"Askvillers - Please comment on the new blog post about Askville's proposed voting system."

So glad to see you posting! That's funny! 

I just thought about this--are you going to rate our answers, Mr. Mayor!? You better! Gotta get those stats up, you know!

So I read the Blog...it sounds good, like it will probably work, but it's going to get complicated! Is that the only way to keep things fair, to add more math to them mix?  Someone spent a lot of time coming up with all of this!

I like that everyone will vote (I think!), and that the voting superpower will be gone... And no minus coins when rated low... But we're getting more and more elaborate configurations figuring into the scoring mix, and getting more complicated...  And it's something that almost has to be actually used first hand to fully comprehend how it will work and what the advantages and disadvantages will be, it seems to me.   

The part that concerns me is the new "weighting" of the answers based on a persons level.  This on the surface seems good, but then I start thinking down the road...we won't really be able to mathematically be able to determine the final score, even if we can see all the votes, without the Askville systems adding into the votes the "weights" of the levels, if I understand this correctly? Unless we can do the algorithm, we won't be able to figure the final results.  Which is one of the things about the Superpower voting that was confusing.  No one knew the impact of the superpower vote mathematically.

Will our "weight" be made public (other than our obvious ranking in a topic)? So that we have some kind of factor, a number, to know what we are worth?  It will be like "hey, throw some weight on this question, will you!". 

"Best Answer ribbons will still go to the highest rated answer that is greater than 4 stars. "  No mention of timing of answer on this, I see. If there are no ties, I guess the timing of an answer will be taken out of consideration.  That will make things more fair, I think.  I know some people hesitate to answer if they aren't one of the first.

I'm trying to figure out if this new plan will help the questioner get better answers, or if it's intended to deal with the answerers.  I don't think it will really change the answers. But it will get more involvement in the voting, and that's a good thing.

If you are not an answerer, and you vote on an answer, will there be a coin for voting awarded?  Many people won't use even the thumbs up voting, as there is no reward or effect for doing so. They can change the outcome for best answer, if they vote, but will they be rewarded for doing so?  Some will anyway, I'm sure, but others won't if there isnt' something in it for them.

Does this mean the thumbs will be taken away?  I'm sure many people will like that.

Voting accuracy rating....will that encourage us to look at the answers, and think which one will be the one that most people like the most, rather than which one we feel is the best, if we want to increase our accuracy rating? And you know we will--something new to compete and excel in--"hey, what's your accuracy rating?? Mines 99%! Beat that!"   I prefer LESS competitions here, other than in the actual questions...as soon as a new "number raising" feature is introduced, a new competition is, too. 

Ok, now I'm trying to figure out how this can be abused...will the strong opinioned people "gang up" on questions to vote en masse??  Will a discussion be created telling people where more votes are needed to sway the vote, now that everyone can do so?  It does seem it would take more effort to sway things, but you know how smart the people are here!  SOMEONE will figure out the way!

I'm sure more thoughts will come to me as I think about the proposed changes some more.

Thanks for allowing us to participate in the decision, Mayor and staff!  And your avatar is cute! 

(now don't forget to rate our answers!!)

I like some of the things they’re proposing - like letting everyone vote - tho’ it takes away some of the "specialness".

The stars sound like a good idea.

The Voting Accuracy, to me, is something that could be easily skewed if people decide to vote an answer "poor" because they don’t agree with it, or don’t like the user.  Of course, that can happen now - it just doesn’t carry the weight it would if they change to the new system.

I don’t like the part about no Best Answer Ribbon if there is a tie...sniff.  I love the ribbons!  Please keep the ribbons!  It's like "positive reinforcement"!

Except for hot religious and political questions where I expect massive voting "piling on" by many askvillers in accordance with the numerically dominant views on this site. This will be self-validating because their votes will determine the final answer ratings very heavily. People with minority views will have the opposite experience and their votes will wind up counting less and less as time goes by.

This will result in hesitancy to express any minority opinion against the known social biases on this site. I am sure opinion questions like the above will be avoided so as to preserve their "voting accuracy scores" so their own votes do not become worthless on every other topic as a result.

I like the five star voting system as proposed very much.

I also think that if you vote you should have to comment and explain your vote in the discussion.

I have not read what others think, but this also means my opinions have not been swayed by anyone. My suggestions are in no way perfect, but I strongly think my general ideas are pointing at the right direction.

1. The most significant change: Anyone can vote? I don’t think this is a very good idea. First, the asker’s vote should still carry greater weight, since it is his or her question, and answerers’ votes should also have greater weight than a stranger vote. We should not take away their priviledges in participation.

I also see potential "mob action" too if an answer is not popular or divisive, and voting for friends is way too easy - think coin-gaming. The randomness of stranger votes might help assuring fairness but it is not guaranteed. Especially in highly opinionated questions.

It seems to me that the proposed system has gone from one extreme to the other. Currently a lot of answers are not getting enough votes, but the proposed system will result in TOO MANY votes.

My suggestion: Not every question needs stranger votes, if it already has 5 votes from the asker and the 4 answerers, we should leave it alone. For answers that have less than 5 votes, the system can solicit more votes from other members RANDOMLY to reach 5 votes. The asker’s vote still weigh as 2, the answerers’ votes still weigh as 1, and the stranger votes weigh as 0.5 multipled by the user’s level in the topic then multiplied by voting accuracy percentage. So if an user is level 2 with 80% voting accuracy, his vote will weigh 0.8, 1.2 for level 3 with 80% voting accuracy.

For example, if an answer has only 1 vote, the system will send voting requests randomly ONLY to 4 other members who are above a certain voting accuracy threshold (more on that later). Each member will have a "voting invitation" box and he or she must respond within a day otherwise the invitation will be rerouted to others randomly, and they get 1 questcoin also for every vote just like the asker and answerers. And only those invited can cast the stranger votes.

This way I think we can keep the best of old and new. And, with care I also support upping BOTH the 5-answer and voting limit to 6, and everything else proportionally. Even higher if the results are good.

2. Removing superpower voting is a good thing. It has not been used often anyway.

3. Removing negative experience points all together might not be a good idea. 2 star is identical to a weak score before, so why would a weak score get 1 questcoin nonetheless on top of the 1 coin he earns by answering? I will agree that the penalty can be less, but not without. My suggestion is that 1 star will lose 3-5 experence points and 1 questcoin, and 2 star gets 0 on both. And I now don't support mandating explanation of negative vote - I used to suggest and support that.

4. This is just like the current system, reasonably unchanged.

5. About weighing. This has to be in the context of point number 1: who can vote. Now if my suggestion is implemented, there is automatic weighting built-in in regard to the user’s topic level and voting accuracy.

More about voting accuracy - my suggestion has a threshold which will qualify a user to be selected to vote. This will penalize those who has bad voting accuracy. The threshold can set, for example, as 70%. Initially everybody will be given 100% (I don’t think we should calculate from history since that was not in the rules to begin with), so everyone can be selected to vote for undervoted questions.

6. Using decimals to break the tie is much better. Totally agree.

7. No best answer selected when there is a tie? I am Okay with that.

Hope I have helped.