[EM] Scoring (was Re: OpenSTV 2.1.0 released)

[Juho Laatu]

On 20.9.2012, at 8.20, Michael Ossipoff wrote:

> You said:
> 
> The idea that there are third candidates but that are never elected,
> and that can act as spoilers  does not fly very well.
> 
> [endquote]
> 
> In what sense doesn't it "fly well"? What does that mean?

I just meant that it is a waste of effort and energy to have "fake" candidates that appear as they could be elected, but they can't, or whose presence may make the "natural" winner not win.

> Let me try to translate what you said:
> 
> "The idea that there are non-Republocrat candidates that have never
> been elected to the presidency, and that an act as spoilers is either
> unsatisfactory to Juho, or disbelieved by Juho."
> 
> Is that what you meant?

Approximately so.

> 1. We don't want nonwinning candidates to be spoilers. That's why we
> don't want Plurality, IRV, or unimproved Condorcet.

It is hard to find methods that have no weaknesses. Luckily we can often use methods whose weaknesses are weak enough. All those three methods may meet that target in some elections.

> 2. It's unsatisfactory (to the public) that only Democrats and
> Republicans ever win, because the public regard the (Democrat and
> Republican) politicians as sharing the same moral level as a
> schoolground drug-dealer.

Depends on if they want "that only Democrats and Republicans ever win".

> But I'm not sure that I've interpreted you correctly. Maybe you meant
> that what "doesn't fly" is the belief that the winner must be a
> Democrat or a Republican. Yes indeed, that doesn't fly.

I meant that using a method that appears to elect any of the candidates but in reality can elect only certain candidates does not look natural.

> So, you see, we are in complete agreement.

Pretty much so. But I didn't want to take position on two-party systems vs. multi-party systems.

> You said:
> 
> If we want to have a two-party system...
> 
> [endquote]
> 
> Whoa. Who is this "we" who want to have a 2-party system. Are you
> saying that you want Finland to have a 2-party system? Or that you
> want the U.S. to have a 2-party system, and that I also do?

That was passive. So "we" could be anybody who wants that.

> But it isn't for me to say how many parties we have, or how many
> parties should sometimes win. That's for the voters to decide.

In a representative democracy perople elect representatives who will then decide. In this case the representatives of the two main parties will decide. This is a general problem of political systems (not only a problem of two-party systems) and also other organizations. Those that are in power have tendency to maintain their own strong position.

> To say that there should be some particular number of parties (like
> two) would be undemocratic.

Unless the voters or their representatives say so. (Or maybe someone just uses his freedom of speech and freedom of opinion.)

> So I'm not denying that some _do_ want a 2-party system. For instance,
> I'll venture a guess that the Democrats and Republicans like there to
> be a two party system.

That may well be true.

> In fact, I'll go farther than that, and suggest that maybe the wealthy
> types who own and bribe the Republocrat politicians also like there to
> be a two party system, in which the people owned by them are perceived
> as "the two choices".

Yes, there may be also such people. But maybe not very many. Maybe there are also many people that make use of the situation but that don't have any philosophical thoughts on what political system would be good. They are just opportunistic, with not much interest in politics nor in theories on two-party systems, Duverger's law etc.

> But seriously, the number of parties should be as many as people want.

I note that also multi-party systems have similar problems. They may have e.g. cutoffs that give no seats to parties below 5% support. Or the size of districts may set some limits (from this point of view a one-member distric is an extreme case). A more general form of the question could be e.g. if in a N member representative body a party with M/N of the votes should get M seats.

> You said:
> 
> , there are also better election methods for that purpose than plain plurality.
> 
> [endquote]
> 
> 
> Apparently, then, you disagree with Riker. I'll take Riker's side on
> that question: Plurality is perfect for making, maintaining,
> preserving a two-party system.

Maybe he should have said "one good" instead of "perfect". Or maybe word "perfect" means that he is very happy with having "fake" candidates that can collect the protest votes without making any harm.

What I meant with "better" was that some methods could allow new parties to rise and replace one of the old two parties, or that would allow those two parties to be internally more responsive to voter opinions than they (maybe) are today.

> Oh don't get me wrong. I'm not saying
> that Plurality can do it alone. No, it needs a little help. It needs
> the help of a mass media system that continually hammers home the
> message about "the two choices", and always reports campaigns and
> elections as if there were genuinely only two parties.

I think the mass media strengthens the two-party system, but the two-party system might survive even if the mass media would request the voters to consider also other parties.

> But still, for preserving a two party system, no voting system does
> the job as well as Plurality.

Maybe a system that would allow only those two parties :-). We have had also systems that allow only one party :-).

> If you think you can come up with a better one, then let's hear it.
> And don't forget to write to tell Riker.

I might have some proposals, but I'd need to know what the target of the reform should be. If the target was to maintain the two-party system, then maybe a system that would not allow other parties - except that it is possible that such a system would lead sooner to a revolution than a system that allows "fake" parties to run and collect the votes of the protesters but paractically never get representatives :-).

> You said:
> 
> Or one approach could be also to have only two parties and two
> candidates (=> meets FBC).
> 
> [endquote]
> 
> None can deny it. "Having" only two parties and two candidates could
> definitely enforce a two party system. What part of the Soviet Union
> did you say you were from?

Almost part, had to fight :-).

>>> As I said, I used to say what you said above. That was before I
>>> observed a progressive lesser-evil Democrat-voter voting in a
>>> Condorcet Internet poll, for a presidential election. Yes, I've
>>> mentioned this before.
>>> 
>>> She's a progressive, and preferred the policies of Nader to those of
>>> the Democrats. But she felt that Nader couldn't win, and that, because
>>> only a Democrat can beat the Republicans, the one and only goal is to
>>> maximize the probability of a Democrat winning instead of a
>>> Republican. We've been over this.
>>> 
>>> So she ranked all of the Democrats over Nader. I couldn't tell her
>>> that she needn't do that, because it was optimal strategy, given her
>>> assumptions and her goal.
>>> 
>>> You see, that's what you're missing. It's what you were missing before, too.
>> 
>> But wasn't that more a problem of the voter than a problem of the method?
> 
> Most certainly. The method couldn't care less. It was a problem for
> the voter only.
> 
> The problem is this: When it's a problem to the voter, then it's a problem. No?
> 
> You see, that's the whole problem: It's a problem to the voter.

A good system would have a good method and a good understanding of the method (and politics in general).

> I freely admit that it isn't a problem to Juho. But then you, Juho,
> don't live here, and so why should you consider it a problem?

I try to speak at a theoretical level (not to take position on what different countries should do).

> Given voters who share her preferences, predictive beliefs and goal
> priorities, those voters will favorite-bury. That you wouldn't is
> quite irrelevant. If lots of people do, then favorite-burial is
> affecting the election result, is it not?

The opinion of other people does influence on what people do. Some methods might even reinforce this behaviour. I believe, in most methods the method specific "bad group behaviour reinforcing" factor is not very strong.

> So what are you trying to say?  ...that we don't need an FBC voting
> system--We just need a different electorate?  :-)

Originally I was saying that good methods should meet good criteria "well enough". We also need good electorate, which could mean continuous education and encouragement (by media, country and fellow citizens).

> Her predictive beliefs were wrong, based
> on fallacious assumptions based on televised misinformation.

Maybe combined with one's own mistaken beliefs. I guess there are always some risks. For example in ranked methods people might keep some rating style thinking and therefore e.g. rank the strongest competitor of their favourite candidate last. Also ranking a strong, only barely acceptable candidate first may fall in this category. ("first position = lots of points, last position = as few points as possible")

> But, if you're saying that you want to tell all Americans that the
> Republocrats _aren't_ "the two choices", and that the Democrat
> _shouldn't_ be regarded as acceptable, then I encourage you and
> applaud you.

I try to avoid telling Americans what they should do. But I try to discuss especially with EM experts on how different political and electoral systems might impact societies (their societies or societies in general).

> But
> the funny thing is that the same people who say that the Democrat is
> acceptable, don't seem to agree with hir policies in office.

I can guarantee that people are often dissatisfied with the policies of the parties that are currently in power, and their own parties also in multi-party systems :-).

> It isn't for me to tell you what's acceptable. But when it wins, then
> _accept it_ and don't complain.

Or accept and influence (sometimes by complaining but maybe more often by proposing).

> But FBC complying methods, by not giving favorite-burial incentive,
> don't cause one person's notion of un-favorite acceptability to
> disgust and revolt someone else. That's because, at least, people
> won't be voting un-favorite "acceptability" over their favorites.

But I think you just proved that people can bury also irrationally :-).

> In ICT or Approval, I could say: "The Democrat is acceptable? That's
> nice, but at least you aren't voting that "acceptable" over your
> favorite."

But what if I want to vote in Approval so that I approve only the "acceptable" that I really want to win? ;-)

Btw, what do you think of SFBC for ranked systems? (StrongFBC = no need to rank your favourite below or equal with the less liked candidates)

>> I tend to think that all common Condorcet methods are likely to perform quite well in typical large > public elections
> 
> And isn't that what EM is really all about? Assertion of unsupported
> opinions like that?

What more would you expect in addition to my belief that they would perform quite well also in most political elections in addition to the less competitive elections in which they are currently used?

> So you, living a long way from here, know more about the psychology
> and self-declared motivations of American voters than I do, right?

This comment was a general comment (with no intention to talk about the situation in the U.S.).

>> (i.e. in elections where there are very many different opinions, it is not easy to guess how people > will vote, it is not easy to control the voters etc.).
> 
> A mis-statement like that is forgivable and excusable when it comes
> from someone who doesn't live here.

When I talk about "typical elections" that have some typical characteristics I aim at excluding e.g. elections where voters generally vote as told by their party, and the paties have some interest in making the voters vote in some non-typical way. I think the U.S. has some problems but it still probably falls in the category where voters have a tendency to make sensible and independent decisions.

Juho