Joe Bernstein  wrote: 
 >In article , Todd Michel McComb 
 > wrote: 
  
 >> If this does continue as an issue, or expands as an issue, we might 
 >> have to consider some real changes to the way voting works. 
  
 >Have any of those with archives of RESULTs analysed them to see what 
 >effect on the voting process diligent blocs of 15+ voters would have? 
  
 >If the Strombolis keep at this, my gut expectation is that it would 
 >have a major effect on future votes.  The number of close votes 
 >doesn't exceed the number of not-nearly-enough-YES votes, but my 
 >*impression* is that the breakdown is vaguely like this: 
  
 >not-nearly-enough-YES   60% 
 >close, either way       35% 
 >way-over-enough-YES      5% 
  
 Well, in 2003, I found 19 votes, of which: 
 5 had more than enough to pass 
 9 couldn't have passed even if they had 20 more YES votes 
 3 might have passed if they had 20 more YES votes 
 2 would have failed if they had 20 more NO votes (or would look like 
    clear passes if they had 20 more YES votes) 
  
 Note, this means 12 votes (65%) resulted in fail, and 7 (35%) 
 resulted in pass. 
  
 So, roughly: 
 not-nearly-enough-YES   50 +/- 5% 
 close, either way       25 +/- 10% 
 way-over-enough-YES     25 +/- 10% 
  
 I cursory inspection of 2002 seemed to indicate similar situation. 
  
 >For the rest of this post, I'll assume I'm right; I'll be greatly 
 >relieved if those with statistics can truthfully tell me I'm not, 
 >though, for reasons that will become obvious. 
  
 ... 
 >So if suddenly the entire "close" category is subject to unpredictable 
 >bloc voting that at least presumptively is not based on any of the 
 >things the votes were intended to measure, then basically the most 
 >numerous category of actually-created groups is no longer being voted 
 >on in a way that even vaguely relates to the purposes of voting. 
  
 I think I see what you are saying.  If the "close fails" moved up 
 to "close pass" and out of "close pass" to clear pass, the number of 
 passes nearly double in your table.  That means half of the passes 
 passed for reasons not resembling the intention of the system. 
 The actual breakdown indicates that if the "close fails" switched 
 to "pass" and so forth, the increase would increase the passes from 
 35% to 50%, which actually is not statistically significant by 2003 
 standards. 
  
 Note, that if the block voting is arbitrary (i.e. sometime NO 
 sometimes YES), the statistical effect is negligible.  Further, 
 since the affected proposals are borderline already, under that 
 condition, those kinds of proposals shouldn't expect much 
 assistance from NAN.  I mean the fate of borderline proposals 
 are by definition unpredictable.  Your point, however, is probably 
 more along the lines of if the block voting is weighted more to 
 one side of the vote. 
  
 ru 
  
 -- 
 My standard proposals rant: 
 Quality, usefulness, merit, or non-newsgroups popularity of a topic 
 is more or less irrelevant in creating a new Big-8 newsgroup. 
 Usenet popularity is the primary consideration. 
  
 --- SoupGate-Win32 v1.05 
  * Origin: you cannot sedate... all the things you hate (1:229/2)