Charlie Crist and others have asserted publicly that fixing Florida’s voting problems is not rocket science. We think that they have not given the matter sufficient thought.
Central to both is the study and control of certain critical rates. For both the issue of control is very complicated. Indeed, for the administration of voting by the populace at large, it appears that one needs to mix in serious amounts of Queuing Theory where throughput, a key rate, arises.
Crist and others holding similar opinions regarding the ease of the solution to the problem might do well to Google “Queuing Theory” and “Queuing Theory and Voting” and follow a number of the links found there to begin to get a clearer picture of the complexity of the problem, mathematically and otherwise. Politics merely add very irritating knots to the mess.
Regarding the situation in Broward County. Amy Sherman’s article in the Miami Herald on Dec. 13, State officials review election problems, was particularly revealing. There we, the public, finally obtained an indirect glimpse at throughput data. What the data cited in the article tell us is that however the staff of the election supervisor, Brenda Snipes, estimated the potential voting rates for early voting sites, it was very seriously flawed.
They asserted that they expected early voting sites to be able to handle between 400 and 500 votes per hour. And what was the basis for that estimate? We do not know, but the actual best average rates quoted were roughly half of the lower expected value quoted. What is singularly annoying here is the interpretation given by the staff. Their comments seem to indicate a failure to understand that the problem lay in the modeling with the given layout, not the crowds.
One would expect for the throughput of the relevant queues to reach their true maximums for their given configurations under heavy demand. The real problem is that it is the modeling and accompanying estimates that control the requests for and configuration of resources. Improvement is needed.
David L. Ritter