2006 ENnies Judge Voting Poll/Thread

[fusangite]

I question if "4000" is a true threshold. There is only 1 candidate with a postcount between 1850 and 4700. While that particular candidate is not doing well, I'm not sure you can prove that it is the poscount threshold that is holding them back. I feel that a candidate with as few as 2000 posts (spread across several areas of the boards) could actually win.

You're probably right here. I suspect that due to the paucity of data, a number of my other thresholds aren't that real either. Just offered as a starting point.

Just curious here, but what factors do you feel are most likely to keep a candidate who meets most of your "floor" standards from ranking in the top five? Given the analytical lengths you've already gone to, I figure this question has already crossed your mind?

Yep -- here's my unstated belief: What you say in your reviews and posts matters a whole lot. If you write reviews and posts that either people don't agree with or don't remember, they won't earn you many votes.

 

[MavrickWeirdo]

As of 8:30 this morning we have 614 votes, over 112% of last elections voters

429 	Teflon Billy (Jeff Ranger)    		        69.87% 
398 	Crothian (Chris Gath)    			64.82% 
345 	JoeGKushner (Joe G Kushner)    		        56.19% 
329 	Cthulhu's Librarian (Richard J. Miller), SC     53.58% 
215 	diaglo (David Temporado)    			35.02% 

157	Ankh-Morpork Guard (Graham Johnson)             25.57% 
140 	Xath (Gertie Barden), SC    			22.80% 
100 	Umbran (Arnis Kletnieks)    			16.29% 

 82 	Keeper of Secrets (Matthew Muth)    		13.36% 

 74 	trancejeremy (Jeremy Reaban)    		12.05% 
 70 	Quickbeam (Kevin Bopp), SC    		        11.40% 
 60 	RavenHyde (Selma McCrory)    		         9.77%

 60 	nakia (Nakia S. Pope)    			 9.77% 
 45	Tarondor (Scott Nolan), SC    		         7.33% 
 42 	Mixmaster (Leslie Foster), SC    		 6.84% 
 36 	Eridanis (Matt Bogen), SC    			 5.86% 
 31 	JediSoth (Hans Cummings)    		         5.05%

 

[fusangite]

There is much less strategic voting than I suspected there would be. The 615 voters so far have cast 2618 votes or 4.25 votes of the five potential votes. This suggests an actual majority are using 100% of the votes available to them. This suggests to me that most voters are motivated primarily by the desire to pick the best overall panel possible more than the desire for their favourite candidate to be assured victory.

It may also suggest, however, that because a minority of voters are used to multi-member plurality voting (most are engaging in single member plurality only if they live in the US or Canada and mixed SMP-list systems in Europe) that many have not got the hang of how to vote strategically in this particular system. Although the point was made early in the thread, it may be that some voters are not thinking through the fact that their votes can often cancel each other out. This might also help to explain why the votes against the incumbents are so inefficiently distributed given that many seem to be motivated by a general desire for change rather than strong loyalty to individual challengers.

 

[Alzrius]

It may also suggest, however, that because a minority of voters are used to multi-member plurality voting (most are engaging in single member plurality only if they live in the US or Canada and mixed SMP-list systems in Europe) that many have not got the hang of how to vote strategically in this particular system. Although the point was made early in the thread, it may be that some voters are not thinking through the fact that their votes can often cancel each other out. This might also help to explain why the votes against the incumbents are so inefficiently distributed given that many seem to be motivated by a general desire for change rather than strong loyalty to individual challengers.

...riiight.

I say, the politics of failure have failed. We need to make them work again. We must move forward, not backward. Upward, not forward. And always whirling, whirling, whirling toward freedom!

 

[freebfrost]

I say, the politics of failure have failed. We need to make them work again. We must move forward, not backward. Upward, not forward. And always whirling, whirling, whirling toward freedom!

That sounds suspiciously more like the Politics of Dancing to me...

 

[MavrickWeirdo]

Just for the fun of it I decided to look at where some of our candidates MIGHT be a year from now (re: postcount) if their posts per month rate remains the same.

(Keep in mind this is the type of speculation which is considered meaningless by serious statisticians, due to the number of variables involved.)

Ankh-Morpork Guard	12967
Umbran 			 7461
trancejeremy		 3425
Xath			 2757
Quickbeam 		 2281
Eridanis 		 1842
Keeper of Secrets	 1732
nakia 			  943
Mixmaster 		  408
JediSoth 		  313
Tarondor 		  155
RavenHyde 		    5

 

[fusangite]

...riiight.

I didn't know that I was being unclear. What was problematic about the post?

 

[CRGreathouse]

I didn't know that I was being unclear. What was problematic about the post?

I thought it was a great post, and I agree with you that it seems likely that most people just don't know how to vote strategically in this sort of election, especially with this many candidates.

 

[Conaill]

I think the roughly 5 votes per voter probably has more to do with Dextra's original voting instructions.

Perhaps a more telling measure in this case is to look at changes in voting behavior over time. Since the results of the poll are not blinded, one would expect the pattern of votes to shift as people take the previous votes into account.

In particular, one might expect to see a gradual decrease in the average number of votes per voter over time, as people see that fewer of their votes will actually "matter" (becaus ethe top places are already more or less locked in). One would also expect to see a shift from voting for the top candidates to voting for the second-tier ones (i.e. those around rank 5, where new votes may still create a change in the outcome) and lower (protest votes, and people using up their 5 "alloted" votes to indicate preferences below the 4-5 candidates who "will win anayway").

So what do we actually see in the data? Looking at the voting patterns at 190 votes, 408 and 614 votes (i.e. roughly 1st, 3nd and 3rd block of 200 votes) we see that...

1) the number of votes per voter has stayed approximately constant: 4.27 vs 4.11 vs 4.39

2) out of the top 5 candidates, only *one* has shown a significant decrease in votes over time, and that's JoeGKushner.

3) in the "second-tier" candidates, i.e A-M G, Xath and diaglo, there has only been a slight uptick in votes over time,

4) the lower tier of candidates hasn't really seen a very significant increase in votes over time either.

Overall, the voting patterns haven't really changed much over time, despite what one might expect for such an open poll. So in that sense, I don't really see any evidence of strategic voting either.

(PS: before anyone hammers me for this, I used "significant" in the colloquial sense here. If I had more detailed voting data, I could randomize vote order and put some actual p-values on there, but I'm not *that* much of a geek that I'm going to do real statistics on a beautiful Saturday morning. )

 

[Alzrius]

I didn't know that I was being unclear. What was problematic about the post?

It wasn't so much that it was problematic, as it just seemed over-analytical. People just vote for who they like and that's the end of it, or so I thought (maybe I was naive). Breaking it down into statistics seems to cross that threshold of looking for meanings that aren't there. Again though, that's just my (rather simplistic) take on it.