Very interesting... i'm guessing how many msx users are in the world (and see mrc) that are making 100+ more votes on the first two entries...
Are we so many?
Of course some of us cheat, that's the only negative aspect of allowing anonymous votes. Like, vote X from home, work, girlfriend's place, parents place, and ask a few MSN buddies to vote X for you.
Is it possible to split the results in the newspost in 1: all votes, and 2: votes from MRC members that have a combined total reactions+forum messages of 3? (2: if an MRC member didn't vote, and the IP that the MRC member mostly logs in with did vote, include that too... eg. I only log in if I do a reaction or forum post)
Yes, the less honest voters (active MSX community is small), the more serious effect a bit of cheating will have.
("combined total reactions+forum messages of 3" in my post above should of course be "combined total reactions+forum messages of 3 or more")
Sure I voted both from home and from work but I don't think two votes instead of one matters much. I actually think the broader audience give a better picture than just the 30 or so active MRC users. Just looking at the statistics, there are currently 3 MSX friends (registered users) and 170 guests online. So I'm not sure that the vote is that bad. I'm sure some people vote at 3-4 places but it still doesn't matter much. I checked some earler polls and the ones that are more interesting get up to 700 votes. I doubt those votes are from one person going to all neighbors and friends just to get his opinion through (but maybe I'm wrong...)
The strange trend of the first two major competitors makes me think there is some tricky people...as i said in previous post.
i do not expect, especially in the last two weeks a countinuos race between the first two main positions while the others are substantially costant in the number of votes they get.
What i expect, in a similar poll, is that after a relatively short time ( a week), the percentages of those entries do not change too much.
Because of the importance of this poll i expect almost all active mrc people to vote in a period of a week, so the picture should be almost clear just after that period...
But seeing at the results, this has never happened....
Yeah perhaps. I was a bit surprised too. After a week UR had over 40% of the votes, and then a week later it was 30%, But I'm not sure if its anything strange, the other three entries have increased about the same amount (give and take a percent or two). My guess is that the active MRC users that voted the first week voted more in favor of UR, maybe because of emotional reasons or because of the importance of that demo, not sure. Later voters are probably less biased. I got interested in the poll the last week (mainly because it looked like Unleashed were going strong) but the result hasn't really changed much. The biggest winner last week was Sphere that gained over 4%, and ML and UR lost about 4% combined, while Unleashed stayed the same.
Or maybe the later votes is simply less interested in the actual thing they vote on and more to get close results. I actually remember the same sort of finish in another poll where it looked a bit un-statistical. But its not that its been 1 vote difference the whole time though. The gap between #1 and #2 has been anyhwere between 0 and 10 votes the last week so maybe its just the average that seems weird. An interesting observation is that the gap often increased evening time in Europe and decreased evening time in Japan (thats morning and afternoon here which is when I typically read MRC).
Btw, are you only using IP address to find unique voters? You could use cookies or something as well... Some ISP's change the IP address of the users, so if you're using IP address only, you can get many votes from one user, probably much more than 3-4.
With an open poll you'll never get an exact result. A lot of websites have polls, including respected newspapers. With big voting numbers I think you'll actually get a pretty good representation of what the readers of the website think and it doesn't really matter if a small group votes multiple times. The issue here is that the total number of votes is relatively small, so perhaps someone could actually get enough extra votes to change the outcome. Its quite a lot of work though so a person trying to change the outcome of an MRC poll, probably has a pretty boring life.
even large scaled open polls don't mean they can't be rigged... a dutch website geenstijl.nl proved that by asking it's readers to massively vote for the name 'geenstijl' when Doritos was deciding what the new flavor of their dorito's should be called in a public vote. They won by a landslide...
(same can be applied for imdb ratings.... many people voting either 1 or 10 to 'compensate' an either too low or too high rating for their personal taste)
Comments (20)
By [D-Tail]
Ascended (8263)
15-03-2008, 20:50
By Manuel
Ascended (19298)
17-03-2008, 19:55
By Samor
Prophet (2165)
18-03-2008, 23:16
By Latok
msx guru (3927)
21-03-2008, 16:11
By poke-1,170
Paragon (1783)
23-03-2008, 02:28
By hap
Paragon (2042)
23-03-2008, 15:14
By PingPong
Prophet (4093)
07-04-2008, 20:57
By PingPong
Prophet (4093)
13-04-2008, 13:29
By hap
Paragon (2042)
13-04-2008, 14:50
By PingPong
Prophet (4093)
13-04-2008, 19:38
By hap
Paragon (2042)
13-04-2008, 20:01
By dvik
Prophet (2200)
13-04-2008, 20:54
By PingPong
Prophet (4093)
14-04-2008, 01:06
By dvik
Prophet (2200)
14-04-2008, 01:12
By snout
Ascended (15187)
14-04-2008, 10:11
By dvik
Prophet (2200)
14-04-2008, 15:37
By dvik
Prophet (2200)
14-04-2008, 15:46
By PingPong
Prophet (4093)
14-04-2008, 19:48
By dvik
Prophet (2200)
14-04-2008, 20:09
By Low_Profile
Champion (425)
15-04-2008, 14:40