Score Voting a subset of Range Voting?

[Steve Cobb]

Has it been previously noted and discussed that Score Voting is a subset of Range Voting, so that the two terms are not equivalent?

The ranges of both 321 and Majority Judgment have named values that are never converted to points.

If my understanding is correct that 321 and Majority Judgment are forms of Range Voting, the Wikipedia article

https://en.wikipedia.org/wiki/Range_voting

is incorrect:

"Range voting or score voting^[1][2] is an electoral system for single-seat elections, in which voters give each candidate a score, the scores are added (or, equivalently, averaged),^[3][4] and the candidate with the highest total is elected."

One way or the other, the article needs fixing.

I noticed this because Consensus Weighted Range Voting would better use Score instead of Range, being a truly numeric method.

[Andy Jennings]

Are you saying anything purely numeric would be "score" and anything evaluational could be called "range"?

I feel like the opposite would be more appropriate.  "Excellent, good, fair, poor" could be "scores", but "range" can used as a synonym for "interval".

https://en.wikipedia.org/wiki/Range

I think this is the sense in which Warren originally used it.

--
You received this message because you are subscribed to the Google Groups "The Center for Election Science" group.
To unsubscribe from this group and stop receiving emails from it, send an email to electionscience+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Steve Cobb]

You might be right about range, but I don't know anything about its origins in range voting.

I think most people would assume that scores are numeric, but I am not confident of that.

But it seems to me that there should be a clear difference between methods using normal mathematical operators on the voters' numeric scores and those that count potentially non-numeric values.

BTW, while I have begun referring to this class of voting methods as "evaluational", I noticed in the Range Voting article that someone has previously used "evaluative" (as opposed to preferential and what I've been calling allocational). 

On Saturday, September 2, 2017 at 8:34:38 PM UTC+2, Andrew Jennings wrote:

Are you saying anything purely numeric would be "score" and anything evaluational could be called "range"?

I feel like the opposite would be more appropriate.  "Excellent, good, fair, poor" could be "scores", but "range" can used as a synonym for "interval".

https://en.wikipedia.org/wiki/Range

I think this is the sense in which Warren originally used it.

[Brian Olson]

As far as I can tell 'Score, Range, Rating' and probably a bunch of other terms are equivalent.

There's probably some minor public education/public relations subtle play that should be market tested and focus grouped for better advocacy, but I think mathematically they're all equivalent.

'Ranking' is right next door and my software that works natively on ratings just converts a ranking by using 'N - r' where N is the number of candidates and r is some ranking of a candidate on a ballot. Ranking where equal-rank is allowed is thus mathematically equivalent to rating/score/range.

If you want to put labels on the rating that are A,B,C,D,E,F or 5,4,3,2,1 or 100.0..-100.0 they're all the same except for quantization effects like when someone really wants to have more than 5 shades of grey but they're given a 5 value user interface.

--

[Steve Cobb]

They certainly have something in common, but I would not necessarily call them “equivalent”. There may be a useful and important distinction, e.g. I would define “rating” more broadly (as a category), and Score Voting more narrowly (as a specific method). To me, the primary categorization of voting methods is the expression:

-Allocative: Voters divide a fixed amount of vote, whether continuously divisible or in discrete tokens, among candidates. These methods are thousands of years old, but would include modern asset methods like Liquid Democracy and GOLD.

-Preferential: Voters rank candidates (equivalent to comparing them pairwise, so maybe Comparative)

-Evaluative: Voters rate candidates *independently* on a scale (possibly non-numeric)

Those expressions may be counted, converted into scores (e.g. Borda) and summed, or tallied in some other way. When evaluative expressions are scores that are totaled or averaged, we have Score Voting. This is the same as the Wikipedia definition of Range Voting. But there are other rating methods that never total.

With that understanding, I would prefer to call SRV/STAR Score-Preferential Voting, but it would be a stretch to call 321 any form of Score Voting: 321’s values are counted, but not really summed.

[William Waugh]

Regarding https://groups.google.com/forum/?fromgroups=#!topic/electionscience/5tgDUb6eOQU

Yes, people often use "range" to mean "interval". But in set theory, we speak of the range and domain of a function, and in these terms, they can be any set, not necessarily a set of numbers.

For a rating that resembles a score but isn't numeric, I had rather use grade.