Is Range Voting the Best System?

[TastyWheat]

Apparently someone made a site completely dedicated to proving that range voting is the best voting system under the sun. I looked at most of it and he makes a damn good case for it (comparing it to other systems). My state GOP is opposed to IRV (it's in their platform of all things) so they're probably just as opposed to this.

http://rangevoting.org/

[sailor]

It can`t be the best. It stil includes voting.

[Truth Warrior]

Go out on the range and just yell who you want. :D

[fgd]

Technically, I think that approval voting is simpler to implement. Range voting has some hacks like "getting above X% responses" on a candidate to be considered for the final tally. Approval voting gets very similar results to range voting, and I think is less susceptible to "gaming".

[slothman]

I agree the RV is the best.
The problem is that it is hard to implement and teach voters about how to use it.
Approval is much easier to use, teach, and set up machines with it.

[Truth Warrior]

I agree the RV is the best.
The problem is that it is hard to implement and teach voters about how to use it.
Approval is much easier to use, teach, and set up machines with it. Just leave those pesky "details" to the Diebold programmers. They can handle it. ;)

[tmosley]

I always liked range voting, although I knew it by a slightly different name (rank voting--similar but not exactly the same).

[TastyWheat]

Technically, I think that approval voting is simpler to implement. Range voting has some hacks like "getting above X% responses" on a candidate to be considered for the final tally. Approval voting gets very similar results to range voting, and I think is less susceptible to "gaming".
See: Approval vs Range Voting (http://rangevoting.org/AppExec.html)

Simpler? Yes, but not as effective. They've also shown (in a real world example) that approval voting is very likely to revert to plurality voting.

[brokenladder]

I always liked range voting, although I knew it by a slightly different name (rank voting--similar but not exactly the same).

Score voting (aka range voting) is a cardinal voting method. Ranked methods are called "ordinal". There are numerous ordinal methods, such as the Condorcet methods, Instant Runoff Voting, and Borda.

Score voting is superior to all ranked methods (http://scorevoting.net/BestVrange.html) - even ones which have never been invented.

[brokenladder]

Technically, I think that approval voting is simpler to implement. Range voting has some hacks like "getting above X% responses" on a candidate to be considered for the final tally. Approval voting gets very similar results to range voting, and I think is less susceptible to "gaming".

Well that "less susceptible to gaming" bit is highly misleading. Here's a strange but poignant analogy.

Say you have to go on an arctic expedition with either coat X or coat Y. They are identical in all respects except for their insulation value. Coat X has an insulation value of 10 when it's dry, and 7 when it's wet. Coat Y has an insulation value of 20 when it's dry, and 11 when it's wet.

Now suppose someone tells you that you shouldn't pick coat Y, because "it's more susceptible to getting wet".

Here are some Bayesian regret figures (http://scorevoting.net/ShExpRes.html) showing the robust superiority of score voting to approval voting.

[brokenladder]

I agree the RV is the best.
The problem is that it is hard to implement and teach voters about how to use it.
Approval is much easier to use, teach, and set up machines with it.

That's basically nonsense I think. There's lots of evidence that people actually find score voting simpler to use. For instance, the web site HotOrNot.com may have been a binary "hot"/"not" vote (approval voting) were it not for the observation that making such a choice tended to belabor the decision by users, whereas the need for rapid clicking (to gather as much data as possible as quickly as possible) led the designers to use score voting. People have an easier time scoring than making an all-the-way-yes or all-the-way-no decision.

In my own score voting experiment (http://scorevoting.org/Beaumont.html) in Beaumont, Texas, I found that the local yokels had no problem scoring the five gubernatorial candidates in 2006. Not a single person asked for explanation or seemed to misunderstand the idea.

Approval voting is simpler than score voting, and it is the only other method that I think is of acceptable quality - but score voting is significantly better according to the best known Bayesian regret figures.

[brokenladder]

See: Approval vs Range Voting (http://rangevoting.org/AppExec.html)

Simpler? Yes, but not as effective. They've also shown (in a real world example) that approval voting is very likely to revert to plurality voting.

This is absolute total nonsense (http://groups.google.com/group/scorevoting/web/degrade-plurality), as illustrated by the fact that 90% of Y2K voters who claimed to prefer Nader actually voted for someone else.

Approval voting was used in contentious elections at Dartmouth and San Francisco State University, and members of both institutions related to me that they saw a large percentage of ballots with votes for more than one candidate.

This "approval degrades to plurality" myth is put out by disingenuous hacks like the members of FairVote.