Policy Review Polling Options

Status
Not open for further replies.

jas61292

used substitute
is a Community Contributoris a Top CAP Contributoris a Forum Moderator Alumnusis a Battle Simulator Moderator Alumnus
In this thread, there are a few issues related to polling that I would like us to address. This includes the types of polls we use for competitive votes, the sizes of slates for competitive polls, and when we should be using preferential block voting (PBV), the vote counting system introduced to count the Topic Leadership Team selection vote, rater than instant runoff voting (IRV). While the first two issues are a lot more linked together than the third, I believe that they all fall under the same general concept of polling issues and are similar enough to address together in one thread.

To give a little background, the reason I wanted to make this thread during this PR cycle was because of something I noticed this past CAP project that I could not remember ever seeing in the past: the use of multiple bold voting (MBV) for the first round of competitive polls. This happened both for the concept poll and the movepool poll. Having looked through the archives, I have found that this was not the only time this has been done, having also been used for the Necturna movepool poll, but, in general, this is something that has not been typically done, so it surprised me that it came up, not once, but twice this project.

Now, what exactly is the issue with this that I think needs discussion? Well, I believe that, in a large slated competitive poll, having unranked votes is a detriment to the project as it overly encourages strategic voting and makes it impossible for people to express more than their vaguest preferences. As I said earlier, the reason I wanted this thread was because of what happened this past project, but to be more specific, it was because of the realizations I had when trying to vote in the first movepool poll. As a lot of you may remember, I was very much against giving Cawmodore all of Acrobatics, Drain Punch, Roost, and other moves. However, I was not equally against all of them. Acrobatics was my number one sticking point, so, if this were a ranked vote, I would have put those movepools without Acrobatics first, followed by those with, ranked based on other moves included. However, this was not a ranked vote. Everything I would vote for would be given equal weight, meaning I would have to decide whether it was more important to vote trying to get rid of those I disliked most, or to stick with my number one most important thing and only vote for the movepools without Acrobatics. I essentially was forced to choose a breaking point and vote based on what I wanted to see (or not see) going forward, and not based on my actual preferences.

To put that in more general terms, an MBV on a competitive poll essentially forces people into strategic voting, something which I believe we do not want in any stage of the competitive process. So, with that said, the first thing here I want to propose is the following:

Proposal - No future concept, typing, stats, primary ability, secondary ability or movepool poll will ever use Multiple Bold Voting

Maybe I'm wrong about this, but I'd like to believe that, for the most part at least, the above proposal will be easy for people to agree on. However, simply saying that we want to do the above is not really a solution. MBV is typically used when we have large slates (around 7 or more) in order to narrow it down, so if we are going to eliminate the use of MBV in competitive polls, we need to do something about situations that involve larger slates.

One easy solution would simply be to use a different type of voting, either IRV or PBV (or something else if anyone has a better idea) in such a situation. Now, I don't know a ton about the statistics of polling, but while, according to the CAP site, MBV "is typically better than IRV for determining the rank of entries between first and last place for polls with very large submission numbers," I believe that for the purposes of competitive polls, no slate that we will likely have will be big enough for any such advantage gained from MBV to be worth the negatives that come with it, nor large enough for the results of an IRV or PBV poll to be significantly inaccurate.

With that said, while I believe that using IRV or PBV for these polls would be better than MBV, I think that simply saying we should do that is ignoring an important issue: why suddenly this project did it come up more than once when throughout the rest of the generation it has only come up one time total? I think the answer can be found in a general trend towards larger slates this past project. Check out the slate size for the competitive steps over the past five projects (all the projects since we made typing into a single poll):

Code:
Poll        | CAP 2 | CAP 3 | CAP 4 | CAP 5 | CAP 6 |
-----------------------------------------------------------
Concept        5        7       7       7        9
Typing         4        6       4       5        5
Ability 1      4        3       5       4        6
Ability 2      -        2       3       5        5
Stats          5        7       7       8        8
Movepool       8        7       7       9       10
While there are a few exceptions, I think it is pretty plain to see that in general, the trend over the past couple years has been towards larger slate sizes. This is especially evident over the last two projects. CAP 5 set multiple high marks, only to have CAP 6 set just as many the next time, and be at least tied for the high on every single stage but one (typing, where the only project with a larger slate was Mollux: the one with the concept all about typing).

So, why exactly have we seen such increases? Personally, I believe this to be because of the TLT system we implemented prior to CAP 5. Within our current system, each section leader lead discussion and sets slates, while the TL can add or remove from the slate base on their interpretations of the concept. The key element here that I believe leads to large slates is the fact that we have essentially told the section leaders to just look for what the community wants, and leave the analysis for how they fit with the concept to the TL. Now I'm not going to sat that this is a bad thing, but I believe that, in having done this, we have lead our section leaders to start slating a much larger number of things than they otherwise would. Since their job is just to see what the community thinks is a good idea, many times I have gotten the impression that TLT members have decided to just slate anything that got a single intelligent supporter.

Now that sounds all nice in theory, but I really believe that, in doing that, TLT members are ignoring an important part of their role. The job of a section leader when it comes to slating is to make sure that all options that have received significant intelligent support get slated, and while I comment all our TLT members over the past two projects for making sure the "intelligent support" part is satisfied for all things slated, I look back and see that there has been more than one case where things were slated which have had only one or two people providing intelligent support, which seems to show a lack of attention to the "significant support" part.

Note: I don't believe anything regarding this "significant intelligent support," or really anything regarding how a TLT member should be making a slate, has actually been codified in any way. My use of the phrase is based on interpretations of the intent of the TLT system from the thread in which it was created, as well as countless discussions on the topic that I have had with numerous users on IRC over the past year.

Of course, it is always possible that I am wrong about this. Maybe the larger slates are just a coincidence and have nothing to do with the TLT system. Either way though, I personally believe that these incredibly large slates are a problem, and that fixing slate size is a much better solution to the problem of using MBV in competitive polls than simply using another voting method, regardless of how effective it is.

I don't want to throw out any specific proposals on this issue yet without at least some discussion, but I'd like people to at least think about the possibility of setting some sort of limits on slate sizes. I find it hard to believe that, in any given poll, there is ever really more than 6-7 options that really have enough support to be considered viable slate contenders in comparison to the other things being considered.

So, what I would hope to discuss here are:

A) if using IRV or PBV for large slates is acceptable
B) whether or not decreasing slate size is desirable
C) positives and negatives surrounding implementing limits on slate sizes
D) any alternative ways people may come up with of reducing the slate sizes without hard limits
E) any alternative ways people may come up with of fixing the problem of MBV in competitive polls that don't involve reducing slate size or using IRV/PBV

-------------------------------------

In addition to all of the above, there is one more thing related to polling that I would like to discuss. When we implemented the TLT system back before CAP5, we had to come up with a new polling method for the selection of the TLT members. IRV was deemed undesirable for this job as second place in an IRV simply says who would have the most support among people who did not support the first place winner, not who actually has the second most support. We wanted the TLT to consist of the 4 people with the most overall community support, so taking second, third and fourth place in an IRV would not give us what we are looking for. So, instead of IRV we decided to introduce PBV. PBV is a form of voting similar to IRV, but specifically designed to pick multiple winners. Rather than having second place be the person with the most votes by people who did not prefer the eventual winner, PBV finds the IRV winner and then, to find second place, removes them from contention and runs an IRV again, finding the most prefered option of the entirety of the community among all the options other that the first place winner. This repeats until you get the desired number of winners.

While this system is perfect for TLT elections, it would also seems to fit in well in almost any poll where we previously would only use IRV. In fact we already have begun using it elsewhere, including in Art Poll 2 for both CAP 5 and 6, Name Poll 2 for CAP 5, and Movepool Poll 2 for CAP 6. As you can see though, there is not really any consistency as to why or when to use PBV rather than IRV, and that is what I'd like us to discuss. So, let me spark such a discussion with this proposal:

Proposal - Preferential Block Voting will replace Instant Runoff Voting in all standard CAP polls

I personally believe that PBV, as a method designed to pick more than one winner, is more appropriate for our purposes than IRV in pretty much every instance that we currently use it. When we have more than 3 options, we never set out to choose only one winner, as IRV was designed to. Now, in the end we do eventually want to pick only one for any poll other than TLT selection, so in theory the winner of the IRV, which would be the same as the winner of the PBV, should always win the final bold vote. But, we know that is not always the case. The difference with narrowing it down with PBV is that the options voted on in the final bold votes will be the three overall most prefered by the community, rather than the communities number one vs the number one of the smaller portion of the community that did not prefer the first place, vs the number one of the even smaller portion that did not prefer either of the above. This could lead to more blowout final votes, as the competition might be everyone's second favorite, but no one's first, but I believe it would overall cause votes to be more accurate to community preference. As a side benefit PBV could help reduce drama by not always postponing showdowns of controversial options until the last poll.

I don't particularly see a ton that is favorable about sticking to IRV, but if someone who knows more about these methods or the statistics thereof and would want to bring up some points, please do.

-------------------------------------

As a final note, if anyone has any other issues regarding polls, polling methods or slates, feel free to bring them up in this thread. If anything seems large enough to merit its own thread, we can deal with that as it comes up, but for now I think we should try and keep any poll related topics in here.
 
Last edited:

Stratos

Banned deucer.
wow for a wall of text that was a surprisingly quick read


The first thing I'd like to comment on is the slate size issue. I... somewhat disagree with you that it was brought about by the TLT. The attitude you described for section leaders... that isn't the attitude I had. I didn't want to see if something "aligned with the concept" before I slated it, and I think that such an attitude was one of the major problems with the Strong TL model. My slating policy was inspired by the thread you linked, Doug's post here in particular. And given how much I stressed this post on IRC throughout CAP5, I wonder if maybe it had an influence on the rest of the TLT too during CAP6 (maybe because they didnt want me to yell at them rofl!) I may have been section leader, but I was still only one user, with one opinion on the concept. It wouldn't have made sense for every single ability to fit my very narrow idea of the concept. CAP is a community project, not a Pwnemon project. So I slated any ability that looked like it had significant intelligent support from the community.

It's important to note how I defined intelligent, though. Intelligent had very little—practically nothing—to do with how well I thought the ability would work on the CAP. Intelligent just meant that the arguments were clearly grounded in a solid understanding of the metagame. Though I originally supported Magic Bounce, by the end I was certain it would ruin the CAP. I also hated Unburden with a burning passion, and thought it would make a CAP that taught us nothing about the metagame. Both were supported, though, by arguments based in an accurate representation of the meta (from fuzz / nyktos respectively). Why the fuck did my imperfect theorymon predictions matter more than theirs simply because I'd won an election and they hadn't? As i saw, it was neither my nor cape's role to judge "concept" in slating. The community gets to pick. My job was simply to weed out the ones posted by users who clearly had no clue what they were talking about and those with no significant support.

Intelligent is easy to judge. Significant is hard to judge, and I wonder about my decision to slate Sand Rush in the primary ability poll. However, I do believe we should err on the side of caution. My rule was that if not slating something would make more than one user on irc angry, I would slate it. This is relatively easy for discussion stage leaders; we could narrow down pretty clearly what was generating discussion and what wasn't, so we knew what had as much chance as a snowball in hell. Submission stage leaders must have had hell on that, though, since there is almost no commentary during their stages, so they had to eyeball it a lot more. What this means is that we did end up with big slates. I don't see why that's a problem. I liked the big slates. It meant it was a community project. I would not like to see limits on slate sizes.

as to preferential block voting: would this require the same "vote for every option" imposition that we have on the TLT votes? I'm curious as to the effect it could have if we did the same thing of deleting all non-compliant posts, so this is an important question.

EDIT @ Below: in that case, full support for PBV. I see no downside, except having to write a new script to read votes.
 
Last edited:

jas61292

used substitute
is a Community Contributoris a Top CAP Contributoris a Forum Moderator Alumnusis a Battle Simulator Moderator Alumnus
Just a quick reply:

The first thing I'd like to comment on is the slate size issue. I... somewhat disagree with you that it was brought about by the TLT. The attitude you described for section leaders... that isn't the attitude I had. I didn't want to see if something "aligned with the concept" before I slated it, and I think that such an attitude was one of the major problems with the Strong TL model.
Hey, would you mind pointing out what exactly gave you the impression that I was saying the TLT members would be looking for things aligned with the concept? I specifically was trying to say that that was NOT the TLT member's job, but rather that it was the TL's job to make sure nothing on the chosen slate was obviously NOT aligned with the concept. I mean, the entire extent of the TL's job nowadays is to champion the concept throughout the project. However, the TLT members are not given that job, and I completely agree with what you say there regarding TLT analysis of concept, so if something in my OP is implying otherwise, I'd like to fix that.

(By the way, can we please, PLEASE, change the name of the TLT to something else. Those super similar acronyms are killing me.)

as to preferential block voting: would this require the same "vote for every option" imposition that we have on the TLT votes? I'm curious as to the effect it could have if we did the same thing of deleting all non-compliant posts, so this is an important question.
Technically, we could do it either way. However, unless we want to do it like we do for TLT and have only one poll ever, there is no advantage to doing it that way, and if we were to do that, well, it would just be IRV anyways.
 
Last edited:

Nyktos

Custom Loser Title
As a final note, if anyone has any other issues regarding polls, polling methods or slates, feel free to bring them up in this thread.
okay i will

First, I would like to see the victory condition in multiple bold vote polls changed. Right now, we normally have the top N move on, where N is an arbitrary number chosen by whoever closes the poll. Setting aside the issue with the abitrariness, moving on a fixed number sucks because it encourages strategic voting. In particular, if you have one favourite option, the "correct" thing to do is to vote for that option and only that one, because every vote received by any other option is effectively a vote against your favourite. This is sad, because we mostly use MBV for art polls, and we want to encourage people to vote for all the pieces they like. A better choice for the victory condition is that all options supported by at least one third (rounded down) of voters move on. This means that it's fine to vote for all your favourites, because whether your number one choice moves on is solely based on the number of votes it gets, not any of the others.

I chose the one-third number because if it were already in use, it would have lead to the exact same second-poll slate as was actually chosen in the art polls for CAPs 2, 3, 5, and 6. The same is also true for CAP 2's movepool poll which was MBV. As such this isn't really shaking things up very much in practice. Now, of course, there were some polls where the result wouldn't have been the same, but I claim that none of the ones I looked at (all of the past five CAPs) are problems. The closing post for CAP 4's first art poll remarked upon the large drop in votes after the third option and suggested only those three would have been moved on had the OP not specified seven. The one-third standard would have advanced the top four in that poll, so nobody who actually had a chance would have missed out. CAP 5 and 6 names would have had significantly smaller second polls (three and two names respectively) using this method, but that actually sounds like a good thing to me. CAP 6's movepool would have had a moderately smaller second-poll slate (five options instead of seven) but the OP proposes that we stop doing MBV for competitive polls anyway, and I agree with that.

Now, there are a few theoretical weaknesses to this proposal, namely the possibilities of no options receiving enough votes or (nearly) all of the options doing so. These are unlikely to ever come up, but for the sake of having official policy for those cases, I'd propose that the highest-ranked option simply be declared the winner in the former case (since frankly it doesn't really matter who wins if every option is that unpopular). To avoid the latter case, we can still have some upper limit on the number that move on, as long as it's comfortably larger than the number we would expect to ever actually advance. Twelve or so, perhaps, or even higher. (Edit to clarify: This number should basically be the absolute maximum we'd be willing to have in a non-MBV poll, whatever that is.) Again, I'd like to stress that adopting these policies is more or less just to appease the paranoid and neither one is likely to come up, so this shouldn't be a major focus for discusion.

ok now for the controversial one

Can we please quit using voting methods that are specifically designed to avoid the necessity of runoffs and then doing runoffs anyway? Seriously, the way we do voting in CAP right now makes absolutely no sense. The reason it's called instant runoff voting is because it simulate the results of immediately holding a runoff. It's true that the winner of the IRV poll doesn't always win the final poll in CAP (though they usually do: to be exact, they did in fifteen out of twenty [i.e. 75%] competitive polls in the past five CAPs, excluding first-poll majorities). However, a different result in the IRV poll and the last poll simply signifies that either people are changing their minds about which option they prefer, or it's not the same group of people voting. Neither of those are a reason to consider the final poll winner any more legitimate than the first poll winner. (And I hope that the PRC doesn't consider single-vote IRV to be a bad way of choosing a winner if we pick TLs that way...)

The use of PBV here is barely an improvement. PBV is an excellent choice for the TLT poll, because that's the situation it was designed for: choosing multiple winners. And when it's used for that purpose there are, again, no runoffs. Using it where we currently use IRV is still a major abuse of the system. What I would instead propose is that all competitive votes should be single-ballot IRV, with a 48-hour time limit. The longer time period allows plenty of opportunity for people who've been swayed by IRC or are just indecisive to change their votes, and allows as many people as possible to vote. Since most competitive votes end up with three polls, this should still ultimately reduce the amount of time CAPs take by a couple of days. Single-ballot IRV is used in the real world to choose governments. Do we really think that CAP movepools are so much more important that we need to do extra runoffs just to be sure?

I seem to recall that the not doing runoffs is the reason voters are required to rank all options for TL/TLT polls. It's actually not that big of a deal in my opinion to do the same for competitive polls, but I also think that it's not that bad if people choose to potentially waste their votes by not doing so. Perhaps put a note in the OP pointing out that voting for a single option does mean that your vote essentially doesn't count if that option isn't the most popular.
 
Last edited:

Bull of Heaven

Guest
I seem to recall that the not doing runoffs is the reason voters are required to rank all options for TL/TLT polls. It's actually not that big of a deal in my opinion to do the same for competitive polls, but I also think that it's not that bad if people choose to potentially waste their votes by not doing so. Perhaps put a note in the OP pointing out that voting for a single option does mean that your vote essentially doesn't count if that option isn't the most popular.
This part actually does look problematic to me, and is the only reason I'm unsure about PBV. Telling the average CAP user that they should rank every option or their votes might be worthless is asking for a lot, especially in complicated stages where there's a lot of information to process. Consider how long just one movepool submission can be, and remember that this discussion is taking place between the small group of the people that are the most dedicated to CAP. Participation from the masses is important to the health of the project, hence all of our discussions about optics, and what you're suggesting does seem to me at first glance to create a potentially unreasonable barrier to entry. I'll probably skip more votes than I do now if this is implemented, and I'm on the PRC. I would expect worse from casual users. On a related note, there will undoubtedly be plenty of people that don't list every option despite the warning, and given that, I don't see how this is consistent with jas' argument that PBV will more accurately reflect what the community wants. Before anyone suggests that this effect is actually good because it weeds out uninformed votes, I'll add that this isn't necessarily the case. It's not too much work to come up with a few informed priorities (e.g. no Acrobatics) and pick a few options based on those; it's much more complicated to weigh every option against all others. Decreasing participation is bad, and so is possibly discouraging informed voting.

On MBV: I share jas' dislike of it, as I'm never comfortable with a vote in which my actual priorities are barely represented. I don't know a lot about voting systems, and I haven't given this the kind of analysis that he has, but it's always intuitively seemed wrong. I wouldn't miss MBV if it were eliminated.

On slate sizes: My inclination is not to limit them. I understand the concerns about large slates, which is why I don't complain about our extremely strict rules for flavour polls, and I'm aware that a limit would improve the situation I described in my first paragraph. That said, limited slates aren't consistent with my views of what the TL/TLT is: a form of quality control that keeps the process from going off the rails, and not much else. Pwnemon's ideas about slating sound like the right ones to me, and I don't think that it's appropriate to force the TLT to make tough choices between quality submissions. I also have a question about this idea: if there were a limit on slate size, and the section leader reached that limit, would the TL be able to use the +1 power? I'd be interested in hearing about any other mechanisms people have to discourage large slates, but I don't think that an actual limit isn't something I can support.
 
Last edited by a moderator:

Nyktos

Custom Loser Title
This part actually does look problematic to me, and is the only reason I'm unsure about PBV. Telling the average CAP user that they should rank every option or their votes might be worthless is asking for a lot, especially in complicated stages where there's a lot of information to process. Consider how long just one movepool submission can be, and remember that this discussion is taking place between the small group of the people that are the most dedicated to CAP. Participation from the masses is important to the health of the project, hence all of our discussions about optics, and what you're suggesting does seem to me at first glance to create a potentially unreasonable barrier to entry. I'll probably skip more votes than I do now if this is implemented, and I'm on the PRC. I would expect worse from casual users. On a related note, there will undoubtedly be plenty of people that don't list every option despite the warning, and given that, I don't see how this is consistent with jas' argument that PBV will more accurately reflect what the community wants. Before anyone suggests that this effect is actually good because it weeds out uninformed votes, I'll add that this isn't necessarily the case. It's not too much work to come up with a few informed priorities (e.g. no Acrobatics) and pick a few options based on those; it's much more complicated to weigh every option against all others. Decreasing participation is bad, and so is possibly discouraging informed voting.
First off, there's no reason that PBV should require people to rank all options. Note that I'm not supporting PBV; I'm supporting keeping IRV and eliminating runoffs. You seem to be mixing up my proposal and jas's.

I do agree that people are likely to skip votes if they are required to rank all options, which is which I don't favour that option. (I probably should've made it clearer in my post that while I do think it's not the end of the world to do that I think it's worse than the alternative.) I mentioned the possibility only because we currently do that in the TL/TLT polls which are presently our only polls without runoffs. I suggested making a note in OPs because I want people to be aware of the fact that not ranking all options means that their vote potentially "doesn't count". Remember that this is already the case in IRV; all that changes in my proposal is that winning an IRV poll matters (much) more, and thus having your vote "not count" arguably does too.

Ultimately I think that if people can't or don't want to rank all the options on the slate, it's okay that their vote might not count all the way through. Effectively, if you don't rank every option what you're saying is "these are my top N in order, and I don't care who wins if they're eliminated". I don't see anything wrong with doing that provided that you're aware that that's what your vote signals. If you take the position that any movepool containing Peck will automatically ruin the CAP to the extent that the rest of the movepool is irrelevant, there's no problem in abstaining from voting in the case that all non-Peck movepools are eliminated.

Also do remember that I'm proposing doubling the amount of time you have to compose your vote.
 

Bull of Heaven

Guest
Nyktos: Thanks for the clarifications. I clearly hadn't read your first post carefully enough. I still have some (admittedly smaller) concerns about what your proposal asks casual users to do. It's one thing to say "I don't care what moves on to the next poll if these are eliminated," and another to say "I don't care which of the other options wins the whole thing". Again, I suspect that this is a bigger deal to a casual user than a PRC member, since going through every option in one big poll (remember that I don't want to restrict slate sizes) looks more daunting than figuring out some early priorities and adjusting them over the course of multiple polls, unless you're already very committed to this project. Your proposed extra time to compose a vote does mitigate my concerns, and maybe it could ensure that most of the competitive votes we actually want are still cast, but I still don't think I like the optics of this proposal.
 
One major relevant point in this topic is that every voting system is flawed in some way. Every voting system is vulnerable to some kind of manipulation, and/or fails to elect a candidate who fits certain criteria of who "ought" to be the winner. So pointing out that a voting system we're using is flawed doesn't really do anything. If we used that reasoning to change the polling system we have, we'd be changing it every time this topic is brought up.

It would be helpful to establish clear goals for what we want to achieve with CAP voting. I don't know every reason behind every decision that led to the current system. Nonetheless, from what I can tell, these are the main purposes of the CAP polling system:
  • To represent community consensus as best we can (obviously).
  • To gather information for the purpose of analysis or something else (like the CAP art gallery).
  • The spectacle: People like the act of participating in these polls in itself, and speculating on the results.
In particular, having multiple runoff polls helps in achieving all three of these. By using multiple voting systems, we gather voter information from different angles, we reduce the vulnerabilities of each particular voting system, and of course we increase the spectacle of the event. This, as far as I can tell, is the motivation behind doing multiple polls per stage. I believe Doug talked about this in one of the post-CAP 4 PR threads, but I can't find it right now.

I actually like multiple bold voting (more commonly called approval voting) a lot. It's simple to explain and tally, so unlike IRV there's far less of a sense that there's some kind of arcane voodoo magic going on. I also think that it gives us a pretty strong indication of what will and should win, in terms of picking a group that contains a rightful winner, especially since we do a head-to-head at the end. This isn't to say that I think that IRV is inferior or that IRV should be eliminated in favour of MBV, but I do feel like, if anything, there should be more MBV polls.

I also don't think that large slates are inherently bad. Everybody has strong opinions and a lot of us like to predict in advance that a certain choice will be an utter disaster. Yet, how many people can honestly say that they can predict such things in advance, and how many people would agree that they were right in the first place? Nonetheless, I wonder whether the TL and the section leader really should have different methodologies for making/modifying the slate.

I also kind of wonder whether we're not cutting options strictly enough between polls. In MBV polls, when I considered advancing options that had more than some fraction of the vote like 1/2 or 1/3, the resulting slate for the next poll was often smaller than what other mods would suggest. I don't think there's been any instance where an eventual winner wasn't from the smaller slate I had in mind from using such a criterion. The second art polls for Mollux and Necturine are particularly vivid examples of this; in both cases, Mos-Quioxte's entry, which was the only one that had majority support in the first poll, wiped out its competition, and it seemed as if I was the least surprised out of everybody. There's a point where it seems like we're wasting time on conclusions that we're not quite ready to admit are foregone, and thus creating a spectacle out of sheer ignorance. Sure, people change their minds and stuff, but I suspect that that is not a very major phenomenon. The occasional differences between IRV results and final poll results can probably be explained mostly (if not entirely) by the fact that IRV is not a Condorcet method, i.e. there's no guarantee that the IRV winner would beat all of its competitors in a head-to-head.

That may or may not be everything I wanted to put out there...
 

Nyktos

Custom Loser Title
Sure, people change their minds and stuff, but I suspect that that is not a very major phenomenon. The occasional differences between IRV results and final poll results can probably be explained mostly (if not entirely) by the fact that IRV is not a Condorcet method, i.e. there's no guarantee that the IRV winner would beat all of its competitors in a head-to-head.
While that's true, it doesn't matter here: conventional runoff voting is not a Condorcet method either. The IRV winner does not necessarily win a head-to-head against all other candidates, but does win one against the second-place candidate. Similarly, the third place candidate comes third in a contest against the first and second place winners. As such, if we advance three and nobody changes their votes, the third-place candidate in the IRV will be eliminated in poll two, and the IRV winner will win the final poll.

Assuming a frictionless vacuum and perfectly spherical voters, advancing the top 3 (or indeed top N) candidates from IRV into conventional runoffs is exactly equivalent to re-running the final (N - 1) steps of the IRV algorithm very slowly.

In any case, looking at your three points: increasing the time limit should help with the first and I'm really not sure how what we do is better than straight IRV for the second, but I guess I can't deny the spectacle argument. It's always kind of been my suspicion that that's the main reason we do multiple polls. I really don't like it as a justification for our method of choosing important parts of CAPs, though.
 
Last edited:
The key point is that in IRV you don't have to put every candidate in your ballot. So when you redo the final runoff or two, you'll have people who voted for none of the options, yet will vote in these runoffs anyway. I suppose that's the reason I should have given rather than the Condorcet thing.
 

paintseagull

pink wingull
is a Top Artistis a Forum Moderator Alumnus
Multiple Polls:
The key point is that in IRV you don't have to put every candidate in your ballot. So when you redo the final runoff or two, you'll have people who voted for none of the options, yet will vote in these runoffs anyway.
I actually quite like this feature as a voter. It is often very daunting in a 7-option IRV poll for competitive, complex options with pros and cons to rank everything, or to even be sure of my first vote. We are choosing things like movepools and stats here, not simply candidates, as in political elections or even TL/TLT elections. I like the chance to see what the community likes and dislikes, and to see the major camps solidify and rally behind a single option. This makes it easier for me to decide my vote in later polls. So while I like the PBV/forced ranking for TLT, I wouldn't want it for competitive polls. I don't support Nyktos' proposal, I think we are good with the IRV to Single Bold series of polls.

Of course a person with a platform has pros and cons too, but they also have many months to campaign and a personality to bring it all together.
----------------------

PBV:
Not sure where I stand on replacing all IRV with PBV. I have to think about it more. My initial thought is that it's a bit overcomplicated (to understand, not to execute) and unnecessary.

BTW, Pwnemon, we already have a script for this and it works fine for both TLT and regular competitive polls that are normally IRV
----------------------

MBV and slate sizes:
Bull of Heaven's post about +1 power is a good point against limiting slate sizes. I know that our general rule of thumb is to reduce large polls down to about 7 options before doing IRV, but is there any reason that we can't do IRV on 8 or 9 options? Seems to me that there is really no reason not to, aside from it being more difficult to vote the more options there are. A TLT's slate should consider intelligent community support but also strive for the simplest poll. Both of these considerations are community-minded but conflict with each other, naturally constraining the poll size.

With MBV itself my biggest concern is actually the ratio of the number of voters to the number of options. For example, in smaller Prevo polls with MBV it can be very hard to see where to cut the slate for the next poll because significant differences are not apparent. This might be a problem in bigger polls too. CAP 6's Movepool Poll 1 had a 10 options MBV with only 41 votes. CAP 6 - Part 10 - Movepool Poll 1 The difference between 12 and 8 votes is really quite small, and that leads me to question MBV's effectiveness. On the other hand, CAP 6's Art Poll 1 had 23 options and 170 votes. CAP 6 - Part 8 - Art Poll 1 Community preferences are very obvious to me here.

I'm not sure what my conclusion is on this yet, I think we need more discussion.
 

jas61292

used substitute
is a Community Contributoris a Top CAP Contributoris a Forum Moderator Alumnusis a Battle Simulator Moderator Alumnus
When I wrote this OP I was making suggestions just to slightly adjust the current system as far as which polling types are used when, and not to completely change how we do polling overall. However, I have always found some of the practices we use rather odd, and to be quite honest, I am 100% with Nyktos in that I think a single IRV would be a vastly superior than what we do now for any competitive poll. In my opinion, what we do now makes absolutely no sense, and don't really help out anyone. Just to reiterate, we have been using methods that serve to eliminate runoff votes, and then running runoffs anyways. That's ridiculous if you ask me.

And why do we do this? What gain does it provide? Sure, people who didn't vote for them all to begin with can have their opinion heard. Except, that ignores the fact that they already had the chance to express their opinions, and chose not to. If they have enough of an opinion to vote in the final bold vote, then they had enough of an opinion to vote in the IRV. If they didn't vote for them all in the IRV, that is no one's fault but their own. Like, I get that deciding between a bunch may not be easy, but if you are able to rank a few for IRV and then are also able to decide between the final few in a bold vote when its only between ones you didn't vote for in the IRV, then you could have shown your opinion on those in the IRV. Either you can't decide or you can. I personally have no sympathy for people who can only decide on their least favorites once their favorites are gone. You either have a preference or you don't. There is no in between, and we should not be shaping our polling structure for the benefit of people who claim to fall in to this non-existent in between area. If we simply let people know that this is the case, and give some more time, then more people will probably vote for them all, and THAT would actually be more advantageous.

Also, in response to what capefeather brought up about the potential influences for the voting system, while I cannot say what went into creating the current system, I can say that the whole "spectacle" thing should absolutely not have anything to do with what we decide here. The excitement and speculation should derive from the poll; the poll should not derive from them. The length of the poll can always be adjusted to provide that same suspense or whatever. A good single IRV can provide all the same advantages as what we currently have, in my opinion.

Less polling also means less gaming of the polls. IRV is more complicated than bold voting and makes it harder for the average person to gauge where the voting stands. This makes strategic voting and bandwagoning less likely than in a single selection poll.

To be quite honest I see absolutely no reason not to go with a single IRV poll for competitive polls. However, with that said, our current method is so absurd to me that I get the feeling that there may be some not-so-obvious reason why we do this. If anyone who was around for and was a part of the decision making process where we decided to use the current system can shed some light on why things are the way they are, I think that would be important to bring up before we make any decisions.

If, of course, there is are real, compelling reasons for the multiple poll setup we have, I stick by what I said in the OP that I believe MBV serves no place in competitive polls, and the PBV is better than IRV for that setup.
 

DetroitLolcat

Maize and Blue Badge Set 2014-2017
is a Forum Moderator Alumnusis a CAP Contributor Alumnus
I've only got a quick point to make here, and that is that we should use voting methods as they're meant to be used. IRV, or instant-runoff voting, is intended to simulate a single-bold vote as well as all subsequent runoffs. It is a single-winner system; IRV is intended to produce a single solitary winner. IRV does an extremely poor job of creating a slate for a runoff election because IRV takes the most popular option of the entire community, then the most popular option out of those that did not prefer the most popular option, then the most popular option out of those who did not prefer either of the top two options, etc. Instant-runoff voting's goal is to find the most popular and only the most popular option, and it does a damn good job of doing so. For example, if a slate has 7 options, then IRV allows us to conduct single bold voting plus 5 single-bold vote runoff elections in 24 hours. We could achieve the same results from an IRV poll as we could from 6 Single Bold Vote polls, removing the last-place option each time. Obviously we prefer the former because there is no point in polling for six days when we can we can achieve equally legitimate results using one poll. IRV works because by the end of the poll, a majority of the community voted for the winner. CAP has always used majoritarian voting systems and IRV is the best way to expediently determine the single option a majority of the community prefers to all others.

Remember, IRV does not try to rank the options in terms of popularity. The only conclusion we can draw from an IRV poll is that the option that comes in first is the most popular option. Coming in second in an IRV poll does not mean that an option was the second most popular.

Notice that Preferential Block Voting (PBV) does not just list the options in the order that they finished in the IRV. When we do PBV, we take all of the options and run an IRV. We take the winner of that IRV and then run the IRV again with the winner's votes removed. This works because every winner receives a majority of the votes cast. PBV can be understood as iterated IRV; only the winners of the IRV matter. That is because coming in second in an IRV poll is meaningless when IRV only intends to produce a single winner. This isn't to say anything about PBV, it's just an example as to why second place in an IRV is meaningless. If you take anything out of the previous three paragraphs, it's that IRV should not have runoff elections because IRV intends to find a single winner and nothing else.

PBV, currently, has been used in two different polls: the CAP 6 Movepool Poll 2 and all TLT election polls. Preferential Block Voting is a multiple-winner system; it intends to determine the X most popular options. It does not discard anyone's opinion for places other than first like the current multiple-winner IRV does. PBV is appropriate when we want to select more than one candidate. This is the best way to conduct a TLT election because we have 4 seats of equal importance to fill. In fact, the TLT election is the only election in which we need multiple winners and, consequentially, the only election in which PBV is absolutely necessary.

The current system is for the TL poll to be a single-ballot IRV, the TLT election to be a Preferential Block Vote, and competitive polls to be some odd combination of Multiple Bold Vote, Instant-Runoff Vote, and Single Bold Vote. I'm going to leave MBV alone for right now and look at the reason as to why we currently use the extremely awkward IRV+SBV+SBV for most competitive polls. The reasons are found in this PR thread from January 2011. These two excerpts capture the reasoning behind the implementation of the IRV+SBV runoff system.

tennisace said:
In regards to your points one and three, IRV is designed to get a winner instantly. That said, I wouldn't mind IRV in the first poll and then taking the top two choices for a one on one vote in the second round, that seems quite fair since in CAP opinions are known to change.
Rising_Dusk said:
IRV is something that has my support, although I definitely like the idea posted later in this thread of picking the top two and having a normal poll between those at the end. I think this is really important because it keeps to the spirit of the competition of CAP and allows people to rally behind one or another, etc. Basically, it just makes it a lot more fun with no cost to the process or the mods. I think we should do it.
I was not a PRC member in 2011, so I will not comment on the validity of these ideas at the time. It's safe to say that the IRV+SBV system was implemented in order to:
1) Allow for people to change their minds over the course of the polling period
2) Maintain the competitive nature of the polls and provide a spectacle to watch.

In 2014, minimizing drama and curtailing the CAP process duration are two of the most important issues facing the PRC and the community. Although this system had good intentions, we should not stick with a system that was intended to draw out the process to accommodate fence-sitting and manufacture drama. Plus, discussion of a topic closes before the poll opens. Should we really be accommodating mind-changing after discussion has closed? If no discussion takes place between Day 1 of polling and Day 2 of polling, what makes someone's opinion on Day 2 any more valid than their opinion on Day 1? Since the only new public information between the two days is the result of the first poll, multiple polls allow for increased bandwagoning and gamesmanship. Consequently, in my opinion, fewer polls are more efficient and legitimate than more polls.

Multiple Bold Voting does have its use, but only for flavor polls. Since we slate all legal options for flavor polls, we can expect to have a lot more candidates for flavor polls than competitive polls. We have a TL and TLT to cull the slates for Typing, Abilities, Stats, and Movepool, but we do not have any culling procedure for Art, Sprites, Name, and Dex Entries (as we shouldn't!). It's hard to construct an IRV ballot for 20+ candidates, and we can use MBV to cull the slate to 7 (or any other number we choose). It is easier for our voters to make a MBV ballot than an IRV ballot.

In conclusion, I believe there are two ways to move forward with competitive polls.

1. We make all competitive polls Instant Runoff Voting with zero runoffs. We could run a one-winner Preferential Block Vote here if we would like to see the options ranked by popularity.
2. We make all competitive polls Preferential Block Voting with one runoff between the top 2.

The former drastically improves the time factor, while the latter allows for people to change their minds during the polling period. I'm more partial to the former, but the latter certainly has its benefits.
 

paintseagull

pink wingull
is a Top Artistis a Forum Moderator Alumnus
I think we should consider whether or not we like the variance that multiple polls brings. For sure, when we do many polls, we do not get the same set of people every time. In addition to this, people are able to change their minds.

Consider today's prevo poll, run from approx. Sun 3pm to Mon 3pm EST, with the following results:
17 healndeal
8 darkie

Now look at the previous 4-member single bold poll, run from Fri 5pm to Sun 12:30pm EST:
13 healndeal
11 darkie
7 px
2 toatahu

When voting numbers are low like this, and especially when one poll accidentally runs long, you can end up with huge variance in votes for the same entry. Darkie's vote dropped by 3 (almost 30%), when all things being equal you'd expect it to stay the same at worst. My guess was that this was simply due to the other poll being open longer and during at a time with higher forum traffic, and darkie's voters happened to be folks who weren't around or forgot to check back.

If you imagine repeating this scenario multiple times but changing the voting times and lengths, especially on small voting turnout like this, I would guess that the cross section of people we'd poll would be so different each time that you could plausibly change the winner.

The best way to reduce this effect is to run one poll over a longer time. Say single round IRV for 48 or even 72 hours.

Do we want to reduce this, or do we think it's a fun/exciting effect of having a large community participating via forum?

Personally I'm thinking I'd like to reduce it, so for that reason and the reasons put forth by Nyktos and DLC, I'm leaning towards favouring single round IRV for competitive polls.
 
To reply to some of the points made:

I'm kind of amused by how much attention the "spectacle" point got when it doesn't really affect the rest of what I said at all. Even though I brought it up, I do see it as nonessential and kind of a bonus. The multiple poll system is indeed too convoluted for that purpose alone. On that note, I don't like the idea of using IRV to intentionally obfuscate the process of choosing the winner. It just feels like the same kind of convoluted reasoning. I'd much rather focus more on choosing a voting system based on resistance to manipulation and immunity to certain really undesirable types of manipulation. Another alternative is to do secret ballot posts like the suspect test voting system, though there still seems to be too many permissions issues for that to be realistic.

I'm also kind of puzzled at how we seem to be trying to decide on voting systems. As far as I can tell, the only major reason not to like approval voting is that it's kind of susceptible to "bullet voting" (you drop upvotes to increase the odds for your other upvotes), which doesn't even seem to be the worst susceptibility. It's not like IRV is perfect. There are people out there who'd gladly tell you why they like approval voting much more than IRV, heh... We could go with Nyktos's suggestion and have options cross a 1/3 or 1/2 approval threshold to advance, or we could just pick the winner from it. I'm not even saying that we absolutely should go with approval voting. I just wonder how much we're really thinking about this, and whether there isn't a voting system we haven't even used before that might be more appropriate.

Onto my main point:

In general, I like the idea of having one IRV poll. However, I'm not sure if we should directly pick the winner every time. I get what jas is saying, but I don't think we can really say that we can fully account for all the reasons that someone might not fill out his/her entire ballot. One possible alternative is to have the PBV #2 fight the #1 if the #1 doesn't have a majority in the last runoff. Maybe even repeat the process for a possible #3, etc. We might need to do more PBVs and/or analyze past IRVs and PBVs before we really decide on this kind of thing, though.
 

jas61292

used substitute
is a Community Contributoris a Top CAP Contributoris a Forum Moderator Alumnusis a Battle Simulator Moderator Alumnus
Another alternative is to do secret ballot posts like the suspect test voting system, though there still seems to be too many permissions issues for that to be realistic.
Just a quick point. A little while back ginganinja brought up on IRC the idea of blind votes (more related to bandwagoning than manipulation or anything). Personally I expressed concern as it seemed that it might require a separate subforum or something, but ginga asked Joim if it would be possible to do what is done in the blind voting subforum on a thread by thread basis, and he indicated that it would be possible, and that he would be willing to implement it (unless Doug would prefer to, him also being an admin who can do that kinda stuff) if we wanted to go that route. Whether or not we do want to do that, I'm not really sure, but I'm just posting this to say that someone who knows stuff said it would be possible, so if people think it is a good idea, we don't have to throw it out for feasibility reasons.
 

Birkal

We have the technology.
is a Top Artistis a Top CAP Contributoris a Top Smogon Media Contributoris a Site Content Manager Alumnusis a Battle Simulator Admin Alumnusis a Senior Staff Member Alumnusis a Community Contributor Alumnus
This has been brought numerous times in the past outside of the Policy Review Committee. The last time I can remember talking about it with DougJustDoug was about a year ago when jc104 brought it up. Although I cannot speak for the entirety of CAP's history, there were certainly logical reasons for setting polls up the way we have them today. Those reasons are largely mentioned in the posts above, so I'm not going to cover them.

I think there is a lot of wisdom to our current polling options. That being said, the logic behind Nyktos' proposal is sound enough, and it seems to be well-received by the CAP PRC as a whole. While I'm not convinced it is a better alternative method, I'm not opposed to giving it a test run for CAP1. We are already sort of "testing" it with DHR's current pre-evolution process, and it doesn't seem to have any major hiccups. If we're all in agreement here, let's do a single round of IRV for all CAP 1 competitive polls. Let's talk it over after CAP 1 and see what we need to tweak or revert.

Note the implications of this. The slate will need to be pretty small in order for this to work, so I'd recommend the CAP TL choose a slate of around six at maximum for every competitive poll. If the TLT decides to add an option, then seven is still a manageable number for one round of IRV. I'm not a huge fan of increasing the time limit to 48 hours. It slows down the momentum (especially in those last 12 hours) and doesn't add that many more votes. I don't think it's unreasonable to start expecting more from our CAP voters. It might help to eliminate some of the "drive-by" voters as well.

That being said, all non-competitive polls will be determined by the CAP moderators. Art Polls will still probably retain their full length, since that is our biggest spectator sport. Others such as Pokedex and Name are sort of irrelevant in terms of time due to the alignment with the playtest.

Let's move forward with this process. If there are any final arguments or discussion points, let's have them. But if we're all in agreement here, let's give it a test run. Good suggestion, Nyktos. Oh, a final thing we might want to discuss is whether or not we require our voters to rank every single option. I think if we're doing one round of IRV, that is an implied rule that we'll start explicitly stating at the start of polls. But if the PRC deems ranking all of the slate as unnecessary, then I suppose we don't need to. Let me hear your thoughts.
 

Birkal

We have the technology.
is a Top Artistis a Top CAP Contributoris a Top Smogon Media Contributoris a Site Content Manager Alumnusis a Battle Simulator Admin Alumnusis a Senior Staff Member Alumnusis a Community Contributor Alumnus
Oh and ginganinja, blind voting is definitely a possibility with Xenforo, yes. I think I even have the administrative capacity to set up such a system! However, let's take this one step at a time. Removing multiple rounds of voting is a huge change-up. Let's review that recommendation after CAP 1. If we implement both and discover difficulty within the process, it might be hard to pin down the root of our problems.
 

paintseagull

pink wingull
is a Top Artistis a Forum Moderator Alumnus
We are already sort of "testing" it with DHR's current pre-evolution process, and it doesn't seem to have any major hiccups.
The only thing from this discussion that's affected the prevo so far has been that we just decided to use PBV instead of IRV on this most recent art poll, but that was in the on-site rules already. So, I'm not sure what you mean by this.

There's nothing about single round IRV that says we can't have more than 6-7 candidates, the only reason to limit it would be to not put an over-large burden on voters, but I don't think the extent of that burden, its pros and cons, or if it even exists has been well established in this discussion.

I have more to add, will do so later.
 

Nyktos

Custom Loser Title
I'd imagine the suggestion to limit slate sizes relates to the idea that everyone will have to rank all options. I'd like to reiterate that I see no real advantage to forcing people to do that. If you fail to include all options, there's a possibility that you "don't vote" in some of the later virtual runoffs, but we don't currently force people to vote in runoffs so I'm not sure what we gain by doing so here. As long as people are aware of the possible consequences of not listing all options I don't think it's a problem if they make an informed decision that they really don't give a rat's ass who wins if all the options they like are eliminated.

That said, if there's a real worry that voters listing only one or two options will lead to inaccurate results, how about requiring that voters rank at least five options? (The exact number is up for debate, obviously.) This should still do a fair amount to mitigate the issues of voter numbers dropping off in the late rounds, but doesn't make large slates infeasible.
 

paintseagull

pink wingull
is a Top Artistis a Forum Moderator Alumnus
To briefly summarize (for myself, at least, forgive me if it's too redundant):
Proposal - No future concept, typing, stats, primary ability, secondary ability or movepool poll will ever use Multiple Bold Voting
- we seem to be in agreement here

A) if using IRV or PBV for large slates is acceptable
B) whether or not decreasing slate size is desirable
C) positives and negatives surrounding implementing limits on slate sizes
D) any alternative ways people may come up with of reducing the slate sizes without hard limits
E) any alternative ways people may come up with of fixing the problem of MBV in competitive polls that don't involve reducing slate size or using IRV/PBV
- Points A and B have not really been addressed. Hard limits keep getting brought up but nobody seems to really love the idea of them.

Proposal - Preferential Block Voting will replace Instant Runoff Voting in all standard CAP polls
- again, I think we agree, that if we use a ranked system followed by a runoff, the ranked system should be PBV, not IRV (in fact this is already part of our on-site rules but never became common practice). This seemed to work well in the most recent Prevo art poll, as our top 2 from PBV resulted in a pretty tight race. IRV would have given a different top 2.

The victory condition in multiple bold vote polls should be changed - all options supported by at least one third (rounded down) of voters move on.
- Tried this in Art, it seemed to work fine. I think we have enough decisions to make and probably won't make this a hard and fast rule, but rather another rule of thumb, at least for now. I will probably try it again next time and I'd encourage other mods to do the same.

All competitive votes should be single-ballot IRV.
- So that brings us to this, and Birkal says we're good to try it out. Before we move forward with it though, we need to hash out the stuff that I don't think we're in agreement on, and that is slate sizes (basically points A and B previously), whether to force voters to rank a certain number of candidates, and the time limit on the vote.

There doesn't seem to be anything about IRV itself that doesn't work for larger slates. I've seen examples explaining IRV that have 8 candidates. I have looked for idealized limits to IRV or places where it breaks down but I haven't been able to find anything on that yet. My gut instinct says that forcing voters to rank 3 out of up to 8 candidates should probably be fair and not a burden on voters at all, but then moderators have to make sure every vote has at least 3 choices.

I think that if we're doing a single round vote we should extend to 48 hours. It might be a bit over-long but 24h is too short, and single round 48h still saves time on multiple polls as we do them now. Better to cut voters off as they are slowing to a trickle, than when they are coming in fairly steadily, in my opinion.

All that said, a 2-rounded system of PBV -> 2 choice SBV is also much tidier than what we're doing now, maybe we could keep it as an alternative if we have a large slate of 7+? I'm not sure if this would ever yield a different result than single round IRV, I suppose it really depends on how many options voters choose to rank. So such a 2-rounded system could be used if we anticipate voters won't rank enough of the options on the slate (assuming we don't force them to rank a large number).

Definitely not on the same page as Birkal as just leaving flavour polls the way they are, by the way, I think these discussions should inform flavour as well, but we can keep MBV for those.
 

Birkal

We have the technology.
is a Top Artistis a Top CAP Contributoris a Top Smogon Media Contributoris a Site Content Manager Alumnusis a Battle Simulator Admin Alumnusis a Senior Staff Member Alumnusis a Community Contributor Alumnus
Discussed this with paintseagull, DetroitLolcat, and bugmaniacbob for quite some time on IRC. I think the four of us came to some sort of a consensus, but I'd like to know what the rest of you think about using the following procedure for CAP 1's polling options:

All Competitive Polls:

1) Preferential Block Voting (Round One)
- Consists of voting on anywhere between 6-8 options.
- Voters may rank as many or as few options as they like.
- Thread always lasts for 24 hours.
2) Instant Run-Off Voting / Single Bold Voting (Round Two)
- Consists of voting on anywhere between 2-3 options.
- Decided based on gaps (how CAP mods currently handle it).
- Will likely always be three options, unless there are two obvious leaders.​
- Two options will be SBV.
- Three options will be IRV. Here, voters must rank all options.​
- Thread always lasts 24 hours.
- The final winner is always decided at this stage.
This definitely isn't the perfect system, but it runs the gambit of being both timely and well-informed. PBV allows users to give a good ranking to all of the presented options, but doesn't require it. Round Two is efficient in that it will only last 24 hours, regardless of the second slate. This leads to minimal poll time loss. Ranking three options isn't overbearing on voters. The beauty of this system is that all votes for any competitive stage will be completed in forty eight hours. That's awesome, in my opinion, and will help the project keep its momentum. It's alright to miss a day of voting (since there are two), but there isn't the overbearing pressure to rank an entire set of eight options in a single poll.

I think this addresses most of what paintseagull says in the post above me. Let's see what the rest of the PRC thinks of this model. If we're in agreement, then let's discuss the flavor implications of this. Thoughts, everyone?
 

Nyktos

Custom Loser Title
I fail to see any real way that this proposal is actually an improvement over current practices, aside from the use of PBV. Having an IRV(-variant) poll followed by a runoff that is itself IRV manages to feel even more bizarre than IRV followed by conventional runoffs, though from a theoretical standpoint it's actually the same thing. (And requiring voters to rank all three makes practice very close to theory here.) I also really dislike the fact that the number that move on is decided based on arbitrary mod judgement, but again that's already the status quo. If the mods want to shorten the time polls take, that is of course their prerogative, but adopting this change is not substantially different from making no change at all.

The heart of this seems to be a worry that single-round IRV makes polls less "accurate" because voters won't rank all options, and forcing them to rank all options is also unpalatable for whatever reason. (I basically always rank all options already and I've never found it that difficult...) I suggested a way to mitigate this issue (must rank N options for some N) which nobody actually argued against in-thraed, so I'm not sure why the whole thing has been completely dismissed. I'm apparently missing some IRC context (I've had computer issues lately so haven't been on #cap much) so if there are logs explaining why my suggestion is bad and we need to backpedal to essentially what we already have, it would be nice to see them. As it is, from what I can see this new proposal appears to be coming out of nowhere.
 

paintseagull

pink wingull
is a Top Artistis a Forum Moderator Alumnus
I don't think we are backpedalling with Birkal's suggestion. Our discussion has been productive, and ensuring that we have a logical, well thought out reasoning behind our voting process is just as important as what that process ends up being. We've ditched MBV for competitive polls and we will no longer allow IRV results to be used to create runoff polls. To me these were the biggest issues.

I'll summarize my argument for the two-round system because I agree that the thread has not really tracked how this came about:
You are right that single-round IRV is unappealing to me because I do worry that it will be far less accurate in gauging the community's opinion unless we force voters to rank a significant fraction of the options. Any extra rule or restriction hampers voters and makes the action of voting less appealing, which is certainly not a good thing in my opinion. From my interpretation of CAP's philosophy, we want the most votes possible from our community. For movepool and stats, ranking even 5 options requires a lot of time and effort, and I would hate to lose someone's opinion about their top 2 because they couldn't decide between their next 3. The benefit of a two-rounded system is that if that someone's top 2 don't make it to the second poll, they still have the opportunity to have their opinion heard, and the task of ranking 2 or 3 options against each other is much simpler. For abilities and typing I can see forcing N working, but we still have the task of moderating this rule to contend with, and throwing out votes really sucks. In most cases, I think our PBV winner (aka our IRV winner) will end up being our final winner, but the second round provides insurance without being a senseless continuation of the voting process for the spectacle's sake.

My opinion that forcing N choices on IRV will reduce voter turnout or be overly hard on voters is purely speculation based on my own experience attempting to rank movepool and stats in the past. That this is a negative thing is also my personal opinion. So if those things aren't agreed upon by everyone, then I'm definitely still open to single round IRV, and I think Birkal is too.

----

As a slight edit to Birkal's proposal, I'd add the contingency of only needing one round if the first count of the top favoured options yields a majority winner. (ie the very first paragraph on any PBV or IRV counting script output, before anyone is eliminated from the count)
 
Last edited:

jas61292

used substitute
is a Community Contributoris a Top CAP Contributoris a Forum Moderator Alumnusis a Battle Simulator Moderator Alumnus
While I don't highly dislike Birkal's proposed system, and do see it as at least a small step in the right direction, I am completely with Nyktos on this one. I fully believe that having a runoff vote at all makes little sense when the entire voting system is designed to pick winners immediately, without runoffs. Obviously, there is the issue that it may seem less accurate if people don't vote for all options, but I fail to see how that is really different at all than we have now. Yes, if you only vote for 2 in a system like this and they both get eliminated, you have the option to vote out of the remaining options, but it was the person's choice to only vote for those two to begin with. Also, people miss votes all the time. Voting for 2 out of 5 and then skipping the second vote if they both are eliminated is completely identical to just voting for two in a single round IRV.

Moreover, the biggest problem I have with the pro- two round argument is that it won't suddenly become easier to decide between two entries because ones you liked more are gone. It just doesn't work like that. Either you have an opinion between two options, or you don't. Again, if it is a time constraint thing, then we can make a single vote last longer. What we can't do is make people actually have an opinion where they do not have one in the first place. In my opinion, the idea of people voting just to vote when they really don't care is more harmful than anything a single round vote coule bring to the table. When people who don't have an opinion vote, they will be much more subject to things like bandwagoning than normal.

We are a community project, yes, and we do want as much of the community to vote as we can. However, we want them voting because they have opinions. Voting for voting's sake is not any sort of advantage whatsoever, and that is the only type of voting that I believe a 2 round system will give us over the one round system.
 
Status
Not open for further replies.

Users Who Are Viewing This Thread (Users: 1, Guests: 0)

Top