[EM] FPP-Approval hybrid

[Chris Benham]

Oops!  The definition I typed wasn't exactly what I meant.   The 
fixed-up version:

I have an idea for simple method that I like much better than plain FPP 
or Approval, and that voters accustomed to FPP might like.

*Voters indicate a single favourite candidate, and also may approve as 
many other candidates as they wish.

If the FPP winner F's FPP score is higher than than the highest number 
of approvals of any candidate on ballots that don't have F as a 
favourite, then F wins. Otherwise the most approved candidate wins,*

Obviously candidates indicated as a "favourite" are also approved. The 
FPP winner is the candidate that is voted as unique favourite on the 
highest number of ballots.

This is my compromise idea for people who like FPP and are not turned on 
by pure Approval or any decent ranking method. I rate it as much worse 
than Hare, but some might not agree.

It meets Participation.  Unlike STAR, it meets "Second-place Favourite 
Betrayal", meaning that the voter never has any incentive to vote their 
sincere favourite below equal-second. Also the voter's incentive to 
Compromise regarding their single Favorite vote is quite a bit weaker 
than with plain FPP.  If the voter's unacceptable Greater Evil is among 
the front-runners, then probably just approving all the other 
front-runners and giving your single vote to your sincere favourite who 
is not a front runner will be just as effective in stopping Greater Evil 
as insincerely giving your single vote to Lesser Evil front runner.

It meets Irrelevant Ballot Independence.

My tentative name suggestion:  "Approval-enhanced First Preference 
Plurality".

With just three candidates I think it gives Condorcet Winning Votes like 
results.

49 A
24 B
27 C>B

In normal FPP (and some other methods that elect A) the C voters might 
regret not voting for B.  But here on ballots that don't have A as a 
favourite B's approval score is 51, higher than FPP winner A's FPP score 
of 49.

So we don't elect the FPP winner A and instead elect the most approved 
candidate B.

Chris B.

On 3/08/2024 5:21 am, Chris Benham wrote:
>
> I have an idea for simple method that I like much better than plain 
> FPP or Approval, and that voters accustomed to FPP might like.
>
> *Voters indicate a single favourite candidate, and also may approve as 
> many other candidates as they wish.
>
> If the FPP winner F's FPP score is higher than F's maximum 
> approval-opposition score, then F wins.  Otherwise  the most approved 
> candidate A wins.*
>
> Obviously candidates indicated as a "favourite" are also approved. The 
> FPP winner is the candidate that is voted as unique favourite on the 
> highest number of ballots. That candidate F's "maximum 
> approval-opposition score" is the highest approval score of any other 
> candidate on ballots that don't approve F.
>
> This is my compromise idea for people who like FPP and are not turned 
> on by pure Approval or any decent ranking method. I rate it as much 
> worse than Hare, but some might not agree.
>
> It meets Participation.  Unlike STAR, it meets "Second-place Favourite 
> Betrayal", meaning that the voter never has any incentive to vote 
> their sincere favourite below equal-second. Also the voter's incentive 
> to Compromise regarding their single Favorite vote is quite a bit 
> weaker than with plain FPP.  If the voter's unacceptable Greater Evil 
> is among the front-runners, then probably just approving all the other 
> front-runners and giving your single vote to your sincere favourite 
> who is not a front runner will be just as effective in stopping 
> Greater Evil as insincerely giving your single vote to Lesser Evil 
> front runner.
>
> It meets Irrelevant Ballot Independence.
>
> My tentative name suggestion:  "Approval-enhanced First Preference 
> Plurality".
>
> Chris B.
>
>
>