David has a new innovation at the sewer. The old comment voting system has been enhanced to, amongst other things, hide comments when they receive enough votes against. I had some fun testing this last week under various logins that I set up some years ago and keep active. It looks like an innovate quantitative social testing tool. You can see how it is useful to an organisation like Curia with its close ties to the National party polling system.
How does it work? Well not as big bruv* expected when I was using my lprent login.
My response to big bruv also wound up being hidden. It said that I was interested in how the rating system was operating, and explained why he actually managed to get banned at The Standard in the past.
In fact all of my comments as lprent got hidden. While sarcastic in parts as is my usual style, each comment contained some substantive valid points or was responding to comments directed at me. See them for yourself. 1, 2, 3, and 4.
This type of innovative hiding of comments based on anonymous voting is interesting as it removes all level of personal responsibility for what is effectively a ban. It significantly reduces the feedback that follow-up comments by other commentators or moderators would give. As such you’d have to question its value in operant changes in behaviour.
Explaining the new rating system, David described it thus (my italics).
Comments with lots of positive ratings get highlighted. I have set the threshold for now to 5. I may need to raise it. I don’t want too many highlighted just those which get lots of people saying this is great.
Comments with lots of negative ratings gets hidden you can still read them if you want by clicking on them, but it allows you to skip over them. Now I have set the threshold at 10 negative comments, as I don’t want lots of comments hidden. I will increase this if people are voting comments down just because they may not agree with a comment, as opposed to it being a comment of poor quality. There is a huge difference.
Well I’d agree in principle, however you haven’t succeeded. Hopefully I’ve assisted in providing calibration data for your system, it looks like a threshold probably needs to be above 75 negative comments before the sewer rats stop voting against the person rather than the comment. However that is a ridiculous level. As it stands it simply entrenches bullying as the dominant social behaviour in the sewer by the majority in-group in a way that I have only seen in Hollywood high school dramas.
If that was not your intent, then you should probably limit the number of votes per period after the comment is made. That would allow your commentators to assess a value per each vote If you set it to say 10 votes per login per 24 hours, then less excitable people would have to think before using their votes up early. The coding would be trivial. Of course that also put a value on having several aliases that they use. But at least there is a cost to logging in and out that will reduce idle block voting.
But there are other interesting aspects to this system. Prior to this I’d left several comments to get a feel for the tolerances of the local in-group. These were from various well-established alter-egos on Kiwiblog. The results were quite interesting as an example of pack behaviours.
You’d have to say that this system, while low maintenance in terms of moderating effort, is very good at stifling robust debate on kiwiblog to a relatively small cohort of possible opinion sets.
I’m not expert at Skinner and related techniques. But this type of technique in operant conditioning (using punishment without clear feedback) is often used to drive test subjects either psychotic or into a hopeless unthinking conformance. Fortunately the commentators are not confined, so do have a choice of moving to other sites that do allow some freedom of expression which have clearly defined behaviour standards. Public Address, The Standard, or even No Minister come to mind.
However I’m surprised that lab-rat usage wasn’t highlighted in David Farrars description of the rating system. The benefits should have been immediately obvious to someone who runs a social polling company like Curia. I intend to spend some time to write code to strip this valuable quantitative data from the public pages. It’d be far easier for David to do from the database, especially as it tracks who makes each vote as you cannot rate a comment more than once for a given login.
Because the bully system that the voting tends to assist, people whose comments get hidden will tend to not comment, leading to a purer test sample of the target audience. This will steadily increase the efficacy of the system to a polling company.
It must be a very happy mischief in creating your very own socially in-bred and highly conformant lab group of wingnuts for testing.
* big bruv under that psuedonym was last banned for poor behaviour in April last year for tiresomely attacking this site rather than talking about anything with any substance. He has done a hundred or so comments since then with a few notes by moderators. Obviously the bans left a suitable, but not deterrent, impression to him expressing himself at odds. Just not repeating behaviours that were unacceptable – which is the intent of the moderation and bans.
updated: Spelling mistake in title.
Powered by WPtouch Mobile Suite for WordPress
