POLL: Potential Changes to Monthly Picture Selection
Added 2019-05-18 15:39:52 +0000 UTCSweet Supportive Voters:
An idea was brought to my attention that I found interesting, so I wanted to present it to you to determine the outcome. I know I've made a lot of adjustments to the voting this year, but I'm trying to land on an option that makes as many people happy as possible. So here is the idea:
The vote will continue as normal, but instead of the winner automatically winning and the loser automatically losing, it enters a lottery based system with odds determined by the results of the voting. So instead of the winner winning and the loser losing, the winner has a higher percent chance to be selected and the loser has a lower chance. Let me provide a real world example based on the last vote:
- Amo and Shai in Bear Form (M/F) (Feral) : 186
- Drogan and Laspis in Dragon Form (M/F)] : 191
- Mordin and Kinar Anal (M/F): 193
- Worgen Sicha and Tauren Sicha (F/F): 202
- Serid and Shai (M/M): 203
I had to invert the points so this would all make sense mathematically: So we end up with this (you can see this was a pretty close race):
- Amo and Shai in Bear Form (M/F) (Feral) : 204 out of 975 total points or 20.92%
- Drogan and Laspis in Dragon Form (M/F)] : 199 out of 975 total points or 20.41%
- Mordin and Kinar Anal (M/F): 197 out of 975 total points or 20.21%
- Worgen Sicha and Tauren Sicha (F/F): 188 out of 975 total points or 19.28%
- Serid and Shai (M/M): 187 out of 975 total points or 19.18%
Now, instead of the 20.92 percent winning because it was .51 percent higher than the next guy, we essentially put them in a lottery based system where each option has that percent change of winning.
For the thought experiment, I randomly generated a number from 1 to 10,000 (to try to be as accurate as possible) and I got: 2575. The way the lotto system would work out is so:
- Amo and Shai in Bear Form (M/F) (Feral) : 1- 2092
- Drogan and Laspis in Dragon Form (M/F)] : 2093 - 4133
- Mordin and Kinar Anal (M/F): 4134 - 6155
- Worgen Sicha and Tauren Sicha (F/F): 6156 - 8084
- Serid and Shai (M/M): 8085 - 10000
(I know those numbers aren't perfect, but whatever). This means that the winner would be:
Drogan and Laspis in Dragon Form (M/F)] based on the random lotto winner instead of Amo and Shai in Bear Form (M/F) (Feral).
Now, that's a lot to digest, it adds a bit more randomness to it, but it also gives things a chance that normally wouldn't see the light of day.
So, open discussion, let me know what you think or just vote (please vote). Thanks!
BTW, here are some free throwback pics for taking the time to read this giant message.
Comments
I for one think it's a great idea! Especially when the votes are so close as they are in the example here. It's obvious there is desire for all of them. At a minimum not loosing the lowest I think would be best unless they are 5% or more lower than all the others, in which case it's an idea that most people are probably not interested in.
Allon the Cookie Dragon
2019-05-20 16:07:30 +0000 UTCyeah I can feel that too. Now, we did talk over splitting things out, but I kind of want to avoid doing more work for the same amount of money, ya know. Maybe having a different / another Patreon to handle M/M content and one to remain M/F content. But that would involve me doing another picture a month, which I'm not entirely sure I could swing given that school is almost up and I'd rather spend my time hanging with the kiddos than drawing porn =). To be honest about the pictures that I'm interested in? I probably did those ages ago. This comic has been going since 2012 - I'm actually eager to get moving on my new comic, to be honest. But I'm going to wrap up Druids first. Thanks for the feed back. As per usual, this vote is split pretty much down the middle too.
Amocin
2019-05-19 00:09:32 +0000 UTCSo i didn’t vote because I’m not sure how I feel. I can see what you’re getting at, because I was gunning for the double Sicha pic personally. I do again that the winners suddenly losing do to the random would feel gipped, but the only ways I can think to have it work is for there to be two monthly pics, but that isn’t fair to you Amo. l will say I’ve wondered from time to time what picture you were most interested in, so that might be a thing? Another thought is letting people pay for one of the other options somehow. No clue on the logistics there, just a thought.
Chance the Wolfhog
2019-05-18 22:21:10 +0000 UTCI did plan on keeping all the options this fine around. But your right I should age some of the out.
Amocin
2019-05-18 19:14:28 +0000 UTCIf I could bring up a suggestion, if all the options are close like they were this month, than maybe don’t drop anything and leave all of them as options for next month and just add only one new option for the next vote. That way the last place option could have a chance to get chosen for another month. And if you have to get rid of options, maybe set a limit on how long the option is in the vote, for example maybe giving it 5 months or so to get chosen, and if it surpasses that then get rid of it, or if its a clean slate and there are one or two options that are clearly not favorites. I don’t know, just an idea, whatever you want to do I will be willing to do. I liked the ranking system but I can also like the suggestion you brought up because it can give more variety to what we get. Again, whatever happens happens.
Doty
2019-05-18 19:10:55 +0000 UTCI see where your coming from. As a thought experiment I decided to count the points a bit different. Instead of 1 = 5 points, I did 1 = 4 points. That means anything that was a 5 was effectively not counted (so 1=4, 2=3, 3=2, 4=1, 5=0). This changed the percents a bit (not massively): Amo and Shai in Bear Form (M/F) (Feral) : 139 out of 650 total points or 21.38% Drogan and Laspis in Dragon Form (M/F)] : 134 out of 650 total points or 20.62% Mordin and Kinar Anal (M/F): 132 out of 650 total points or 20.31% Worgen Sicha and Tauren Sicha (F/F): 123 out of 650 total points or 18.92% Serid and Shai (M/M): 122 out of 650 total points or 18.77% But a bit. So in the method in the original post. Each voter would account for 1.5% of the total, each "point" being about 0.1%. With 5's equaling zero, each "point" 0.15% - still 1.5% of the vote (that part ain't changing. Now, lets do another check. Let's say a 1 is worth 5 points, a 2 is worth 3, 3 is worth 2, 4 is worth 1, and 5s are worth 0 (just throwing this out there). Here is how it changes: Amo and Shai in Bear Form (M/F) (Feral) : 150 out of 715 total points or 20.98% Drogan and Laspis in Dragon Form (M/F)] : 146 out of 715 total points or 20.42% Mordin and Kinar Anal (M/F): 143 out of 715 total points or 20.00% Worgen Sicha and Tauren Sicha (F/F): 133 out of 715 total points or 18.60% Serid and Shai (M/M): 143 out of 715 total points or 20.00% Now the winner doesn't change but the loser does. Also everything is even more crazily closer. Anyways, I'm not sure where I was going with this. But weighted randomization wouldn't be a bad idea either. The ultimately point to this is, I to try to find a balance where people still feel like they have a chance of their vote winning even though it only lost be a few percent. Of course....look how that worked out in the 2016 election -_-. So maybe there isn't an option that'll make everyone happy all the time. Now, I don't want to kick the bees nest here, but if I only count the 1s, which would be the result of the pick one vote option, here would be the results: Amo and Shai in Bear Form (M/F) (Feral) : 11 out of 65 total points or 16.92% Drogan and Laspis in Dragon Form (M/F)] : 12 out of 65 total points or 18.46% Mordin and Kinar Anal (M/F): 11 out of 65 total points or 16.92% Worgen Sicha and Tauren Sicha (F/F): 10 out of 65 total points or 15.38% Serid and Shai (M/M): 21 out of 65 total points or 32.31% In the old ways (before ranked voting) the M/M option would have won by a landslide. ANYWAYS. What was I talking about again?
Amocin
2019-05-18 18:38:19 +0000 UTCI don't like the idea, at least not in the state you presented it. It makes voting almost obsolete in most cases, as the precentages of the images in the poll are basically the same, might as well just make a random poll of one of the ideas. Especially with that voting system, since you can't really weigh your votes or chose not to vote for an idea at all. at the moment I get presented 4 to 5 ideas. Maybe I like 3 of them a lot and the other 2 not at all, still the other two ideas get only 1 and 2 points more than the idea I like the 3rd most, which only results in 0.01% difference in voting, if you take a sample size of 10000 tickets as you presented in your example. The point of majority votes is that you get the idea that the people like the most, even if it is just by a small amount. There are always people that will be disappointed with the outcome, but that's how such votes work. So you either need to reqork the voting process, giving people the chance to weigh their votes, like give one idea they really dislike more than 5 points, or two ideas the same amount of points. (which I would like in genere, regardless of the random poll idea) Or make stronger weighs in the random poll idea. Like for example the idea that got the most votes, gets 5 times the tickets than the least favored one. So you have a sample size of 15000 tickets, and the highest voted idea gets 5000 tickets, the second place 4000 and so on and the least favored gets 1000. There is still a chance for the least fave idea to win, but it isn't as high anymore as having 21% vs 19% for the highest and lowest idea. Due to the nature of our voting system there will always be pretty small differences between the results, since you have to give points to ideas that you don't like at all.
Lenina
2019-05-18 18:07:24 +0000 UTCHa. Meh is a good description. Maybe I need to wipe the board of some of the middle of the road pictures. Thanks for the input!
Amocin
2019-05-18 17:41:46 +0000 UTCPart of the issue, as you saw, is some content, like M/M is highly polarizing, while some is overwhelmingly "Meh". I think the Ranking system will continue to throw mostly-even results.
Sartharina/Scow2
2019-05-18 17:38:33 +0000 UTCOdd, I thought I responded to this. Whoops. I think the only reason why I'm presenting this deal is the votes were soooo close this time around that it seems a bit unfair to boot out something that only lost be like...2%. If it were different, like a 10-20% swing I doubt I'd be having this conversation. So maybe only randoming it when it's real close...that could be an option too
Amocin
2019-05-18 17:27:03 +0000 UTCI think i like the ranking system much more. If several are really close at the top, then maybe you could just "/random" one of them, but if somethign is "voted out" then it should not get another chance
Tankana
2019-05-18 17:13:59 +0000 UTCI see what you're saying and this was one of those things that stuck out in my mind. For the above example, to me, it seems fine as they were all like...literally 2% of each other (which is weird, but hey, thems the odds). So I don't think there would be too many folks butt hurt if any one of those options one. However, if we have like...90% and a bunch of 2%s and a 2% won, then I can see a bunch of grumpy cats. So maybe what we could consider is, if one option has say...the VAST majority of votes, then it will win and we won't randomize it. Just a thought. Thanks for the input, I'd rather this be an open discussion (maybe even on discord, idk)
Amocin
2019-05-18 15:51:31 +0000 UTCThis seems the best way to irritate the most people. I thank you for considering us who lose, I have been on the losing side more often than not, but this system... won't work. To explain why, briefly, this way you now have a pretty solid chance that the thing the majority fo people voted for will never be the thing made. Now the majority vote is disregarded, and more people than not are irritated by it. Your ranked system would work better, though if you really wanted to make things a bit more complicated and try to get the thing the majority of people want, ranked voting + runoff. Do the ranked voting, and if none get a clear majority then do a runoff between the highest three.
AnselmBlackheart
2019-05-18 15:48:28 +0000 UTC
