Voting Mechanics

Voting Mechanics is concerned with the polling system itself as opposed to the methodology used in selecting between various choices. There is a certain irrelvancy as to the method of selection (IRV, Ranked Choice, Condorset, or whatever) when one focuses instead on the integrity of the actual system employed to gather and count the ballots. And these "mechanisms" are the subject of this article. Stalin said "It doesn't matter who votes. It matters who counts the votes." And that is the best introduction we can have concerning the subject of this article and, perhaps further articles concerned with specifying an incorruptible mechanism for ascertaining the will of the people.

Counting the votes
Computer "touch screen" voting has soured every rational and reasonable person to the idea of computerized voting. And that is a justifiable response to a system that registers the voter's choice and counts all the choices in a single authoritative black box. That will not ever be acceptable. Such a mschanism is "Stalanist" because it is authoritarian and all authority can be and often is corrupted. An incorrupible mechanism for polling is therefore as free from centralized authority as possible.

In an incorruptible polling system the "votes" (ballots) are housed in many all of which contain the same data/ballots. All persons have read only internet access to these databases such that any person can "count the votes". There is no reason to lock the ballots away in a safe where only the "authorities" can see them. This is an advantage of technology. It simply is not possible for everyone to count the votes if the votes only exist in paper form. In current systems all the people can't count the votes because they are currently paper and/or they are currently in a machine that only the "auhorities" can access.

To make this clear let us assume paper ballots just like we have right now. These are scanned and an electronic representation is entered into the ballot database(s). Counting the ballots (votes) is a trivial matter and is done by anyone who cares to do so. A lot of people will "roll their own" software to do this and several versions from "trusted" sources will also be available. Perhaps you would trust the local Boy Scouts or the local church, or the Young Republicans or some other organization to provide you with vote counting software. Such software (even in the case of multiple selection balloting) is trivial. There is no reason to believe that browser "plugins" would not be prolific. At this point the role of the voting authority is only to insure that the paper ballots and the database ballots are identical. When all persons are satisfied as to their counting (try several different plugins if you want) then the results of the poll are absoultely known by all persons who care to know. There will be no charges of counting fraud or "hanging chads" or judgements about "voter intent".

Puting Ballots into the Box
Let us again assume paper ballots in our explanation of the proposed voting system. These paper ballots must be scanned so as to convert the images of the ballots to electronic form. Then the data is transmitted to the database(s). Right away we have some serious problems in that the scanner used for this purpose can be rigged and corrupted. And if the data (the electronic form of the ballots) is actually broadcast over the ariwaves or over the internet and received by the databases (i.e. the database systems are "receivers" always listening and recording the "votes" as scanned and broadcast), then it is possible to introduce counterfeit votes and/or to pick ballots out of the internet stream or out of the air, alter them, and reinsert them in the stream. The iommediate reaction is to "secure" this connection between the scanning system (the scanning authority) and the database system(s). A secure encrypted connection. This will not be the solution becase it is an authoritative solution.

The Ballots themselves
(1) We must be sure that the ballots (and hence the data in the databases) reflect real people that are to be included in the poll, and that only these real people are included in the poll. (2) We must be sure that the ballots (and hence the dats in the databases) are not counterfiet ballots substituted for the real ballots. (3) Every Voter should be able to "see" his/her ballot in the database(s).

At present we "register" to vote. That puts us in a particular precinct such that we get a correct ballot that reflects our geographic location and the people running for office in that location. The process also, supposedly, insures that only registered voters get ballots and that they only get one ballot each. Imagine this process to be unchanged and that you receive your ballot in the mail as most of us currently do. The alteration in this part pf the system is only that the ballot has a place for the voter to affix a 10 digit number which will be his/her voter ID known only to the individual voter. There will be collisions even though the typical precinct is less than 10000 and the 10 digits can express 10,000,000,000 different numbers. Those who use the number 0123456789 will probably collide with the other morons that use 0123456789 and the same for the idiots that would use 252525... or 55555... or other stupidities. Such ballots will not be lost or discarded but they will be referred to as idiot ballots and if you are an idiot one of the ballots you find in the database with your ridiculous number will be YOUR ballot. It will be interesting to see how many of them are Republican and how many are Democrat. The purpose of this is to allow the rational voters to actually SEE their ballot in the ballot-box (database), but it also yields a check on integrity. The number of collisions should be very, very small once the idiot numbers are taken into consideration. If the collisions are high then the election is suspect.

Certifying the ballot box
For the nonce we will assume that there are paper ballots because that is what most people are comfortable with. I assure you that you will not be comfortable with them when we see how easily they can be corrupted. But that comes later. So the immediate problems with the "everyone counts the votes" meachanism are: "How do we know that the ballot database is a true represntation of the paper ballots?" and "What is to prevent hackers and private interests from manipulating the data?". Sucinctly: There must be multiple scanners and multiple databases controlled by totally different administrators. And every official ballot (after it is cast and separated from the voter(anonimity)) must be serialized. That is correct. There must be (at this point) an authority (actually there are several) that stamps each cast ballot with a unique identity. Each scanner also has a serial number unique to the particular scanner. Any given ballot (identified by serial number) must be read by no less than 3 different scanners administered by 3 different scanner administrators. If paper ballots are used then I see a big yellow bus that runs between no less than 3 precints where people who would have counted the votes will ride the bus delivering the ballots from one polling place to another, around and around we go. If by mail the same scenario occurs: The receiving authority at each mail drop (just like the receiving authority in any precinct) assigns a serial number to any and all ballots that are to be placed in the ballot box/database. The serial number cannot be traced back to the individual voter just as the ballots cannot be traced back to an individual voter. The ballot is scanned and the balot information along with the scanner serial identity is broadcast to the various database systems. The ballots are then pysically taken to the next mail drop or pricinct by a big yellow bus full of the people that would ordinarily count the votes (if you would have trusted them to count the votes you can surely trust them to gaurd the ballots). The people insure that the ballots do not get lost and that extra ballots are not put into this "official" batch of ballots as the batch is hand delivered to the next pricinct/scanner. A second scanner serial number is affixed to the data as the document is scanned again and the data broadcast to the database(s). At least 3 scaners must be used and the database systems must insure that the document has been received at least 3 times from at least 3 different scanning authorities and then the ballot data is placed in the database(s).

dissimilar databases
There must be multiple types of scanners and multiple databases in any event such that no central authority/vendor can be corrupted. As a matter of fact, when you look at the ballot box over the internet you will be looking at one of many different ballot databases all of which contain all the ballots and all of which are identical with regard to the ballots themselves. The voting "authorities" (who have no more authority than you or I) insure that the contents of all the databases are identical. Most likely the database you are looking at will be the one nearest you. It doesn't matter because they all contain the same data. Each database system is administered by a completely different database administrator organization. The different organizations will use one of several different types of computer hardware (e.g.Apple, McKintosh, Amd, Intel, and others) these systems will run various operating systems (Linux, MS Dos, Windows Xp, Sun OS, Free BSD, etc.), and various types of database software (Oracle, Sybase, PostgresSQL, MySQL, mSQL, dbm, Informix, and of course ye ole "roll your own"). They will present the information using the various types of "web" servers (Apache, Sun Java Web Server, Abyss, Xserve, etc) with applications coded in python, php, perl, apt, or whatever. It is virtually impossible to "break in" to the database and "mess with" the ballots or the deleivery system because there are many of them and they are all different. The attacking party would need to compromise many different computer systems all being administered by different organizations/people. The cost of one complete system to perform the database function is less than $500. It is not a major investment and anyone with a decent desktop computer system (and that is a lot of people) has sufficient computer hardware to do this data storage task.

I gotta go to sleep. More tomorrow. --TheTrucker 19:43, 26 July 2007 (PDT)