Explaining STAR almost as easily as Score


Continuing the discussion from Feedback from Fairvote:

I started off with that thinking, but I no longer think STAR is harder.

STAR pulls on the intuition of primary → general elections. People have an intuition about a single “I voted for X”. And STAR is easy to express as:

we use scoring to determine the finalists out of a pool of candidates, and then your same scores show which of the 2 finalists should get your vote in the end.

And then just show them any example from https://star.vote/ and it just clicks.

Score isn’t harder to describe, but it feels in some ways like more of a shift and gets people thinking even more about all the questions and ramifications (and ending up feeling more hesitant/doubtful in some cases).


Pinging @Sara_Wolf to see this


I don’t think it’s that STAR is harder to explain one-on-one by much, but when you’re spreading an idea to the masses, generally the less steps, the better. I agree that with Score Voting, you have the (almost subconscious?) fear that the intense minority may beat the naive majority and hurt them, and that (mostly) evaporates with Approval and more with STAR. But I question whether it’s wise, in general, to push for STAR over Approval, as it seems like a massive change to existing procedures, and it’s not so easy to communicate, from this point in time, why one complex system (STAR) should beat another (IRV) with proven success. (Of course, that “success” is entrenched two-party domination, but it’s familiar and processable, compared to STAR’s promises to the voter that is making a cursory inspection.)


In practice, I don’t find it difficult to spread the STAR message either in general context or one-on-one. There are contentious spaces with people who are generally skeptical about anything new, but Approval doesn’t suddenly just work there. Anyone open-minded can be pretty readily open to the basic messages about STAR. Seeing it in practice at https://star.vote/ is huge.

It is a little harder to get more people to the point where they could readily say the clearest things. It’s easier to tell someone about approval and then have them succeed at turning around and telling the next person. STAR needs to get really solid and clear messaging and resources, but with the right foundations it is fine.

I agree Approval is simpler. But it also feels less comfortable to many voters. When I’m looking at an approval ballot, finding the right line of where to cut off marking approval feels like an uncomfortable strategic speculation. STAR feels slightly similar, but since I can express more than all-or-nothing, it’s just easier. I can say “I don’t like this candidate as much as that one” and score lower. Whereas with Approval I feel like “I’m not sure I want to approve this candidate because that erases my expression that I really feel this candidate is not as good as that one, so should I not approve them just to show that difference, or should I approve them anyway since they are better than that other candidate??” and I sit there feeling awkward and not liking my end ballot regardless of what I mark.


I definitely agree with this.

1 Like

I should add to the core issue that “expressive” is a valuable focus point. Independent of the outcomes, people really like to actually express our judgments. Rank and score systems offer that pretty decently, though it’s not perfect. STAR can be considered effectively a way to be expressive and get good outcomes, and the rest of the details are for wonks even though it’s simple enough anyone can understand the core concept. I wouldn’t say “the details are wonky” early on since that makes it seem more complex than it is. But I also wouldn’t say “it’s simple!” because that’s almost always bad to say about anything. If something is simple, just state the concept and people will get that it’s simple without the assertion.

1 Like

I don’t dispute that that expressiveness is a bonus for STAR. But I suspect that the voter who’s just casually reading about voting reform is more easily swayed to a unique-sounding proposal rather than all the methods where you use scoring or ranking. A lot of people have a fundamental confusion about those two (you hear some people say “rank the options 0 to 5” when discussing Range Voting), so I suspect the reform that most differentiates itself while being simple spreads faster. Fortunately, this isn’t just an idle claim - we will find out soon enough! But as a matter of outpacing IRV, it’s not just winning jurisdictions, but spreading the idea to a voter who’s never heard of it, and on that point we agree Approval (may?) be marginally better. But I am interested to see if STAR speeds up with your suggested improvements as well.


To follow up on this, the tension of Approval can be a good thing. By encouraging voters to think outside the box, you get them looking at the problem of Plurality itself and its vote-for-one limitations. While I don’t think Approval is the be-all and end-all of discussions, I think it’s a tool, a Socratic gadfly that can awaken democracy, one that opens voters’ minds to all sorts of possibilities they hadn’t analyzed before. What Approval does is, in the simplest way possible, show you that there’s a wildfire, and that there’s an array of fire prevention tools you might not have considered. Even if you don’t like the idea of throwing buckets of water (the simple way), it still gets you looking into the biggest problem in our world: two-party domination of politics, and how that weaves its way through every part of our society. I don’t believe we have tried enough approaches to getting the average person to consider why politics is so bad, or how it can be fixed, and while Approval isn’t the only tool for advocacy, I think it’s surprising how little attention it gets compared to STAR and Score, especially before the Center for Election Science was around. It’s a practical flaw when well-meaning, but informed people collide with apathetic voters who don’t want knowledge on the subject, without a lot of simplification. I mean, is it really possible after a decade of the Center for Range Voting that people didn’t grasp the serious problems… or is it that Range itself isn’t going to spread? I’d like to see evidence for either side, if anyone knows more :stuck_out_tongue:


Part of my concerns with complexity come from Lane County, and the world at large. The urban areas of Lane County passed STAR by a near-70% margin, that too, with the campaign running on half as much money as Fargo. So that’s a sign that voters who understand it, who get the proposal, will pass it (or perhaps that liberal-looking marketing will appeal to liberal voters, but that’s unavoidable) But STAR failed dramatically in rural areas, and while part of that is the liberal appearance of marketing signs and the campaign (from what I understand), a lot of it is due to constitutional concerns and FUD raised around STAR. I believe Approval will not have those concerns (as long as you embrace “one person, one vote” into your marketing materials :smiley:); something that radically simple just can’t be FUD’ed or turned into an issue, as its worst-case scenario is just FPTP, and it is so easy and cheap to implement. Where I also think of complexity is the chance of catastrophic system failures in the world itself. I think we’ll have to wait a few years to see, but I’d rather get people on board with a reform that’s probably going to pass narrowly a lot, than a reform that might pass well with voters who understand it, then fail with voters who don’t. Not to mention, there is some benefit to people who aren’t in the ballot-measure-campaigning area understanding the proposal from news and word of mouth - that too helps spread it in a way that can, on its own, add to number of victories over time. But it is close, so for now, pushing for either should have a huge impact on system failure risk in the world, in politics, economics, climate, etc… (another link on the topic, and one model of how to spread ideas that might be useful somewhere… which I think is another reason to focus on simple: you can more easily get majority consent/indifference towards simple, which lets you actually try it out and get people against FPTP)
Here is one way you could characterize Approval:

The fact we saw from the renormalization group the “veto” effect as a person in a group can steer choices. Rory Sutherland suggested that this explains why some fast-food chains, such as McDonald thrive, not because they offer a great product, but because they are not vetoed in a certain socio-economic group –and by a small proportions of people in that group at that. To put it in technical terms, it was a best worse-case divergence from expectations: a lower variance and lower mean.

Approval has a lower “worst-case” scenario in most people’s minds, and more easily expectable, calculable results than something more fine-grained and uncertain, which you need a background in voting theory to predict. So I expect some voters to push us towards Approval, as a way of tamping down on this fear. When people like Score, those who might dislike it because they don’t understand it can’t as easily express that frustration (they’ll get criticized on points they may not fully grasp), so I feel that their frustration will get expressed at the ballot box, rather than directly. As for a campaign, while it may not enthuse people to push for a less superior reform, not having constant resistance or uncertainty around whether you can convince people to at least slightly like the reform/not turn out and vote against it is a bonus for attracting people, both volunteers, supporters, and neutral observers like the media (from what I can tell in media releases on the Center for Election Science.)

(another paper and a book on this “minority propagation of ideas when majorities are open to it”)
Essentially, spreading a small change that a majority can evaluate and be comfortable with is better than a large change that some of them may actively resist out of fear, which will then lead to clashes between the pro and con sides, all while the majority just sits around, and doesn’t absorb the change as fast. But I am not aware of how much progress STAR is making behind-the-scenes, so…

(and finally, some additional readings on the man behind some of this probability-based/social psychology and physics stuff)


So my pitch is:

With STAR voting, you give each candidate, independently, a score from 0 to 9*. The scores are added, and the top two candidates are finalists. Your full vote then goes to whichever finalist you scored higher. It acts like a primary and general, in just one election.

*I like 0-9 more than 0-5.

1 Like

0-9 doesn’t fit with our idea of 5-star voting and quality.

1 Like

You can have that preference, but please don’t pitch that to anyone. The whole STAR campaign went through a long process before settling on 0-5. It’s still STAR with any range, but the clear pitch consistently is 0-5 and that’s that.

Most research shows that 0-5 provides the best balance of seeming simple and accessible vs expressive. If you want to open a new topic to discuss the election-theory of that detail, I wouldn’t object (not that I have any power besides expressing objection, but I think it’s a totally valid topic to discuss, it just doesn’t belong in a discussion of how to pitch/explain STAR).

That’s not necessarily a bad thing as there are pros and cons about making it “familiar 5 star rating”, but as I said above, that ship has sailed as far as the STAR campaign and movement goes. STAR is embracing 0-5 at this point. Discussing that is tangential to discussing explanation and pitch of the system.

1 Like

This is not quite correct. STAR passed everywhere we did any canvassing and had any lawn signs. It passed by the strongest margin (70+% in the precinct where most of us live) where we did a single pass of door to door canvassing on the main drags. (A traditional successful canvass is 3 touches on the house of every likely voter.) It passed by ~54% in precints where we did some canvassing and organizing, but not a lot. It failed in towns that we didn’t go to, at all.

We had red/white/and blue lawn signs and less progressive looking marketing for more rural areas. Marketing was somewhat targeted, though distribution was an issue. Parts of more conservative Springfield where we did some outreach voted yes.

We only ever had one negative article locally published. (The Register Guard had a number of positive articles but was sold to corporate media and then endorsed no at the last minute.) Aside from that the only bad press was FairVote. Nobody raised constitutional or logistical concerns. The Democratic Party bosses locally spread a rumor that STAR is the notorious California Top-2-Jungle-Primary system we already have, and some people are still confused by that to this day. That deliberate bit of misinformation made the rounds in political circles and on social media. Outside of Lane County in Oregon Democrats are some of our biggest supporters, interestingly.


Thanks for the explanation!


My pitch:

STAR Voting uses a 5 star ballot. You score the candidates from 0-worst to 5-best. If your favorite can’t win, your vote automatically transfers to the finalist you preferred, so your vote is never wasted and you can vote your conscience.


The only issue I have with STAR is the difficulty of explaining on what occasion it would yield a different result from Score voting.
I still don’t have a pithy explanation for that one.


My attempt at pithy “outcome different than score”:

“For example: in a large pool of candidates, maybe the majority expressed only a weak preference between the two finalists (like scores of 1 vs 0). That might be outweighed in the scoring by a minority who expressed a stronger difference the other way. So, the majority-preferred candidate still wins, even if they had lower score.”

Could add: “Of course, they still needed a high enough score to get to 2nd place.”

None of this is a complete explanation given various scenarios, principles, and ramifications. But it’s one way to just pump the intuition that cases can happen where the outcome is different from score.

The other key point: “If everyone votes honestly, plain score and STAR will usually give the same results. But the runoff stage in STAR reduces the impact of dishonest strategies so the scoring is more reliably honest. So, it’s less about the runoff changing the outcome directly and more that including it gets voters to give more accurate and expressive scores in the first place.”


It allows the runner-up from the Score election to win if they’re better for the majority.

1 Like


My tweak: is because it’s a statement about the system, not whether voters’ preferences actually serve their own interests or not.


The main problem I have with saying the majority preferred the candidates… it implies a certain level of deliberation and choice that utilitarian systems don’t have. It is something I’ll agree to differ on, though. I believe the only time we can really say the majority preferred some other outcome over the utilitarian one is when they were actively debating and thinking about the minority in their decision, but it isn’t far off to say that they passively prefer a certain outcome until they’ve thought more on it.