An important note for Hugo voting.

[theweaselking]

More so this year than most years, it's important to understand how voting "No Award" works.

The Teal Deer: If you want to vote No Award over something, put No Award at the end of your ballot and DO NOT list the things you're voting No Award over.


Basically, if you put something on your ballot AT ALL, you risk your vote going to that thing, especially since "No Award" is most often the first thing eliminated. The reason it works that way is a little counterintuitive, until you realise that No Award is just another candidate on the ballot.

For an example showing the problem, let's take perennial No-Award-voting favourite category, "Best Doctor Who Episode Written By Stephen Moffat". In 2012, the candidates were Community, some dude having a meltdown at the previous year's Hugo awards, and three largely indistinguishable Doctor Who episodes written by Stephen Moffat, Neil Gaiman aping the style of Stephen Moffat, and Tom MacRae.


Let's imagine your ballot. You are a sensible person of good taste! You want Community to win and all the other options to die in a fire.

You SHOULD vote:
1. Community
2. No Award

The naive model of Hugo voting that a great many people have might make this mistake, though: Because they want the other options to die in a fire, but ESPECIALLY hate that one dude's meltdown and the particularly bad Stephen Moffat episode, they might vote like this
1. Community
2. No Award
3. Some dude having a meltdown onstage
4. Stephen Moffat.

And, y'see, that's bad. Because of how the vote counting works, many people THINK they're "leaving off" Tom MacRae and Neil Gaiman but making extra sure to "downvote" Moffat and Some Dude, when actually they're voting *for* Some Dude and Stephen Moffat *over* MacRae and Gaiman.

Because the voters in this category historically have poor taste, let's imagine the first-round of ballots runs:

Gaiman:50
Moffat: 40
MacRae: 30
Some Dude: 30
No Award: 10
Community: 1. You are the only person with taste this hypothetical year, hypothetical Hugo Voter.

So, Community is eliminated, and all the first-place Community votes (yours) now go to their second choice: No Award. Which is now last and *it* is eliminated, which dumps all the votes for it (including yours) to the next choice down. In your case, Some Dude.

Let's pretend the 10 people who stuck No Award first really did mean it and didn't list anything, so their votes now vanish. This leaves the current voting as:

Gaiman:50
Moffat: 40
Some Dude: 31 <- your vote went here!
MacRae: 30

... and MacRae is now last, and is eliminated.

Congratulations, Hugo Voter. You just eliminated Tom MacRae by throwing your support to Some Dude, when you *meant* to say that Some Dude was so terrible that the only person he should lose to this year is Stephen Moffat and you were actually somewhat okay with Tom MacRae even if you didn't think Doctor Who should be mistaken for "best of the year".


The point is, voting No Award is a useful tool! But anything you list after No Award is going to get your vote and your support BEFORE things that aren't listed at all. So don't do that. If your ballot goes:

1. No Award
2. Chlamydia
3. David Duke

and you leave off the rest of the possibilities because you haven't read them or don't care, then when No Award is eliminated (it almost always is eliminated first, or second if one of the nominees is L Ron Hubbard), your vote goes to Chlamydia. And when Chlamydia is eliminated, you've now voted for the Grand Wizard. So *do not* list people you genuinely do not want to get the award below No Award. List No Award last, and do not list them at all. Things you list under No Award can and possibly *will* get your vote.[1]

And now you know! And don't fucking vote for Doctor Fucking Who, seriously people. It's okay, but it's certainly not the best TV the year has produced.

For everyone who doesn't care about this style of voting or nerd awards but has still read this far, here's a ninja astronaut airsurfing a Draculabot:



[1]: I am ignoring the No Award Test, because it complicates the example unnecessarily.

Comments

[zenicurean.livejournal.com]

Dr McNinja!

[supergee.livejournal.com]

Blogging this; thanx

[torrain.livejournal.com]

I find that this comic is an excellent summary of it, too.

Also it features Plom McFartle looking into... "hugs".

[theweaselking.livejournal.com]

The Australian system requires that you rank all candidates and does not have a "No MP" option, so it's not exactly like the Hugos, but yes.

[thomas monaghan (from livejournal.com)]

Your name is very apt!

[theweaselking.livejournal.com]

The waffle strikes at midnight? I'm not too sure what you're trying to say, here.

(Did you have something to say, or was this just a contextless driveby nonsequitor on a clarification of the differences between Australian Elections and the Hugo awards?)

No award

[livejournal.livejournal.com]

User supergee referenced to your post from No award (http://supergee.livejournal.com/3289243.html) saying: [...] How to vote against [...]

[theweaselking.livejournal.com]

It gets even worse, from my example's perspective, if all the Tom MacRae voters *really liked* Some Dude. Not enough to vote him first, but certainly enough to vote him SECOND. Which means that, because of your hate-vote for Some Dude, Tom MacRae gets dropped and all those votes go to... Some Dude.

Leaving the next round of the ballot at:
Some Dude: 61
Gaiman: 50
Moffat: 40

Uh-oh! Instead of eliminating Some Dude by putting MacRae on your ballot *anywhere* above him, you've not only eliminated MacRae but you just hate-voted Some Dude into first place, eliminating Moffat! Better hope Moffat fans like Gaiman more than Some Dude, otherwise your hate-vote just gave Some Dude the Hugo.

[nancylebov.livejournal.com]

I hope some version of these instructions are included in the voter's packet.

[kevin-standlee.livejournal.com]

There are instructions, and copies of the WSFS Constitution, distributed with the ballot. Most people don't read them. Most people seem to have difficulty with anything other than "First Past the Post Voting." And not many people even know that the No Award Showdown exists because it's never been invoked -- No Award has never had more ballots ranked higher than the Tentative Winner.

The Hugo Awards Web Site has an article about The Voting System (http://www.thehugoawards.org/the-voting-system/), which does include the No Award Test.

While the instructions seem awfully complicated, the gist is really simple: If you like it, rank it higher than things you don't like. When you get to the point where you don't care about who is left on the ballot, or run out of candidates, stop ranking candidates. The system is built to simulate the process of multiple rounds of runoff elections done in person, which is why it's sometimes called "Instant Runoff Voting."

[theweaselking.livejournal.com]

Indeed, I linked that exact page in my original post.

But your point, Most people seem to have difficulty with anything other than "First Past the Post Voting." is why I made the post I did: I, and presumably you, have seen a number of people suggesting that they'll vote all the works that were good, then No Award, then the works that sucked - leaving off the works they didn't get to or don't care about. And my point (and yours) is that that's WRONG. That ranks "don't care one way or another" below "that sucked". So I'm trying to correct that.

[kevin-standlee.livejournal.com]

But that's exactly what the system is supposed to do: if you disliked a work, but have a relative preference, rank it. Only ignore works that you don't care about.

You're presumably thinking that "don't care" is somehow weaker than "dislike." "Don't care" is an abstention. It's like you didn't vote at all. And I hate systems that force people to make a preference if they really don't care. Voting should always include the right to abstain.

In the example you give, if one of the candidates you disliked won, but your preference helped it win, that's appropriate. "Dislike" is a preference. "Don't care" is not a preference -- it's the absence of preference. If you would be happier if "I don't care" had won instead of "that sucked," then you should have ranked the "I don't care" work ahead of "that sucked."

[theweaselking.livejournal.com]

Again, you aren't disagreeing with me.

You're presumably thinking that "don't care" is somehow weaker than "dislike."

No, because I know how it works. The point of my post was for people who *don't* know how it works.

If you would be happier if "I don't care" had won instead of "that sucked," then you should have ranked the "I don't care" work ahead of "that sucked."

Yes. Which, again, I want to point out, is something that A LARGE NUMBER OF PEOPLE have expressed confusion about, and is the error I'm trying to correct.

[kevin-standlee.livejournal.com]

You do know, right, that even if your No Award vote is eliminated, it comes back for the "No Award Showdown test" at the end of the process:

After a tentative winner is determined by normal Instant Runoff Voting, you compare the number of ballots that voted No Award above the Tentative Winner (or didn't list the TW) versus those ballots that listed the Tentative Winner above No Award (or didn't list NA). Ballots that list neither the TW nor NA don't count for this test. If by some chance No Award comes out on top in this test (it's never actually happened), No Award wins.

The important thing here is that your votes below No Award are saying, "I don't really want this candidate to win at all, but if it must win, I prefer it over the others I've ranked lower." Your vote ends up showing a preference between the candidates you don't like that much while still having a technical possibility of knocking it off completely when it comes to the No Award Showdown.

This Showdown is not part of normal Instant Runoff Voting. It's Section 6.5 of the WSFS Constitution (http://www.wsfs.org/bm/const-2013.htm#article6).

[theweaselking.livejournal.com]

You do know, right, that even if your No Award vote is eliminated, it comes back for the "No Award Showdown test" at the end of the process:

I am indeed aware - hence my footnote about skipping the No Award test because it would overcomplicate the example.

The important thing here is that your votes below No Award are saying, "I don't really want this candidate to win at all, but if it must win, I prefer it over the others I've ranked lower." Your vote ends up showing a preference between the candidates you don't like that much while still having a technical possibility of knocking it off completely when it comes to the No Award Showdown.


Yup. But if your No Award is not widely shared - for example, when you're voting No Award over Doctor Who in the "Best Doctor Who Episode Written By Stephen Moffat" category - you aren't going to knock it out at the end. And you *are*, likely inadvertently, expressing a preference for your "downvote" over all other things you didn't mention. Which is my point.

[kevin-standlee.livejournal.com]

Well, yes, and that's the point. You're saying that you don't like anything below No Award, but if one of them must win, you would prefer A over B. Now if you genuinely don't care whether A, B, or C wins, but hate them all equally, then yes, leave them off your ballot. But the NA test is designed to stop a winner actually winning when more people voted for No Award than for BDWEWBSM, even if their sub-No Award votes converged on that candidate.

There are some people who think that only ballots that rank BOTH No Award and the Tentative Winner count; this isn't true. Ballots that rank EITHER OR BOTH count in the showdown. In the Showdown, it's as if there are only two candidates: the TW and NA, running head-to-head against each other.

[cartesiandaemon.livejournal.com]

I thought I understood this, but now you've confused me again.

It seems like some people think that leaving off works is "neutral", and putting them below "no award" is a vote against? I'd never heard of that misconception, but I agree it's wrong. AIUI, basically works you don't vote for are treated as ranked equally at the bottom of your ballot (below no award if and only if you listed no award on your ballot).

In fact, it might be nice if there were a way of saying "I haven't read these books, and if people who HAVE read them think they're better than my #1 I don't disagree", but there isn't.

But your proposed solution only seems to make sense if you've heard of that misconception. I would have phrased it as "don't leave anything out of your ballot in the middle, if it's better than anything you did list". In fact, I think my explanation of the ballot would just be something like a sample ballot:

1. Gandhi
2. Someone fairly good
3. Someone else fairly good
4. No award
5. Someone mediocre
6. Someone else mediocre
7. Someone who's friends with Hitler
8. Hitler.

Leaving off everything below "no award" seems completely wrong if you don't want any of them to win, but if they do, you really, really, really want the mediocre people to win over friends-with-hitler and want friends-with-hitler to win over hitler?

[theweaselking.livejournal.com]

AIUI, basically works you don't vote for are treated as ranked equally at the bottom of your ballot (below no award if and only if you listed no award on your ballot).

That's (mostly) correct - the only quibble I have with it is semantic differences about "ranked equally at the bottom" versus "not ever preferred" and those really are semantic. That's also not how a number of people have expressed an intention to vote.

Leaving off everything below "no award" seems completely wrong if you don't want any of them to win, but if they do, you really, really, really want the mediocre people to win over friends-with-hitler and want friends-with-hitler to win over hitler?

Correct. In your example, if you rank every last possible entry, your vote will go to every person other than Hitler before it goes to Hitler, and it will only go to Hitler once every other participant is eliminated - which will never happen. You have correctly voted as you intended: "Preferred nice people, then don't give it out at all, but if you are going to give it to an asshole who doesn't deserve it then please rank the assholes in this order".

If you leave any entry off, however, you're voting for everything you *do* rank over everything you *don't* rank. Even below No Award.

And this matters in the Hugos because of the number of people who've expressed a preference along the lines of "Well, I don't know anything about a couple of the entries and haven't liked the author's other work, but I'm sure it's fine - but I really hated that one and/or call shenanigans on the nomination, so I'll leave off the don't-cares but put 'hate' below 'No Award'."

[cartesiandaemon.livejournal.com]

Thank you! OK, I think we agree about how the voting actually works. I'm confused because I've seen OTHER misapprehensions, but not this particular one, so what you were saying didn't immediately make sense to me.

[nojay.livejournal.com]

A quick summary --

1. Write your preferred choices in order.

2. If you don't care if something else wins, don't write their name on the ballot.

3. If you actively don't want something to win, write No Award as your last choice after the ones you like AND NOTHING ELSE afterwards.

Does that sound about right, Kevin?

[kevin-standlee.livejournal.com]

Yes, that works. The only reason to rank works after No Award is if you don't like them as a group, but do want to say, "If you must give the award to one of these pieces of junk, pick X over Y over Z."

Interesting Links for 14-05-2014

[livejournal.livejournal.com]

User andrewducker referenced to your post from Interesting Links for 14-05-2014 (http://andrewducker.livejournal.com/3090537.html) saying: [...] ) A brief reminder on how ranked voting works (in this case, for The Hugo) [...]

An important note for Hugo voting.

[livejournal.livejournal.com]

User lsanderson referenced to your post from An important note for Hugo voting. (http://lsanderson.livejournal.com/2914174.html) saying: [...] An important note for Hugo voting. How to vote for No Award as your last best choice [...]

[frith]

That's not a ninja, ninja are silent and unseen, not loud and sticking out like a sore thumb. The surfer is in fact my brother (there's a VR screen in the helmet so he doesn't realize how far above the ground he is).

[malkaesther.livejournal.com]

I think I finally understand how the voting works. This is my first real year voting. Thanks for the great post.

[kevin-standlee.livejournal.com]

I've always thought people over-complicate the process. The way I approach the ballot is to look at the choices and say, "Who do I want to win?" and put a 1 by that choice. Then I say, "If that wasn't on the ballot, who would I want to win?" and put a 2 by that choice. The repeat this process until I get to the point where I either run out of choices or say, "I don't care whether or not any of the remaining choices win or not."

The voting process simulates a case of in-person voting where we count a show of hands for each candidate, and if anyone gets a majority, s/he wins, but if not, whoever comes in last place drops out and we re-vote, repeating the process until someone gets a majority. That's why it's also called "Instant Runoff Voting."

[theweaselking.livejournal.com]

The confusing factor is the addition of the "I would rather NOBODY win than ANY of the people left on this ballot" option. And the accompanying weird cognitive twist that leads some people to think that voting someone "below NOBODY" is a downvote.

[malkaesther.livejournal.com]

I agree that people have over-complicated how "don't want to win/hate", "don't care", and "no award" play together. That's what has had me confused with all the explanations. Knowing to leave them hate/don't care off the ballot and put no award as my last vote helps a lot.

I've found as a technical writer people tend to overthink directions. I try to run instructions past 8-year old kids for anything software related... Well for 90%+ of stuff I write. If a kid can't follow my instructions I figure at least 50%+ of my users won't be able to as they are only using help/docs when at wits end/angry/frustrated/stressed. Keep it simple stupid :D

[theweaselking.livejournal.com]

Also: I really appreciate you popping up to clarify and help. I know you've got a Google Alert or something set to ping you on terms like "Hugo Voting", but your contributions, here and everywhere else I've seen you pop up, have always been helpful.

[kevin-standlee.livejournal.com]

Thank you. Not everyone appreciates the Invoke Kevin spell. You were actually referenced from elsewhere. Also, anything that generates a pingback to TheHugoAwards.org will cross my e-mail because I'm one of the administrators of the web site. (Which is not the same thing as being one of the people who administers the Hugo Awards themselves, I hasten to add.)

And I've found that some people really resent being shown the facts of how the awards work, how the rules are written, etc. They'd much rather have convenient conspiracy theories. They're more comforting than anything that amounts to, "Not everyone likes the things I like."

[theweaselking.livejournal.com]

I know there were a bunch of places this one got linked. Livejournal's been so relatively dead that I'd kinda forgotten what it's like to get an influx of strangers.

And it's not the conspiracy theories that bother me as much as the perfectly reasonable people who have missed a critical point and are about to vote *against* their intention. The fact that not everyone likes what I like is a different problem - a known issue that will be corrected in the next major release.

[silmaril.livejournal.com]

Thank you, I really appreciate this explanation.

Of course, if they don't release the package soon I'm going to have to vote for Best Fan Writer and nothing else, because I won't have time to review.

[kevin-standlee.livejournal.com]

The Hugo Voter Packet was released yesterday (http://www.thehugoawards.org/2014/05/hugo-voter-packet-released-2/).

[silmaril.livejournal.com]

Thank you!

(Didn't get an e-mail, wasn't checking every day.)

[calico-reaction.livejournal.com]

Thank you. So much. I was wondering how NO AWARD worked with the ranking system, and I wonder why they simply don't disable your ability to rank once that's selected. Oh, and THANK YOU for the particular example you chose. I'm still bitter over that Community episode not winning.

[kevin-standlee.livejournal.com]

The reason they don't make No Award "sticky" (which is what you're suggesting) is because there was a large debate on the subject years ago at a WSFS Business Meeting between people who wanted it to be "sticky" (because they want No Award to win more often) and those who want it to be considered the same as any other candidate (and therefore something that can be eliminated like any other candidate). The current rather complex "No Award Showdown" rule is a result of a compromise between the two opposing camps.

As I recall, the last time WSFS debated this was 1991 (http://www.wsfs.org/bm/minutes-1991.html), and nobody has brought it up again since then, so the current system stays. Anyone who wants to change it has to go through the effort of proposing a constitutional amendment and convincing people to vote for it, a subject on which I've just introduced a brand new proposal (http://kevin-standlee.livejournal.com/1350599.html) for consideration at this year.

Remember, that when it comes to matters of WSFS, if you have a Worldcon membership, it's not "they," it's "we," since every member has the same legal right to introduce proposals. There's no Board of Directors who makes decisions. We run the official business of WSFS like a New England Town Meeting, with all of the hazards of such direct democracy.

[calico-reaction.livejournal.com]

I could be misunderstanding, but I was always under the impression that supporting voters (which I've been for the past few years) didn't get to introduce proposals or anything like that. If I'm mistaken, then I happily stand corrected.

Also, thank you for that wonderful detailed explanation. I don't have a huge dog in the fight about wanting "No Award" to win more often; however, it's shitty that my first place vote could go to something I don't want to win at all if I'm not ranking things properly, you know?

I look forward to reading your proposal! I wish you luck!

[kevin-standlee.livejournal.com]

Thanks for the kind words.

Supporting members are granted the right (it's actually not a general right under parliamentary law, but it has become established by WSFS precedent) to introduce proposals to the Business Meeting even though they can't be there to debate or vote on them. It's rare, but not unheard-of, for non-attending members' proposals to pass. (To be fair, it usually happens when a well-known, respected member who has attended past meetings isn't at a Worldcon but sends a proposal. At the very least, the members present are likely to debate it rather than spike it as a no-hope job.)

My proposal to which I linked is to make the ratification stage of changes to WSFS rules a vote of all members (including the supporting members) of the following year's Worldcon, rather than having the proposal come back before the Business Meeting a second year. It's like how many US states' constitutions are amended: the legislature proposes amendments, and the people vote on them. This would give you, as a supporting member, a direct vote on anything that passed last year.

[theweaselking.livejournal.com]

It is worth noting, as other people have in the comments here and over at Scalzi's, that if you *are* going to rank everything, ranking the stuff below No Award is potentially useful: "None of these deserve to win, but if one of them has to, I hate that one more".

It's only if you don't rank each an every option that options under No Award are problematic, and then only if you don't really mean to rank the ones you chose over the ones you don't choose.

[calico-reaction.livejournal.com]

But long story short, if I hate something, I shouldn't rank it at all, let alone under my ranking for No Award?

No title

[livejournal.livejournal.com]

User marthawells referenced to your post from No title (http://marthawells.livejournal.com/604256.html) saying: [...] this again, as it's getting closer to the Hugo voting deadline: An important note for Hugo voting. [...]