At 09:59 AM 12/31/2009, Bruce Gilson wrote:
>On Wed, Dec 30, 2009 at 11:25 AM, Abd ul-Rahman Lomax
> > ['...]Note that much of the damage done in Instant
> > Runoff Voting is due to eliminations.
>For once, a comment with which I agree. The problem with IRV seems to be
>because a good candidate, with a lot of second-preference support but not
>much first-preference support, gets eliminarted early. This is why I tend to
>think that Bucklin would solve most of IRV's defects.
And we agree on that, as well. Basic slogan of mine:
Count All the Votes.
That's exactly what IRV does not do.
Bucklin, once the smoke clears, and with enough candidates (where the
winner really isn't obvious, it's a real contest), unless a majority
candidate appears sooner, does count all the votes, making it quite
equivalent to Approval. But it is likely that it will encourage more
appropriate addition of lower-preference approvals, by allowing them
to be expressed that way, as lower preferences.
The inability of pure Approval to do this is the major objection
raised; Bucklin addressed this a century ago, before the
strategy-free characteristics of Approval Voting were known, before
Approval Voting had a name. The claim that Approval has never been
used for a modern public election is simply false; the many Bucklin
elections were really Approval elections, in substance.
There is a history, and we should know it.