This optimization tool takes previous data from a particular polling location as input and estimates the number of poll workers for check-in and voting booths needed to keep approximately 99% of voter wait times less than 30 minutes.
Arrivals: This data should be number of voters arriving per hour for each of the 13 hours (6AM to 7PM) that the polls are open.
Check-in times: Check-in times in the model represent how long it takes for voters to be checked in by a poll worker. This data must be measured in seconds for the simulation. These values should be measured from the moment a voter steps up to the counter until the moment the next voter steps up to the counter (if there is a line) or the original voter leaves the counter (if there is no line). If you do not have any data or an estimate on the mean and standard deviation of check-in times, a default option with Williamsburg check-in times from the November 2015 election is provided.
Voting times: Voting times in the model represents how long it takes people to vote at a voting booth. This data must be measured in seconds. These values should be measured from the moment a voter steps up to the voting booth and occupies it until the moment the next voter steps up to take the booth (if there is a line) or the original voter leaves the booth (if there is no line). If you do not have any data or an estimate on the mean and standard deviation of voting times, a default option based your polling location’s ballot length with Williamsburg voting times from the November 2015 and March 2016 election is provided.
NOTE: All files MUST either be plain text or have a .txt extension.