MitD XIII: Learning is happening

[Jasdoif]

That feels wrong to me. If someone tells me they want A & B, and don't tell me that they want A, I don't think I should convert their vote from "A and B" to "Only A". If they had wanted A as an option, they'd have voted for it. Doing it to the Any(A, B) is fine. But doing it to the A+B is not.

That's kind of what I said (or at least, was attempting to say) in the bulleted list: If option B is removed, "A+B" cannot be met, so the entire "A+B" combination is removed from the ballots.

Speaking of the instant runoff voting, I keep trying to think of scenario where the instant-instant runoff I did (i.e. discard all combinations that weren't capable of reaching a majority) could lead to complications, and I keep coming up empty. I appreciate it is not the way it is supposed to be done by the theory, but I don't see how it would have made any difference. No matter how the votes were arranged, any combination that couldn't reach 50% of the votes was going to be discarded before the end, so by discarding them from the start and centering on the ones with a chance, and how their vote distribution ranked them, it simplified the runoff without endangering it. Is there something I am missing? And more importantly, is this getting off-topic enough we should move it to Friendly Banter or Mad Science? (I consider it more meta-topic than off-topic, but I appreciate not everyone will be interesting on sausage making)

If an option on over 50% of the ballots nevertheless has the smallest number at the top line (perhaps a very popular second-place choice, with one first-place supporter), it will be removed from the ballots. This is why IRV fails the monotonicity criterion: the option was removed because it had first-place supporters instead of second-place supporters, so rating their first choice higher was responsible for its loss. (A ballot with all its options removed will mean the entire ballot is removed, lowered the number needed for a majority of ballots; so IRV will still come up with a winner).

Only using the options that appear on 50%+ of ballots is closer to Alternative Smith. It's not invalid, just different enough from IRV (and the video explaining it) that the variation should be mentioned in advance.

I find that "I don't care if A gets voted, it's just not getting my vote" and "I definitely don't want A" are separate stances that deserve separate listings. More importantly, it is a expression of intent that came up at least twice in the last vote, so this isn't even a hypothetical, but something I had to patch in last time, and that therefore I want clearly spelled out in advance.

Now, it is true that !A is equivalent to Any(B, C, D, ...) only as long as new options are not added, and that I do allow options to be added, so technically, if someone PMs with "not A" when options are {A, B, C}, and I list them as !A, and then option D is proposed that they also don't like, the onus is on them to alter their vote at that time, while if I listed them as Any(B, C), they'd not have to do anything once D came along to also exclude it. But on the flipside, if they did like D, in that it is still not A, then with !A they don't have to do anything, whilst with Any(B, C), they'd have to change their vote to Any(B, C, D). I feel that my approach of transcribing a "I don't want A" as, literally, not A, should in principle be a more accurate representation of their intent.

Oh...so !A would be like antimatter A votes? You subtract them the count of votes in favor of A?

[Grey_Wolf_c]

That's kind of what I said (or at least, was attempting to say) in the bulleted list: If option B is removed, "A+B" cannot be met, so the entire "A+B" combination is removed from the ballots.

Again, that feels wrong to me. If there is a majority of people who only want A and B when they go together, but don't want either individually, then converting their vote to an option they didn't want feels wrong. Now, in practice, from what I remember, anyone that wanted A+B last time (for any value of A, B) did tend to then vote for them individually. But it is not a given. Voters may not want B individually, they may want B and A together or not at all, so I shouldn't convert their A&B to A simply because B got voted out - I should allow that B individually might be unwanted, but A&B together exactly what a majority feels should be done.

If an option on over 50% of the ballots nevertheless has the smallest number at the top line (perhaps a very popular second-place choice, with one first-place supporter), it will be removed from the ballots. This is why IRV fails the monotonicity criterion: the option was removed because it had first-place supporters instead of second-place supporters, so rating their first choice higher was responsible for its loss. (A ballot with all its options removed will mean the entire ballot is removed, lowered the number needed for a majority of ballots; so IRV will still come up with a winner).

Aaaah. Yes, you are correct. Damn. Fair enough, I'll be more careful with ballot resolution.

Oh...so !A would be like antimatter A votes? You subtract them the count of votes in favor of A?

No, I don't subtract a vote. Instead, at resolution, !A does become equivalent to Any(B, C, D...). A vote for !A is a vote for any extant combination that doesn't include A - i.e. it's "Anything but A". Admittedly, I think mathematically the result is the same, but I do the latter and not the former.

Grey Wolf

[Jasdoif]

Again, that feels wrong to me. If there is a majority of people who only want A and B when they go together, but don't want either individually, then converting their vote to an option they didn't want feels wrong. Now, in practice, from what I remember, anyone that wanted A+B last time (for any value of A, B) did tend to then vote for them individually. But it is not a given.

That's what (I think) I said: If they only want A and B together, and they can't get A or can't get B, the "only want A and B together" option is thrown out and their ballot goes on to the next combination in their ranking. There's no transformation with the exclusive combination, since there's no room to legitimately transform.

Aaaah. Yes, you are correct. Damn.

Yes; this is my major problem with IRV.

No, I don't subtract a vote. Instead, at resolution, !A does become equivalent to Any(B, C, D...). A vote for !A is a vote for any extant combination that doesn't include A - i.e. it's "Anything but A".

Okay...So either you'll transform that into Any(B, C, D....) when you get PM, or the voter will need to be adjust/accept if/when new options they disagree with come out. That works.

[Grey_Wolf_c]

That's what (I think) I said: If they only want A and B together, and they can't get A or can't get B, the "only want A and B together" option is thrown out and their ballot goes on to the next combination in their ranking. There's no transformation with the exclusive combination, since there's no room to legitimately transform.

OK, I get what you are saying, but I still disagree that just because an atomic component of a vote is discarded, that the compound vote should be discarded. I consider each combination independent from every other. Sure, A lost. But if A+B has a majority, why would that matter? All that tells me is that A was half of a good idea. I'd vote for more salt while not voting for more sodium or more chlorine. A combined vote is sometimes better than the sum of its parts.

Grey Wolf

[Jasdoif]

OK, I get what you are saying, but I still disagree that just because an atomic component of a vote is discarded, that the compound vote should be discarded. I consider each combination independent from every other. Sure, A lost. But if A+B has a majority, why would that matter? All that tells me is that A was half of a good idea. I'd vote for more salt while not voting for more sodium or more chlorine. A combined vote is sometimes better than the sum of its parts.

Well...if the A+B combination has a majority, then A+B is the winner. If A+B doesn't quite have a majority....then the A and B options are going to be expressed on at least as many ballots as A+B is; unless there is some other majority, neither A or B are going to be the least expressed option (so the A+B votes will endure).

That said, this was an idea to get you out of expanding all the subsets into explicit combinations (and then determining how to order them)...one I haven't done a lot of analysis on to see how/if final results are distorted. There's nothing wrong with going the full expansion route.

[lio45]

Now, in practice, from what I remember, anyone that wanted A+B last time (for any value of A, B) did tend to then vote for them individually. But it is not a given.

The crux of our disagreement is right there. You're willing to pay a very heavy price - forcing people to vote for every single one of the combinations/permutations of what they'd support, with the very real risk that some won't bother to do that or realize they have to - just to ensure that in the very rare case where someone would only support something when paired with something else, they can express that preference.

Why not just eliminate the idea of pairings, and treat any pairing that gets brought up / supported / suggested - I have yet to see an example of one of those, for the record - as its own entry? This way, there would be no more pairings ever, just individual fully independent suggestions that can either be implemented or not. (And then there would be no good reason anymore to insist on using a system (IRV) designed for circumstances where we have to democratically select one, not less, not more, option among several. We could just vote in a binary manner on any suggestion that meets the threshold for a vote whenever that happens.)

The only way this is sounding like less work to me is that you have theoretically reduced the number of votes occurring entirely by virtue of the higher barrier to entry, and that is entirely separate from your binary voting system. It also doesn't really do anything for those situations where you do have, say, 5 different proposed changes that have all been called for a vote, unless you want to make the amount of support so high as to be impossible to meet anyway (which seems to defeat the purpose of even allowing for a vote).

Except that you've already noted you have 5 or 6 things you'd like to see voted on. If you were running the vote, 9 feels like a likely number of things to have to vote on. Sure, if you set high thresholds, you cut that down, but who sets the thresholds and how are they determined? You, as the vote holder doing it arbitrarily?

It feels far more arbitrary and vulnerable to manipulation by the person running it, unless you have voting ahead of time on where to set the thresholds and other details, but that requires you to have a vote on the details (ok, how many people say 60%? How about 40%? etc). When you have to set up a committee to decide how to set up the committee, it usually means you're heading in more-cumbersome direction.

These two posts have the same point so I'll address them together.

We have an approximate idea of how many voters vote (last time, it was at least 20; another possible definition would be "the number of individual active thread participants in the last x days" or "in the last y pages"; or we could historically look back and say "the absolute minimum number of voters in thread history was 15" - which implies that unless you have at least 8 people supporting your idea it has almost no chance of passing).

This number can give us a good sense of a threshold below which there's no sense in even calling a vote. At minimum, I'm comfortable saying that if someone suggesting a modification can't even get five separate people to agree that they would want to see that one specific modification implemented, then there's no point in even going to the voting stage on that one. Kish's two suggestions, for example - unless at least five different users chime in to say "yep, good idea" for either (separately, obviously) of them, then there's no point in voting. Don't you agree?

And the nice thing is, usually those discussions would double as argumentation over the modification in question (the pros and cons of it, and what it seeks to achieve), which is something we'd do anyway.

So, Crusher, no, no way there'd be 9 things to vote on... maybe 1 or 2, tops.

[Squire Doodad]

Snip

There's a difference between what has happened and what can happen. Just because you've never had to deal with more than 15 soldiers attacking doesn't mean that you won't someday have 50 attacking you. It may be unlikely, sure, but it could happen.
This isn't a circumstance that has a formula capable of describing it, with maximums and minimums. All we can look at is see what has happened in the past- but as we all know, past examples will eventually become obsolete. You're suggestion may be functional in an event of a small scale, low number scenario. But if when it has larger numbers, it fails terribly, then the only way it can work is if it doesn't ever have larger numbers. But the issue is, you haven't actually suggested anything to handle larger numbers. At risk of using a ridiculous analogy: you've suggested increasing the distance the soldiers need to run, you've suggested making the walls bigger- but you haven't come up with any defense plan for what happens if the soldiers get into the city.

Status quo is not god, but the current status quo handles things far more reliably than your idea does.

[lio45]

Is this sarcasm? 'Cause it seems to me it's always been pretty evident that the more participants, the more of an edge a simple binary system has over the current IRV-style one. If we were only a few voters, then even with the current system it wouldn't be that much of a problem to ensure everyone expresses their preferences correctly {scrubbed}

[Squire Doodad]

Is this sarcasm? 'Cause it seems to me it's always been pretty evident that the more participants, the more of an edge a simple binary system has over the current IRV-style one. If we were only a few voters, then even with the current system it wouldn't be that much of a problem to ensure everyone expresses their preferences correctly {scrub the post, scrub the quote}

I think you're missing the point.
I'm out of energy to deal with this.

[Aspheric]

So, Crusher, no, no way there'd be 9 things to vote on... maybe 1 or 2, tops.

You do know that there were nine things to vote on in the last voting session, right? I'm not sure if Crusher made that statement with that in mind; but his claim, and the criticism of your proposed method, isn't without merit. Your proposed system seems like it provides fewer options, makes it more difficult to actually bring anything to a vote to implement changes, and has a harder time accomodating the nuance of what people actually want. I won't pretend the current system isn't without faults, but it's hard to see how your proposal fixes any of its imperfections.

[Schroeswald]

You do know that there were nine things to vote on in the last voting session, right? I'm not sure if Crusher made that statement with that in mind; but his claim, and the criticism of your proposed method, isn't without merit. Your proposed system seems like it provides fewer options, makes it more difficult to actually bring anything to a vote to implement changes, and has a harder time accomodating the nuance of what people actually want. I won't pretend the current system isn't without faults, but it's hard to see how your proposal fixes any of its imperfections.

What exactly is being discussed? I'm completely lost but I feel like voting for whatever of the however many rules we're voting on individually is the most reasonable (to quote O-Chul and all that).

[Squire Doodad]

What exactly is being discussed? I'm completely lost but I feel like voting for whatever of the however many rules we're voting on individually is the most reasonable (to quote O-Chul and all that).

Ah, right.

Grey Wolf is about to host a vote, and lio is fussing about how we're using our current system regardless of its current functionality, relative celerity and usefulness in processing votes.

I'll get back to you on what we're voting on, I forgot XD

I think we're voting on changing the CR requirement to (18, but you can suggest alternative numbers before the vote ends) and on whether or not "no mind-control immunity" should be changed to "no mind-control immunity on the scale of a god" or something.

[Grey_Wolf_c]

I feel like voting for whatever of the however many rules we're voting on individually is the most reasonable (to quote O-Chul and all that).

And if you feel that way, every voting system probably allows for it - the one I intend to use certainly allows it. But if you'd prefer to cast a vote of the style "both of them, or neither", then an individual vote won't do.

Worse, individual votes can end in paradoxes. Say that we have two options, A & B. A receives a majority, as does B, but A & B together fail, so now I have to implement both and not implement them together. Fun times.

And before you tell me that I shouldn't allow the vote "A & B together, but neither individually", let me stop you right there and state, clearly and unambiguously, that as long as I am the curator, I will never restrict the votes someone can express. And yes, I have had people who wanted combination of options but not their individual components. This is not theoretical.

ETA: see this post for the upcoming vote

Grey Wolf

[Squire Doodad]

And if you feel that way, every voting system probably allows for it - the one I intend to use certainly allows it. But if you'd prefer to cast a vote of the style "both of them, or neither", then an individual vote won't do.

Worse, individual votes can end in paradoxes. Say that we have two options, A & B. A receives a majority, as does B, but A & B together fail, so now I have to implement both and not implement them together. Fun times.

And before you tell me that I shouldn't allow the vote "A & B together, but neither individually", let me stop you right there and state, clearly and unambiguously, that as long as I am the curator, I will never restrict the votes someone can express. And yes, I have had people who wanted combination of options but not their individual components. This is not theoretical.

ETA: see this post for the upcoming vote

Grey Wolf

Well put! Let's get to voting.

Have there been any additional suggestions? If not, I'm going with No Changes for the time being.

Assuming No Change is A, CR18 is B, and Criteria 7 (Mind Control) is C, I'd be doing just A.

[Grey_Wolf_c]

Well put! Let's get to voting.

I'll probably do it official later - my child is being a bit of a handful, so waiting until nap is likely advisable at this point in time - but please be assured that I will retroactively assign votes for those who have already expressed a preference.

I will probably allow non-PM votes, too. I've clearly lost the battle against that particular approach.

Grey Wolf

[Squire Doodad]

I'll probably do it official later - my child is being a bit of a handful, so waiting until nap is likely advisable at this point in time - but please be assured that I will retroactively assign votes for those who have already expressed a preference.

I will probably allow non-PM votes, too. I've clearly lost the battle against that particular approach.

Grey Wolf

I think it's better to allow non-PM votes, but don't mention that you would allow them (just say "please PM me your votes" or something to that effect), thus not losing out on votes and minimizing the space lost to them.

Oh wow, you have a kid? I guess it's one of the side effects of being on a forum but I had no idea.

[Grey_Wolf_c]

I think it's better to allow non-PM votes, but don't mention that you would allow them (just say "please PM me your votes" or something to that effect), thus not losing out on votes and minimizing the space lost to them.

I think I'll just mention I allow both, and strongly encourage those who just want to vote without further details, discussion or explanations to PM them rather than consume a post with just a couple of bolded letters and some white-letter padding.

Oh wow, you have a kid? I guess it's one of the side effects of being on a forum but I had no idea.

Yeah, they're 4 these days. I announced it in this thread... well, 4 years ago, I'd say

Grey Wolf

[lio45]

You do know that there were nine things to vote on in the last voting session, right?

Exactly, and it was pretty clear from the start that most of them weren't going to pass. One advantage (and it's by design) of my proposed method is that it would have greatly thinned the herd as most of the nine from last time would not have had sufficient support to reach the voting stage.

Your proposed system seems like it provides fewer options, makes it more difficult to actually bring anything to a vote to implement changes

Yep, and that's (again) on purpose, because it's pointless to bring something to a vote if there isn't at least SOME appetite for it. If we're >~20 participants in here and you can't manage to find a mere five people willing to support your idea, there's no point letting your idea move through a formal voting process.

... and has a harder time accomodating the nuance of what people actually want.

I wouldn't say that; but yeah, we can say it shifts the way nuances would be accommodated. And a side effect is indeed that the nuances/variants which have nearly no support in the community would not be accommodated, but again that's by design.

I won't pretend the current system isn't without faults, but it's hard to see how your proposal fixes any of its imperfections.

Its main imperfection IMO is that it's needlessly complicated to a degree that defies imagination when compared to the application here (I'm barely exaggerating), and the same results can be achieved with much less effort and wasted time.

{scrubbed}


ETA: Also, I'd definitely want to add "Must not be an Embodiment of Evil" to the list of things to vote on (more detailed reasoning available upon request). It's the only one of the ~6 modifications I support that I believe has a good chance of getting the green light, so it's the one I feel would be the most relevant as a proposal. I could (and almost should, as that's how votes are supposed to function) suggest them all..... a side effect of that would be highlighting the flaws of the current voting method.

[Squire Doodad]

{scrub the post, scrub the quote}

{scrubbed}

[Grey_Wolf_c]

Voting Time

Topic: altering the FBS requirements.

Result: Voting is Closed. No Change (Z) has it, due to absolute majority in the first round

Rules:

• Semi-public vote: votes can be posted in the thread or be PM'ed to me, the curator. If you are only interesting in expressing the vote, rather than giving reasons, I encourage to PM rather than post, to prevent the thread from being spammed by multiple short "vote for X" posts. I will compile all votes in the original voting post for verification and record keeping. To prevent fraud, please be ready to reaffirm your vote if someone else PMs you to ask about it.
• Multiple votes per person: you can vote for more than one option. Do so in order of preference. You can vote on entries individually, or combined. So, for example, if you like change Y, and only want change X if Y is also implemented, you can vote for X&Y together, followed by Y (or the other way round, if you'd rather have Y, but can live with Y and X together, etc.). Do not be concerned about the number of columns in the table below - there is no limit to how many votes you can make. I'll add extra rows for those who need it.
• Vote counting: will not be first past the post; instead, alternative vote system will be used.
• Voting is binding: if you vote for extra work, be prepared to do the extra work. If extra work is voted for, but no-one does it, the option will be discarded and runner-up will be done instead.
• Voting options: If you feel the options are not to your liking, you can vote for a new option by describing it in your post. I'll add it to this post as time permits, and assign it a letter.
• Changing Votes: You can change your vote at any time before the end of the voting, by either posting or PM.

Notation

• + indicates both selections (i.e. A + B + C indicates a vote for A and B and C, but not for any subgroup thereof)
• Any() indicates any subgroup of the selection (i.e. Any(A, B, C) indicates votes for A + B, A + C, B + C, A, B, C).
• ! indicates a vote against a selection (i.e. !A is a vote to NOT implement A). It will be treated as identical to Any(B, C, D...) when the vote ends (i.e. a vote for any combination that does not include A)

---

Collected voting options:
A) Add "must be minimum Challenge Rating of 18, if a D&D creature"
B) Replace "7) Is vulnerable to mind-affecting effects (SoD)" with "7) lacks god-level mind-affecting effect immunity (SoD)."
C) Add "must not be an embodiment of Evil" requirement
D) Remove "7) Is vulnerable to mind-affecting effects (SoD)" requirement
Z) No Change

---

Collected votes:

Voter	PV 1	PV 2	PV 3	PV 4	PV 5	PV 6
Grey Wolf	Z	A
Kish	A+B	Any(A,B)	D
Doug Lampert	A+B	Any(A,B)	Any(A,D)
Throknor	Z
Squire Doodad	Z
Keltest	Z
Sir_Norbert	D	Z
woweedd	C	C+A	C+B
Darth Paul	Z
lio45	C	C+Any(A,B,D)
Aspheric	Z	A
CriticalFailure	A+B+C	Any(A,B,C)
The Aboleth	Z
Crusher	Z
Jaxzan Proditor	Z	C	Any(B,C)

---

Reason for my vote: We've been using CR18 as a gross rule of thumb for minimum power level for MitD for forever, so I wouldn't be completely opposed to adding it in, but I have heard often enough in this thread the opinion that CR is horrendously mismanaged, with some creatures having been given too much or too little to be trusted that I do not feel it should be more official than a gross rule of thumb. I won't mind it, but it definitely takes second place to not adding it.

B, on the other hand, I feel is way more punctilious and specific than any of the FBS rules should be. I barely understand what is aiming for, and I suspect too many people would come asking what exactly it means for me to want to face it. And as I said, I don't like the idea of voting again on something that already was voted down last time - it strikes me as wasting everyone's time- so even though I do like C, I stand with the results of the last vote.

Grey Wolf

[Keltest]

I vote for Z. I find hard coding CR18 to be a fairly arbitrary change with no actual benefits, given the lack of hard data from the comic about it. B on the other hand seems like its changing the rule to be too specific, and is suddenly removing non-deific sources of mind-affecting immunity as a disqualifier, and C likewise seems like its trying to codify an assumption as a hard rule.

[Squire Doodad]

I just want to say that having "Z" being the no change choice feels oddly appropriate to me Also makes adding in additional options easier to register.

To clarify my vote: All of the given choices at the moment, while probably all true (with some margin of error for the CR18), are also assumptions about the makeup of MitD's character based on other assumptions. For instance, we're not sure how strong the most powerful creature Xykon expected to be in the tower was (though it was somewhere beneath an Ancient Silver Dragon).

[Peelee]

For instance, we're not sure how strong the most powerful creature Xykon expected to be in the tower was (though it was somewhere beneath an Ancient Silver Dragon).

To be fair, Ancient Silver Dragons are, like, the best.

[Kish]

B, on the other hand, I feel is way more punctilious and specific than any of the FBS rules should be. I barely understand what is aiming for, and I suspect too many people would come asking what exactly it means for me to want to face it.

Okay, I have an alternative here:

D) Strike the "must not be immune to mind control" requirement entirely. Rationale: There's no such thing as "truly immune to all mind-control" in the D&D rules; the closest it gets is "immune to spells with the mind-affecting tag," which doesn't include all spells that produce effects that look a whole lot like mind control, as previously discussed.

I think "being able to list creatures like, e.g., the zodar as FBS" outweighs the downside of "might have to list amnesiac Zeus."

However, since I genuinely believe that a creature like the zodar (normally immune to mind control) is a possibility and amnesiac Zeus is not, please put me as voting for that option only in an "if option B is defeated" contingency.

[woweedd]

I'm voting for C, then C+A, then C+B.

[Grey_Wolf_c]

Okay, I have an alternative here:

D) Strike the "must not be immune to mind control" requirement entirely. Rationale: There's no such thing as "truly immune to all mind-control" in the D&D rules; the closest it gets is "immune to spells with the mind-affecting tag," which doesn't include all spells that produce effects that look a whole lot like mind control, as previously discussed.

I think "being able to list creatures like, e.g., the zodar as FBS" outweighs the downside of "might have to list amnesiac Zeus."

However, since I genuinely believe that a creature like the zodar (normally immune to mind control) is a possibility and amnesiac Zeus is not, please put me as voting for that option only in an "if option B is defeated" contingency.

Added - please verify that the votes look to your liking, I was after all guessing your preference - but please be aware that Zodar not being an FBS predates the mind control requirement and what keeps it out is the circus scene.

Grey Wolf

[Kish]

Yes, they do. And the zodar was just an example; "it doesn't qualify because no one's going to vomit at the sight of a suit of armor" works a lot better than "it doesn't qualify because Xykon couldn't possibly have a spell that could mind control a sapient construct."

[Sir_Norbert]

I change my vote to D, then Z. As I understand it, the mind control scene is not one of the "big" scenes, which refers to the circus, tower and escape.

[Darth Paul]

Holy Zarquon, we're voting on something???

Er... I just went through a 3 month relationship thing... there was some dysfunctionality which isn't strictly relevant here, suffice to say that you can't help a person unless they're willing to help themselves, and I finally realized it. I'm single again.

Anyway- about the vote- I'll read and figure out what's what then update this post or PM, whatever's appropriate.

EDIT: Okay, I'm all caught up, and I'm voting for No Change. That's Z as in Zykon.

To be fair, Ancient Silver Dragons are, like, the best.

Yes they are, my friend. Yes they are.

[Kish]

I don't quite follow woweedd's vote. It seems to me that there is no world in which their second or third choice will come into play. What am I missing?