Normalizing the Voting for Recalling City Council Members

In purely technical and in extremely general terms, normalization is the process of eliminating duplicate data. Though duplicate data is not always “bad”, knowing that there are duplicate data helps the informatics specialist derive valuable information.

So, how does this explanation apply to the on-line voting by the citizens interested in Marco Island as related to recalling the city council members? Is there “duplicate” data in the voting? Has the voting been compromised?

The answer to the last question is NO. The voting has not been compromised. Safeguards are now in place preventing the miscreant(s) that voted approximately 3,500 times to remove Mr. Kieser and only Mr. Kiester from performing another such dishonest deed.

Is there duplicate data? Technically speaking, no.

But here is the rub. A household with one computer can have several legitimate voters. As an example, one household can have a wife, a husband, a child over 18, the husband’s mother-in-law, and a dog. In this case there are five legitimate voters (only kidding – there are only 4 – the mother-in-law does not count). Hence “one” computer can vote several times – and the votes be sincere and of different sources (different voters).

However, and in another example, one household can only have one voter but yet vote many times, say five times.

So how do informatics experts distinguish the first five (real) votes from the second five (four are fake) votes?

How? Look carefully at the fifth word in the above sentence – EXPERTS. Informatics experts have at their disposal tried, tested and proven algorithms by which to detect with a great degree of certainty these “fake” votes.

Detecting these “fake” votes, and taking them out of the calculations, is in a very general sense, normalization.

Hence, the new graph on the recall voting site reflects these algorithms.

Why were these algorithms not implemented from day one? Because, in part, it takes analysis of the data to design the proper algorithm by which to account for the problematic data. Hence you need something to study before you figure out how to fix it.

What exactly is the logic to the algorithms? This will not be revealed. Why? Because we want to continue trapping these “fake” votes. Knowing who is attempting to cheat, and in what context, is information that is exceedingly valuable – as priceless as knowing what the real votes are.

Don’t you find it valuable to know that the person(s) that went to such lengths to write a robot to flood the votes did so to recall one and only one council member? I do.

Thank you for your interest and please continue encouraging the voting! But only American style voting (vote early, vote once) and not Latin America voting (vote early, vote often).