A Cappella and Voting

As Evan mentioned in an earlier post, a cappella groups tend to vote on everything, so I am just going to focus on how my group handles solo auditions.  My group has a first round (like a primary) where each member auditioning sings a portion of the song, followed by a blind vote.  During this round, each of the members (including the candidates themselves) can vote up to 3 times and can only vote once for each candidate.  The three candidates (or more in the case of a tie for third place) with the most votes move on to the second round, where they sing again, and then the group, minus the candidates, re-votes with one vote a piece.  In this round, a simple majority is required, and you can also vote as an “undecided” to hear further conversation about each of the candidates, given that a simple majority has not already been reached.

In analyzing our system with two of the three desirable properties of voting systems, areas to improve upon become apparent, particularly in the primary.  You are allowed to vote for yourself or vote fewer than three times, which incentivizes strategic behavior in the first round.  We also do not take strict advantage of ordering information, considering that the votes for your second and third favorite candidates (if used) are weighted equally to your first candidate.  One potential improvement would be using a Borda count in the first round to make members’ first choice votes count more than their third; thereby facilitating preference revelation.  Unfortunately, taking the time to tally up these differently weighted votes would be costly, especially in the wake of our fall concert.  In fact, when considering the increased decision costs associated with using a new system, a move towards this alternative kind of system seems to be more of an appealing way to reduce decision costs and save time.  Even though it might look a little bit different than this in practice, having learned a thing or two from Curly, Larry, and Moe, it might be in the our group’s long term interests to leave the currently functional system in place (even if it is a little imperfect), because it minimizes the total of external costs and decision costs better than I would as the dictator.