Re: [lojban] A ranking system for Lojban speaking proficiency?

[jkominek@miranda.org]

> This sounds cool. But what about multilateral conversations?

They can be treated a one conversation for every unique pair of
participants.

> And should spectators be allowed to vote? (Maybe you can let them
> vote and then somehow weight the results based on what percentage of
> people voted which way, so you can have win, kinda-win, draw,
> kinda-loss, loss.)

The underlying math I implemented would be ok with that, but I think
it would probably produce better results to simply have the results
of more conversations recoded. (The spectators should go have their
own conversations. :)

-- 
Jay Kominek <jkominek@miranda.org>