Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Draco18s]

Also, I added the visible movement between tiles suggestion to the list.

...Limitations of the graphics engine aside...

[Faces of Mu]

Dear dear. How did Adventure-mode skills get so low? Do we not need more things to do between releases? What about voting for things that increase Df's longevity, things that multiply the possibilities to the greatest extent? 

In hindsight, I'm still glad we've got this system here.   

[Draco18s]

It's only at #13.

5 or 6 of the things above it reduce annoying micromanagement in fortress mode, three add things directly to fortress mode (more content/challenge), and two are detail things (one of which is so often requested).

Other great ideas far down the list:

#21: rope ladders (building down)*
#41: doctors and hospitals*
#42: underground diversity*
#45: less noise (building chairs shouldn't cause "disruptive noise" in a 32x32x32 cube)
#74: true arena
#90: magic

*going in in the next version

[Footkerchief]

Dear dear. How did Adventure-mode skills get so low? Do we not need more things to do between releases? What about voting for things that increase Df's longevity, things that multiply the possibilities to the greatest extent? 

It's not really worth voting for since it's already on dev_next and has been selected, along with sieges and the top 10 ESV entries, as a priority for releases after the upcoming one:

This also turned out, as far as the timing of 'when are we going to see this kind of thing', there's the return of the guilds, or I don't remember exactly what it's called, up in the eternal suggestions voting up on the suggestions list and it's doing pretty well. I said I'd take a look at the top ones there for next time - now by next time I mean the next series of releases because I really hope I never have a release this long again, it's going to be over a year - so in the short term we're going to be looking at this stuff, it's not like it's just something I'm talking about; we're going to look at this along with the adventurer's skills stuff and the improved sieges; job priorities; improved hauling; I don't remember if improved farming is up there; increased tile support for graphics ... whatever things were up there that people wanted we're going to be looking at.

[zagibu]

How bad would it be if the new script also meant a wiping of all votes and/or suggestions? I can probably avoid it, but it would help the new design if I could start completely anew.

[Footkerchief]

How bad would it be if the new script also meant a wiping of all votes and/or suggestions? I can probably avoid it, but it would help the new design if I could start completely anew.

Wiping votes a while AFTER the next release isn't such a bad idea, but it took the current top 10 a while to work their way up there.  Are you planning to change the way votes are allocated?  I can't imagine the suggestions themselves would be that hard to carry over?

[zagibu]

There will be three additional methods allowing people to vote for suggestions, all with separate rankings. I'm doing the global thumb up/neutral/down thing, a mode where you can assign fractions of 1 vote to any suggestion and the direct compare mode, that pairs random suggestions and lets users decide which of them they like better.

It wouldn't be terribly hard to reference and modify the old tables. I just thought it would be a good occasion to relaunch the whole affair and purge the dead weight that has been accumulated during the past year. A lot of the current suggestions don't even have a linked discussion thread, so there is probably a lot of potential for purging. A vote purging could also be necessary to make the results of the "pick3" voting directly comparable with the other voting methods.

But yeah, I also realize that some people might not like having to re-vote for or even re-create their suggestions. Maybe I should leave the suggestions and votes alone and add the new script along the old one, instead of replacing it.

[Footkerchief]

I just thought it would be a good occasion to relaunch the whole affair and purge the dead weight that has been accumulated during the past year. A lot of the current suggestions don't even have a linked discussion thread, so there is probably a lot of potential for purging.

That's true.  Some people are being proactive about marking obsolete suggestions for removal, but a purging might still be in order.

[zagibu]

I'm also considering to change the "pick3" voting method (the one in the currently active script) to a simple vote for your favourite suggestion, because pick3 is a subset of the new "distribute" voting method, where people can support any number of suggestions as strongly as they like.

I'm not sure whether this would lead to a pronunciation or an attenuation of the current voting assignments...

[Footkerchief]

I'm also considering to change the "pick3" voting method (the one in the currently active script) to a simple vote for your favourite suggestion, because pick3 is a subset of the new "distribute" voting method, where people can support any number of suggestions as strongly as they like.

Isn't the "favorite suggestion" also kinda subsumed by the direct compare mode, though?  The three methods you mentioned before (direct compare, up/down/neutral, and fractions/distribute) sound like an ideal solution to me. 

[Toady One]

I wonder about the direct compare one.  Does it give higher importance to people that vote for more pairs, just by virtue of the fact that they have more votes?  How much information does it store (since there could theoretically be lots and lots and lots of these)?  Does it have to tax the server to calculate the winners if there are lots of votes any time somebody queries it or does it just store the number of yes/total votes for each suggestion?  If the latter, how can the importance assigned to people be controlled fairly (there's also the matter of controlling abuse)?  Also, how are suggestions compared?  It seems like it would have to be some combination of the win-loss ratio and the total number of wins, maybe, to stop new candidates from being at the top but also to give newish candidates a chance to arrive at or near the top.

Since the votes have gained some official importance for the next series of releases, I don't think the current votes should be purged unless they are saved in some way.  Of course, I could just do that, and use those as the list I'm working from while the new systems attract votes.

[Footkerchief]

^^^ Unless I misunderstood zagibu, the direct compare method is equiv. to preferential voting, which has various methods of choosing a winner, only some of which seem suitable for computing an overall ranking:

The Borda count is a single-winner election method in which voters rank candidates in order of preference. The Borda count determines the winner of an election by giving each candidate a certain number of points corresponding to the position in which he or she is ranked by each voter. Once all votes have been counted the candidate with the most points is the winner. Because it sometimes elects broadly acceptable candidates, rather than those preferred by the majority, the Borda count is often described as a consensus-based electoral system, rather than a majoritarian one.

Instant-runoff voting (IRV) is the American English term for a voting system used for single-winner elections, in which voters rank candidates in an order of preference. If no candidate is the first preference of a majority of voters, the candidate with the fewest number of first preference rankings is eliminated and that candidate's ballots are redistributed at full value to the remaining candidates according to the next ranking on each ballot. This process is repeated until one candidate obtains a majority of votes among candidates not eliminated.

The Condorcet candidate or Condorcet winner of an election is the candidate who, when compared with every other candidate, is preferred by more voters. Informally, the Condorcet winner is the person who would win a two-candidate election against each of the other candidates. A Condorcet winner will not always exist in a given set of votes, which is known as Condorcet's voting paradox.

(various systems that fit this description)

Apparently it's a nontrivial problem, in regard to both voting theory and computational complexity.

Incidentally, the up/down/neutral method could also have some problems with voters playing the system, assuming down votes are just subtracted from up votes -- there's no incentive to vote neutral on anything.

[Toady One]

At least here, ties won't cause nearly as much grief, since we can have more and more winners over time in terms of me actually doing something, so we don't have to worry too much about those situations.  I guess the main thing to focus on is making the system as different from the fractional method as possible while keeping the speed/mem costs in line.

Hmmm....  maybe you can only have as many down votes as you have up votes?  Maybe that's too much like direct compare.  If there are no down votes, but just up and neutral (so it's a pick-as-many-as-you-like method), maybe that's different enough from fractional voting.  The lack of normalization is what makes it complementary to the fractional method rather than inferior, I guess.

[Footkerchief]

If there are no down votes, but just up and neutral (so it's a pick-as-many-as-you-like method), maybe that's different enough from fractional voting.  The lack of normalization is what makes it complementary to the fractional method rather than inferior, i guess.

Yeah, that sounds ideal.

zagibu, the direct compare method would be cool, but if it turns out to be a boondoggle we can probably do without it.  Thanks for continuing to work on this, by the way.

[Toady One]

I wonder though again...  an up/neutral/down is really just a 1 to 3 rating system (assuming we are just going to tally things), and a pick-as-many-as-you-like is a 1 to 2 rating system.  By taking away the ability to make nuanced decisions, is anything really gained?  People can still attempt the same game regardless of the scale by top rating a few and downgrading the rest, whether it's 1-2, 1-3 or 1-10.  Assuming I understand it, the fractional vote system is just that game in a formalized fashion, where it pays off to express strong preferences, whereas in the rating system, expressing strong preferences is much more likely to land you something you don't like (because you didn't favor any of the other options).

Say we take the 1-2 rating system (up/neutral).

          T1  T2  T3  T4  T5  T6  T7  T8  T9  T10
Voter 1:  2   1   1   1   1   1   1   1   1   1
Voter 2:  2   2   1   2   1   2   2   1   1   2


Voter 1 is doing the game, as it were, or maybe just has a strong preference, while Voter 2 is going by general preference.

However, is anything lost if we allow:

          T1  T2  T3  T4  T5  T6  T7  T8  T9  T10
Voter 1:  2   1   1   1   1   1   1   1   1   1
Voter 2:  2   1.5 1   2   1.5 2   1.5 1   1.5 2


Voter 1 is still doing the same thing they'd be allowed to do in any rating system, with the same effect, and Voter 2 is expressing their preferences better.  It's in a 1-2 scale, but this is just a rescaled 1-3 rating system (that is, up/neutral/down = 2/1.5/1 = 3/2/1).

I could very well have missed something, but if this is correct, then a 1 to 10 or even 1 to 100 (perhaps too clunky) rating system is always more informative than an up/down system, even taking cynical voting into account.

Speaking of trying to game the system, I remember one site had a 1-10 system and would divide down the value of somebody's votes during summation if they had a high standard deviation (since they are just voting high-low), but that might be a little rough, he he he.