Problem regarding polls design

Internet Go Nova League
1

Hello there,

First, allow me to thank the Nova League team for all the effort that makes this awesome league possible. I’m sure no one will disagree with me saying that we all enjoy it a lot !

Second, I’d like to raise an issue about the polls : the way they’re made, only a handful of lucky participants have an actual say about the time settings that are chosen. Let me explain :

  • one chooses an amount of time for the time setting type one has previously chosen only.
  • hence, when a time setting type is picked by the community, only those who chose it have a say about the amount of time, because those who preferred another time setting type picked an amount of time for the latter only.

The result is that the poll is biased in favour of the time setting type preferences, to the detriment of the amount of time preferences.

One way to fix this is to divide the poll in two parts :
i. pick the time setting type ;
ii. pick the amount of time for every possible time setting type.

I think this is fairly easy to fix, as it only requires a few additionnal questions (which already exist, btw : it’s just that not everyone gets to answer them).

Also, there’s an overall problem with the vote system. The way it currently works is pretty much like a majority one-round election system (such as in the UK). In the end, the chosen settings may get picked by a minority (e.g., 30 - 20 - 20 - 18 - 12). The largest minority, true, but a minority nonetheless. One way to fix this would be to allow voters to pick any number of possible settings. The time setting that would gather the largest number of votes would be chosen.

This way, picking several settings indicates that one is indifferent to them, and does not want those one did not pick. Personnally, I think I would pick Fisher and Byo-Yomi, for example, thus indicating my indifference towards this choice, but rejecting other possible settings. Some would have a strong preference for Fisher, and may pick this setting only. Others would be radically indifferent and pick every possible setting.

This way, the chosen setting would not be the one the largest minority picked, but the one everyone actually agrees on. The problem is that I do not know whether such a system exists. I’ve created a google form to check, and I think that a multiple choices question allows that.

Anyway, thanks for your attention ! :smile:

3 Likes
2

@hqrpie Thanks for your feedback! :smile:

This is in fact the current design of the poll; a respondent first chooses the time setting type, and the form directs the user to pick the amount of time for that chosen type.

Great idea! I will implement it in the next poll (for June 2015’s TNL.) :heart_eyes_cat:

1 Like
3

I believe preferred time setting is currently only chosen for one time setting type, not every possible time setting type?

4

That’s right, @Samraku.

5

Thanks for your answers, @xhu98 and @Samraku.

@xhu98, I’m glad that you find my idea interesting and wish to implement it next month. I’m sure it will improve the quality of the poll :smile: BTW, it’s a great thing that OGS (and TNL) allow a bottom-up flow of suggestions, many thanks for that.

As for the amount of time, I’m afraid I did not express myself clearly : the current design of the poll makes one choose an amount of time for one time setting type only. This prevents those who did not choose the one eventually picked from having a say in the amount of time. The solution would be to ask everyone which amount of time they wish for every time setting type, regardless of their choice.

Taking into account my different proposals, next poll would be designed as such :

  • players can pick several time setting types ;
  • players can pick several amounts of time for every time setting types.

Thanks again for your attention !

3 Likes
6

As an example of hqrpie’s point, I voted for a Fischer time setting in March. Since a byo-yomi setting won the poll, I did not have any influence on which byo-yomi setting was chosen.

2 Likes
7

I see, @hqrpie and @saxmaam; too bad I already started the May poll! >.<

1 Like
8

@saxmaam : exactly my point.

@xhu98 : TNL shall last for a long time ! One month early or late does not matter :wink:

3 Likes
9

Right!

2 Likes
10

Not sure exactly what you have in mind for final solution, but there’s a danger of having an outcome that the majority of the population doesn’t like. Would it be possible to have a runoff if the most popular option does not gather 50% of the vote?

1 Like
11

@saxmaam - We will see if such event happens with the new multiple-choice voting system.

1 Like
12

@saxmaam Basically, if everyone only picks one time setting type, the result will be the same as the current system. However, voters typically pick 2 to 3 choices with the system I propose, because they usually don’t have strongly differentiated preferences. Hence, the final result is something the majority agrees on. This method allows compromise.