Goonhub
09-28-2019, 01:54 PM
(This post was last modified: 09-28-2019, 02:08 PM by phyvo. Edited 6 times in total.)
Hey, so now that Manta is released I think the weakness of Goonstation's map voting system is quite apparent. I've voiced this opinion in private feedback channels but in less detail and I'd like to share it in public to promote discussion. It's well known that a winner-takes-all system promotes two party systems and discourages third parties in democratically organized countries. Now that Goonstation has 7 viable maps for rotation (since cog 2 and atlas are mutually exclusive) that means that usually there's a lot of votes that get "wasted". There's often cases where one map does not get a majority vote but wins anyways because those who vote for other maps are not coordinated enough to defeat it. Look at this example from today:
Cogmap 2 - 0.5 total votes
Oshan Laboratory - 2.5 total votes
Cogmap 1 - 5 total votes
Horizon - 1.5 total votes
Manta - 4 total votes
Despite having only 37% of the total counted votes Cogmap 1 was the map selected for the next round. 63% of the votes would have preferred a different map. Even if you remove the passive voters you're still talking only 41% of voters actively wanting to play on Cog1 (it's possible there were more passive voters but then they become harder to discount anyways). If the dissenters never coordinate then Cog1 would win this every time and no other map would be played. Now, in practice people vary their votes a bit and sometimes the dissenters DO coordinate. But all that does is bring up the "third party" voting issue, where the 3-4-5th map runner up votes are completely meaningless, except maybe the 3rd runner up as the spoiler "candidate" between 1st and 2nd place. And why are we letting this kind of game theory prevail when players should be able to just vote for the map that they want to play on and feel that their vote helps make that happen?
I am of the belief that Goonstation has great maps that the mapmakers obviously poured a lot of effort into. And I shouldn't be left thinking about how Oshan and Manta "compete" with each other as underwater maps and so make other maps a more likely option or how one migh supplant and engulf the other once people realize that one or the other has the best odds of beating Cog 1.
So what's the solution? In the real world democracies there are other ways of handling this kind of thing, but we're a video game and we're picking a map to play on for the next hour, not a government that runs things for years. So for a good solution I'm looking at a server that has already dealt with this problem quite well: /vg/.
On /vg/ the maps are chosen with a somewhat random *weighted* system. Now there *are* thresholds at the low and high ends. One person voting for destiny can't randomly win a 5% chance die roll and force everyone to play destiny, and if 90% of people want Cog 1 it automatically wins without any random die rolls. But in situations like my example above where Cog 1, Manta, and Oshan all had decent support, the weighted die would be cast. Cog 2 and Horizon would probably not qualify, leaving us with a 43% chance for Cog 1, a 34% chance for Manta, and a 22% chance for Oshan to be the next map chosen. I think this is an extremely fair system. Instead of one or two maps dominating everything (Pre-Manta that was Cog 1 and Oshan) all maps that show enough interest have at least a shot at being the next map. And since rounds generally last 45-65 minutes the die roll isn't a huge deal and there's time for rounds to even out the random results. As far as map voting goes I think this is the best system we can hope for given the number of maps that Goon has. Simple winner-takes-all can work when there's very few options but on Goon we have a glut of options and I've seen a lot of people comment on how this or that rarely played map is their favorite.
Anyways, that's my perspective on the problem. There's not a whole lot of point to having lots of maps if the map voting system discriminates against anything below the top 2. And maybe I'm influenced by the US politics of 2016 but I'd rather not constantly vote for whatever I perceive the other most popular map to be simply because I don't want to play on Cog 1.
Cogmap 2 - 0.5 total votes
Oshan Laboratory - 2.5 total votes
Cogmap 1 - 5 total votes
Horizon - 1.5 total votes
Manta - 4 total votes
Despite having only 37% of the total counted votes Cogmap 1 was the map selected for the next round. 63% of the votes would have preferred a different map. Even if you remove the passive voters you're still talking only 41% of voters actively wanting to play on Cog1 (it's possible there were more passive voters but then they become harder to discount anyways). If the dissenters never coordinate then Cog1 would win this every time and no other map would be played. Now, in practice people vary their votes a bit and sometimes the dissenters DO coordinate. But all that does is bring up the "third party" voting issue, where the 3-4-5th map runner up votes are completely meaningless, except maybe the 3rd runner up as the spoiler "candidate" between 1st and 2nd place. And why are we letting this kind of game theory prevail when players should be able to just vote for the map that they want to play on and feel that their vote helps make that happen?
I am of the belief that Goonstation has great maps that the mapmakers obviously poured a lot of effort into. And I shouldn't be left thinking about how Oshan and Manta "compete" with each other as underwater maps and so make other maps a more likely option or how one migh supplant and engulf the other once people realize that one or the other has the best odds of beating Cog 1.
So what's the solution? In the real world democracies there are other ways of handling this kind of thing, but we're a video game and we're picking a map to play on for the next hour, not a government that runs things for years. So for a good solution I'm looking at a server that has already dealt with this problem quite well: /vg/.
On /vg/ the maps are chosen with a somewhat random *weighted* system. Now there *are* thresholds at the low and high ends. One person voting for destiny can't randomly win a 5% chance die roll and force everyone to play destiny, and if 90% of people want Cog 1 it automatically wins without any random die rolls. But in situations like my example above where Cog 1, Manta, and Oshan all had decent support, the weighted die would be cast. Cog 2 and Horizon would probably not qualify, leaving us with a 43% chance for Cog 1, a 34% chance for Manta, and a 22% chance for Oshan to be the next map chosen. I think this is an extremely fair system. Instead of one or two maps dominating everything (Pre-Manta that was Cog 1 and Oshan) all maps that show enough interest have at least a shot at being the next map. And since rounds generally last 45-65 minutes the die roll isn't a huge deal and there's time for rounds to even out the random results. As far as map voting goes I think this is the best system we can hope for given the number of maps that Goon has. Simple winner-takes-all can work when there's very few options but on Goon we have a glut of options and I've seen a lot of people comment on how this or that rarely played map is their favorite.
Anyways, that's my perspective on the problem. There's not a whole lot of point to having lots of maps if the map voting system discriminates against anything below the top 2. And maybe I'm influenced by the US politics of 2016 but I'd rather not constantly vote for whatever I perceive the other most popular map to be simply because I don't want to play on Cog 1.