Board 8 > Fair Voting

(+) Yeah! (+)

I created an event where 12 people create a game. Over the course of a few days, all 12 people play all 12 games.

Some people create multiple games, and some people share creating one game.

At the end of the event, all 12 people rank each game from their most favorite to least favorite. However, to remove bias I came up with a rule that states you must rank the game that you created last (i.e., least favorite).

However, I dont want the results skewed. Realizing that some people share one game, if they both rank it at the bottom then thats an unfair disadvantage. Additionally, for the people who create multiple games, how are they supposed to rank given the criterion set?

Any ideas?

I was thinking for the people who share creating a game, I can have one of the creators rank it last, and the other rank as s/he normally would without any restrictions.

Just let people vote what they think is the best regardless if they made it or not. You kinda threw complete objectivity out the window when you set the format up allowing mutiple games per person and teams to vote as individuals.

---

Why not have them rate the games instead of ranking them? Then don't let them rate their own game?

hockeydude15 posted...

Just let people vote what they think is the best regardless if they made it or not. You kinda threw complete objectivity out the window when you set the format up allowing mutiple games per person and teams to vote as individuals.

Complete objectivity was thrown out when people biasedly voted for theirs as number one in past tournaments. Also, no teams can vote as an individual. I do not know where you got that from.

Buffy_Summers posted...

Why not have them rate the games instead of ranking them? Then don't let them rate their own game?

Interesting. But the issue would still remain for the people who shared a game. Because that would be one fewer person who accrued their game points.

Moonroof posted...

Interesting. But the issue would still remain for the people who shared a game. Because that would be one fewer person who accrued their game points.

You could just take the average instead of the point totals. But I guess the problem would be people voting other games really low to try to inflate their own score.

I wont think anyone would do that. They would rate their very highly though.

Moonroof posted...

Complete objectivity was thrown out when people biasedly voted for theirs as number one in past tournaments. Also, no teams can vote as an individual. I do not know where you got that from.

He's saying one vote per team would have been better. Otherwise larger teams have an inherent advantage of having more voting power. Even with the "rate yourself last" restricting you are still giving larger teams more power because tray can strategically give low scores to their perceived rivals and have a unduly large influence on the scoring.

---

Exactly. That's why if you cared so much about how the standings would go you should have had a one game per group/person and one ranking from them as well. Once you didn't care about that there really isn't any restriction you can put in that doesn't completely favor one group or another.

---

I see. I dont want to have one team get only one vote because theyll have different preferences. Ill look to break up the teams. Thx.

I can think of something that wouldn't be too hard if you could work it into a spreadsheet.

1) Everyone ranks every game that they did not participate on. (example: someone worked on 2 games and there were 15 total games, so they rank the other 13 games from 1 to 13).

2) Each vote gets counted as a decimal. (So to take the previous example the game ranked the best would be worth 13/13 or 1.000 pts, and the game ranked fourth would be worth 10/13 or .769 points)

3) Average all the votes for each game to determine it's weighted value.

So if a game got nothing but 1st place finishes it would the maximum possible score of 1.000, but more likely the scores will all end up being decimals.

I'm not sure if that makes sense, but I could probably throw something together in Excel.

---

@Moonroof

Here's what I have in mind. (Keep in mind I'm an Excel Noob so the formulas being used are pretty janky looking, but imo they would work)

Here is the sheet that is blank and ready for input:



And then you fill in the first two columns and the "Total Votes" and "Total Points" and "Average Points" will automatically calculate.

Here are two examples of me filling in some scores. On example with nothing but 1st place finishes, and another example with middle of the road finishes:





You would just start with a blank sheet, fill in the first two columns and see what the Average Points worked out to. And Idk how many votes you are anticipating but currently it is set up to take up to 100 votes. This could be changed if necessary. And I'm sure there is a way to set it up with better formulas so it could take an undetermined amount of votes, but I don't know how to do that.

Edit: This still allows the POTENTIAL for large teams to vote in strategic ways, but the most they would be able to do is drive down the average points of a rival, or drive the average points across the board closer to 0.500 and thus reduce the amount of good votes they need to excel by comparison. But I think it's a lot less prone to abuse then allow everyone to give themselves a first place finish and it weights the value of each vote based upon the number of things they had to vote on.

---

I would also like to hear @hockeydude15 's thoughts on this.

---

Thats a great idea @Forceful_Dragon !! Specifically, having everyone rank only games that they did not create, rather than force them to rank their own games last.

The voting will be in-person though and the results should be immediate. I therefore wont have the ability to compile everyones votes into Excel. I will be able to calculate the averages (rather than the sums) pretty quickly though, thereby eliminating the skew. Thanks for the idea!

Also, for what its worth, all voting is done independently and silently. There is no conspiring during voting. That will eliminate any tactics and strategies.

One more item of concern.

For the people who create multiple games, are their results going to be skewed? If there are 15 games, and one person created five of them, then that means theyd rank games 1-10 (because their five wouldnt be included). But in that scenario, their least favorite game gets a value on 10 whereas everyone elses gets a value of 14. Am I missing something?

That is why my system accounts for that by making each vote a decimal that is relative to the number of games in the vote. I don't see any other way that won't inherently favor either a larger/smaller dev team or creators with more or less games.

---