Picture of the author
Topic :
Thread Created at Invalid date | Started by
Number of Post in this thread: 2Please Sign In to comment on this Thread
mattman1971 replied 3 years ago (Sep 14th 2020, 2:26:25 pm)

Given the explanation, I agree that the method is the best shot we could hope for affixing a "good enough" label to it. I'm unable to find the statement that I objected (softly) to, so I'm not sure if it's something that could be fixed with words for reading by future skeptics, without the need for personal intervention to expand on the details. At the same time, I'm always hesitant to suggest more words because that's my personal downfall, too many words. XD

This thread doesn't really help me get to exactly the thing I objected to -- I'm pretty sure that even though I asked a question, I still voted in support of the statement, but now I'm not so sure.

Brent_Allsop replied 3 years ago (Sep 13th 2020, 7:19:22 pm)

Hi Mattman1971,

I just noticed that you objected to the statement we submitted for this camp. With stated reason: How can objective evaluation of expertise be based on popular vote?

 

Thanks for the participation, but I have a few problems with this.

 

  1. This is just the factual description of how this particular algorithm is currently coded, which is also the same algorithm that was used and documented, in the old 1.0 Canonizer system.
  2.  The popular consensus part of the algorithm is just the minor first half.  It is a bootstrap resolution to the, chicken and egg problem.  Once you have at least a popular consensus about who the experts might be, you can then determine the true expertise by having these experts rank each other.  So, it is the second half of the algorithm that really counts, not the popular part.
  3. We have had at least one skeptical person review this algorithm, who finally admitted that it actually works quite well, given the chicken and egg problem.
  4. If you have a better way to get around the chicken and egg problem, feel free to propose it.  And if some people like this Canonizer algorithm better than yours, you can always help us create another different Canonizer algorithm, so people can chose whichever one works best for them.


I'm going to resubmit this, given the above explanation of how we can get what you want, without destroying what other's want.