At Mon, 1 Apr 2013 12:01:13 -0400,
David J. Fiander wrote:
> So, I just voted for the Code4Lib 2014 location. There are two possible
> venues, and I was given three points to apportion however I wish.
> While having multiple votes, to spread around at will, makes a lot of
> sense, shouldn't the number of votes each elector is granted be limited
> to max(3, count(options)-1)? That is, when voting for a binary, I get
> one vote, when voting on a choice of three items, I get two votes, and
> for anything more than three choices, I get three votes?
> I mean, realistically, one could give one vote to Austin and two votes
> to Raleigh, but why bother?
You actually can vote 0-3 on any option, for as many total votes as
The optimal strategy, assuming that you actually prefer one option to
another, is to vote 3 for the option you prefer and 0 for all others.
To slightly change the subject, systems are a policy decision, not a
technical problem. In the case of voting for presentations (more
important to me that conference location), different voting systems
will generate a different mix of presentations. Think of the
difference between the American congress and a parliamentary system.
The question is, does code4lib want conference presentations that are
more “first past the post”  or more representative of the diversity
of interests of the code4lib crowd (like a parliamentary system). The
existing system reduces to a first past the post system, which means
that the presentations which more people prefer win, rather than
presentations that a smaller group of people might feel strongly
This is a question that shouldn’t be decided by the technology; the
policy should decide the technology. A google form might work, and
certainly hand-counted emailed votes would, given the relative
smallness of the c4l community.
Those who are interested can read more here:
Sent from my free software system <http://fsf.org/>.