X-Message-Number: 32438
From: David Stodolsky <>
Subject: Re: Continuing Message Suppression with CryoNet's Rating Syst...
Date: Sun, 28 Feb 2010 12:00:41 +0100
References: <>

On 27 Feb 2010, at 11:00 AM, CryoNet wrote:

> Why do you continue to ignore the "vote" of those who do not rate
> messages at all?

The major problem here is that with a mail system it is impossible to  
know whether someone has read a message. In the publication model I  
presented earlier, this isn't a problem, since the main role of  
feedback is to determine who should be invited to contribute the next  
message in a thread.

> The fact that someone has not thought it worth their
> time to register a specific rating of a message, even though they get
> the digest daily, is worth something in the algorithm - I do not see
> that this happens. This failure in my estimation distorts the entire
> system and reduces the value of CryoNet to me, and likely to many  
> others.

What we need to assess the performance of the current Rating System is  
some participation statistics. If the participation rate is low, that  
means that incentives for participation are not working and then only  
those who want to censor are affecting the ratings.

I performed a random sample of 10 of the 32XX range of CryoNet  
messages. Of these, 2 were my own, so I couldn't see any ratings.  
There was only 1 rating contributed in response to the remaining 8  
messages. We can conclude from this that the Rating System is not  
being used significantly at this time, except by those who wish to  
censor. This explains why we are seeing censorship.

The quick fix for this is to limit rating to a few people who are  
thought to be impartial. Ideally, they would be willing to regularly  
evaluate messages just before the Digest is sent out. This fix will  
not improve the quality of messages, but it will stop censorship and  
spammers, etc.


David Stodolsky
  Skype: davidstodolsky

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=32438