[EM] Voting in small committees : a live test

[Hugo Harth]

Voting and counting ballots in small committees may be very different from 
well organized elections.
Typically, ballots have to be counted quickly, there is noise of the people 
around ....

I did a little test under such circumstances, television-news on and some of 
my family members trying to interrupt me from time to time.

The test itself : there where 16 voters (ballots, as sheets of paper and 
randomly generated orders) with 5 candidates all ranked and only one 
candidate per rank. The task was to elect 1 chairman or alternative out of 
these 5.

Tried first STV (quotum of 50%) : task was performed in 1 minute 45 seconds. 
There was one winner.

Second, tried Ranked Pairs : only to tally all pairs in a matrix : 7 minutes 
40 seconds were required.
When I tried to find a winning order, I was plagued by several ties that I 
forgot to note the time. Finally there were two winners and there was a 
Smith-set of size 2, it was necessary to look deeper to select a winner.

What I believe is, at least for small committees :

1) Methods that require to count pairwise victories require too much time 
and may be error prone.
2) Be prepared (and train yourself first) to handle ties, equal heaps in STV 
and whatever simple but surprising exceptions because they will occur (most 
likely in small committees)!
3) Ranked pairs (or another method that uses pairwise comparisons) for 3 
(perhaps 4) candidates may be feasible by handraising vote but I would 
recommend only to accept yes or no for every pairwise comparison (not to 
double the number of pairs)

yours sincerely,

Hugo Harth
_________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.