View Single Post
  #235  
Old 04-27-2010, 04:10 PM
galena galena is offline
 
Join Date: Feb 2010
Posts: 7
Quote:
Originally Posted by safak View Post
<3 people who have no clue about a subject. hey if you make a post that is condescending, then it makes you sound right? sorry doesn't work all the time.

it IS a statistic impossiblity in a poll, for even a 50% choice get 20 votes in a row(in 20 seconds no less).

now when we are talking an option that previously was at 10% getting every single one of them, its not even a question. sorry. sound even meaner next post and you might trick us.
Just no, statistically impossible is not a statistical term. Run a statistical analysis that gives actual P values and i will talk with you about it. The term statistically impossible is completely misleading. While i think there is something fishy with how the votes changed, especially over a small amount of time. But that is my OPINION, not a statistical analysis.

Considering this is a forum based poll that was based off a vocal number of people WANTING a vote, it can be argued that a analysis of the incomplete votes would be inaccurate as the early voters may be more likely to be those in favor. But that is a whole different point all together. All i am asking is not to use an erroneous term like statistically impossible. If you must throw the term statistical into your arguments about this vote at least use statistically improbable instead of impossible.

*edit as i have seen safaks second reply to me*
Your analysis of the odds of people voting for options is fallacious as you are assuming that the distribution of the votes that have happened are the odds of it happening again. Also following that assumption leads to the idea that how votes are cast are random which they are not.

Like i said, something fishy was probably going on with people making extra accounts, but it does'nt really matter as the devs have chosen not to wipe not because of the result but because of people voting multiple times. Also the quick swing in votes just before it closed only pushed the voting below 66.6% not 66% which was the cutoff for the vote to pass as stated by the first post. Also as i stated before my edit, i just see statistically impossible as a false statement as even if i agreed with your methods for calculating the probability of 20 votes in a row for no, as you stated there is a probability of it happening and as such cannot be impossible. Impossible does not = improbable.

Last edited by galena; 04-27-2010 at 04:23 PM.