Festus1965 wrote: ↑Sat Nov 14, 2020 11:53
The SAMMY AWARDS as depending on votes, and voters are registered in forum, and most gamer are not ... so a admin just keep advertising (forcing) for the own gamer to register and vote and we would have no change against like Multicraft mass.
In most cases people use a bad or the wrong voting system and then get frustrated when the results turn out bad and conclude that voting does not work. (Literally, the most commonly used voting system (plurality) is also the worst one, and the second most commonly used is the second worst.) Designing voting systems is like engineering. You might build something that resembles a big bird, wonder why it never takes of and conclude that flying is impossible. But it's not, you just have to know how to do it.
We can learn from previous mistakes and change the rules accordingly.
So, I looked into the SAMMY AWARDS.
First, it isn't really clear what voting method is used and how votes will be counted. That's not only bad, it's very bad, because it allows whoever counts the votes to interpret them to their liking.
Second it is stated that nobody can vote for their own server/mod/game, but no way to check this. So honest voters miss out, while dishonest can push their candidates.
Third, nobody ever counted the votes, so I did. Assuming that first vote is 10 points, second vote is 7 points and third vote is 3 points.
Brain Craft: 60
MTG: 43
XaEnvironment: 30
Mineclone 2: 10
Voxelgarden: 10
Farlands: 7
IKEA: 7
Tolkien Test: 7
While the newly registered voters pushed Brain Craft to 60 points, that's only 36% of total points.
If we instead count every vote as full vote then we get this:
Brain Craft: 6
MTG: 5
XaEnvironment: 3
Mineclone 2: 1
Voxelgarden: 1
Farlands: 1
IKEA: 1
Tolkien Test: 1
Brain Craft with 6 votes makes for 32% of votes.
Either way, the voting system used is an odd form of plurality voting.
Plurality voting is also known as First Past The Post and it's the usual "mark one" method most people are familiar with. Voting experts agree that this is the
worst voting method there is. And we can see it by the above example. Even if 68% of votes did not go to Brain Craft, it still had "the most votes" and would have won under plurality. Shouldn't we make sure that the winner is at least supported by 50% of votes? Or even more that the winner is preferred by a majority over every other candidate? What about a game that everyone likes, but doesn't make it to the first preference often enough to win?
There are voting methods that are way better than plurality in every way and which fulfil the above requirements.
approval voting: Vote for
as many candidates as you like. Write-ins are also accepted.
score: Score each candidate on a scale of 0 to 5. In principle is the same as approval voting but more nuanced.
STAR voting: Like score, but for the top two candidates count the votes again and count how often one was scored higher than the other.
But that's for single winner elections. In the case of electing the server of the week it's different, as each week should feature another server. Simply electing the best would result in the same server winning every time. Excluding the winner from the list would only result in going down the list of servers by popularity. After the first few it would become boring, and later it would only ever elect new or completely unknown servers.
This is why I use the
random ballot method. It gives every server a chance by the proportion of popularity. This means, even if some group is flooding the topic and making up 36% of the votes, they will only have a 36% chance of winning. Even more, if some coordinated group makes up 70% of votes, then there is still a 30% chance another server is elected. To allow for more variety I added the rule (see first post), that votes for winning server get exhausted. So if the 70% group succeeds in getting their server elected, they would need to renew all their votes for the next round, while the other votes remain valid.
To answer your daughters questions
Festus1965 wrote: ↑Fri Nov 13, 2020 11:22
There is a Post missing from my daughter, she had good question:
* as 300 server every month one mean 25 years for revisit the same
or every weekend another just 5,5 years for repeating
See above. Servers aren't excluded after winning, but past votes for them. So a server could win several times in a row, but that's unlikely. I hope that we get a healthy mix.
Festus1965 wrote: ↑Fri Nov 13, 2020 11:22
* and for that, is it just visiting, looking around ? - or build something as memory been there
and so how long would that builds, areas granted that exit when come again after 25 or 5.5 years haha
That's up to you. It could be an opportunity to just visit and explore something new, or to meet and hang out - depending on how many players participate.