paper a few days ago.

Mexico is atm on a political tsunami because of this.

We are not going to judge anything, we are just making some aritmetical exercise.

ricardo · Post by **ricardo** » Tue Sep 12, 2006 1:22 am

Dare

22th yes, its a party.

Pupil · Post by **Pupil** » Tue Sep 12, 2006 6:47 pm

Which party won? Might be a fun exercise to see if the extra votes ended up on this party's votes..

To check this, one could subtract all votes for each party, except the winning one, from the total votes, also subtract the blank votes and the cancelled votes. Then check this value against the discrepancy that you get with the tests that you've already implemented..

ricardo · Post by **ricardo** » Tue Sep 12, 2006 6:59 pm

Pupil wrote:Which party won? Might be a fun exercise to see if the extra votes ended up on this party's votes..

To check this, one could subtract all votes for each party, except the winning one, from the total votes, also subtract the blank votes and the cancelled votes. Then check this value against the discrepancy that you get with the tests that you've already implemented..

PAN won.

The main claim of the other parties is that this elecction was a mess, and then its impossible to determine who really won, because with such amount of ballots missed or added, it was an election that in some point loose confiability.

Thats what i was trying to determine: if this really is a mess.