Thats 915 votes total, I didn't realize we had so many people. On a side note: Vote results followed the order in which they appeared on the voting page, was that coincidence or did they self order? Are both the top two vetting procedures going to be initiated? They both can co-exist, Shireen 19:57, 18 May 2007 (CEST)
- I call for revote without prizes. I have a huge suspicion that many people just wanted the prize, sockpuppet or not, and just clicked on the first policy (Real vetting), which I think is why it won... and why they are in the same order. -- Nova
-- (contribs) 21:36, 18 May 2007 (CEST)
- Hmmm, I'm not really sure. It's actually a rather "attractive" police. It could very well be the only one that kept a reader's interest long enough to get read. It also gives people they feeling that their creative builds have more of a chance to make it agains cookie cutters. (which may or may not be true) Eronth 21:45, 18 May 2007 (CEST)
- I agree with Nova, a vote needs to be done that isn't entered for a prize, I have very strong feeling that many people just selected a random or the top name on the list and click vote to be entered into lottery, I doubt if the majority of voters even read each policy, or even clicked the links before voting.--Sefre File:Sefresig.pngT*C 23:50, 18 May 2007 (CEST)
- I agree with Sefre, I think a revote is needed without a prize this time. Misfate 00:10, 19 May 2007 (CEST)
- I agree with Enroth's suposition. But, I said it on Cardinals talk page and I will say it here: Perhaps the admins should sit down and take the best pieces of the proposed policies and made one that worked. Almost everyone wants a simple list of things:
- (1) A defined ratio of favored to unfavored votes to determine level of favor.
- (2) A simple way to give builds a hard rating outside of the totalitatian favored/unfavored,
- (3) a superior sorting and organizational system,
- (4) An emphasis in the sytem that encourages testing and/or actual critical thinking about a build.
- We should take Cardinals (the winner's) as the skeletal template (simplicity) and incorporate those features from the other policies that make sense. If not, a LOT of people are going to be shaking their heads at this for some time. His sytem is far simpler than mine, but does need to be tweaked in a few areas. Shireen 00:23, 19 May 2007 (CEST)
- If you are interested, I posted a response on Cardinal's user page.
*Defiant Elements* +talk 00:25, 19 May 2007 (CEST)
- If you are interested, I posted a response on Cardinal's user page.
- We should take Cardinals (the winner's) as the skeletal template (simplicity) and incorporate those features from the other policies that make sense. If not, a LOT of people are going to be shaking their heads at this for some time. His sytem is far simpler than mine, but does need to be tweaked in a few areas. Shireen 00:23, 19 May 2007 (CEST)
I think we should take a servey of what people would vote for. with no reward. then after seeing the results if it looks like two many people who dont care/sockpuppets voted we can revote.--Coloneh 00:47, 19 May 2007 (CEST)
- So in other words, a vote to see if the majority want a revote? That does sound like a good idea. --tsukan 01:03, 19 May 2007 (CEST)
- Gah, put a space in front of that. tsukan 01:05, 19 May 2007 (CEST)
- How much longer are we going to Lock the whole site down? Lets just get something implimented and argue along the way. Real Vetting has the skeletal structure for a good system, and sock puppets or not, won the vote (even above mine *sniff*). Lets just get it up and improve it as we go. I'm tired of the gridlock. Shireen 01:21, 19 May 2007 (CEST)
- Gah, put a space in front of that. tsukan 01:05, 19 May 2007 (CEST)
Prettiest picture?
To me it looks like people voted for the one with the prettiest picture, I read though all of the discussions on each and well... there are some very valid criticisms of the chosen scheme, but it is the one with the best picture. --Nela 01:24, 19 May 2007 (CEST)
- Zactly what I said, it was the one to keep the readers interest longest. Looked effective enough...and, stuff. Eronth 01:27, 19 May 2007 (CEST)
Results
Just to end all discussions about cheating, changing or of new/spam/anonymous users changed or influenced voting results I decide to filter results the way some people wanted voting to be.
Originally there was 915 votes. That includes all registered users on this site. We all know results.
How ever, for this proff-of-the-vote I decide to take only votes from people who contributed (edited anything at all) at least twice.
In MySQL : user.user_editcount > 1
Then I counted all results. Adding this filter removes a lot of votes and lefts us with 116 votes. And results are:
- Percentage Favored : 33.
- Real Vetting : 37.
- Test Before Voting : 21.
- True Build Ratings : 9.
- Ranked User Vetting : 8.
- Rating Based on Tag : 9.
I am sorry to tell you, then even if only old-timers voted, only people who contributed voted, if there was no prize and only 100+ users voted: Real Vetting is a winner.
Admin can request database dump any time and double check this. GCardinal 01:33, 19 May 2007 (CEST)
- I suppose we should now wine how first past the post is a broken electoral system... ;), well there's no accounting for taste is all I can say --Nela 03:40, 19 May 2007 (CEST)
- I'll do it for you, Nela.
- Assuming that 915 users did in fact actually vote, only 33.9% (just over one-third) of the voters felt that "Real vetting" was the best policy. Assuming that only 116 users actually voted, only 31.9% (just under one-third) of the voters felt this way. Put another way, something like 68% (two-thirds) of your editors felt that real vetting wasn't the best solution.
- There are a couple problems with plurality/first-past-the-post electoral systems. Mathematically speaking -- since that seems to be your strong point, GCardinal -- plurality voting violates the Condorcet criterion, the Condorcet loser criterion, and the IIA criterion. I'd bring Arrow's theorem in to defend you, but you chose plurality voting, which doesn't fall under Arrow's theorem.
- Beyond this, on your voting script page, you presented the alternatives in a static order, with your plurality winner on the top of the list. Basic sociological knowledge or polling theory makes it pretty clear that this gives it an advantage. Additionally, you included a prize for voting, which makes your results suspect.
- Finally, what's your margin of error? Since not everyone who visits this site voted, your results are more correctly interpreted as a poll of a sample of the population. Now, I don't feel like looking through my local stats book -- I hate stats -- but I can promise you that your small sample size -- especially if we only include old-timer votes -- doesn't get you more than 90% confidence. I'm abusing math a little bit here, but let's say each of your values differ by +/-10% from the "real" opinions of the populace. Using your complete sample, this gives you the following percentages:
| Policy | Votes |
|---|---|
| Real Vetting | 33.9% |
| Test Before Voting | 24.5% |
| True Build Ratings | 12.1% |
| Percentage Favored Vetting | 11.9% |
| Rating Based on Types and Goals | 10.3% |
| Ranked User Vetting | 7.3% |
- Now, let's drop 10% off the top choice, to see what the "actual" thoughts might be. I'm not adding that 10% anywhere else, because I don't need to do so to make my point.
| Policy | Votes |
|---|---|
| Test Before Voting | 24.5% |
| Real Vetting | 23.9% |
| True Build Ratings | 12.1% |
| Percentage Favored Vetting | 11.9% |
| Rating Based on Types and Goals | 10.3% |
| Ranked User Vetting | 7.3% |
- Even at 95%, by bumping down real vetting by 5% and upping test before voting by 5%, you obtain the same ordering -- that is, even with these poll results, your userbase might very well prefer Test Before Voting to Real Vetting, even by a plurality scheme.
- In short, the results of this vote are completely worthless.
- Tanaric 03:14, 27 May 2007 (CEST)