--- Urzahil <urzahil@d...> wrote:
Mike,
Thanks for taking the time to respond.
If you want a rating that most people will pay
attention to, then you
want to keep it simple. This one seems too complex
for what it purports
to accomplish.
[clip]
I'm not sure
it's worth the effort
just to show how experienced a player is, when a
strict "number of games
finished" would be almost as useful and a whole
lot
easier, at least to
me.
Now you sound like me. The Nazgul rating exists
because I thought that simply listing # of Games
Played, Games Won, Positions Dropped and Win % would
be better than all the other ratings combined.
Little did I know there would be a hundred players with 1
or 2 wins out of 1 or 2 games.
Clint really wants a more robust rating system. I
think there is some room for improvement over the
system currently implimented.
While the math is a bit difficult, I've tried to
keepthe concept VERY simple. Win % over last 10
positions played. Game Activity that gives one point per
position played, per game season, aged at 5% a
month.
Multiply Win % by Game Activity to get base rating,
then adjust for the game difficulty. x% more
nations costs you x% rating per year. x% more experience
costs you 1/10 x% rating per year.
I really can't think of a more simple system that
rewards people for playing a lot of challenging
games, and winning a lot.
Something this complex is also bound
to take a lot of
time to implement and track as well.
I started coding last night. I expect to have it
done in 2-3 weeks. (Quicker if I don't shower, shave,
sleep or eat this weekend).
They should be able to update the results by simply
telling it to load that day's xml files, then tell
it to dump out the html to ftp to the web site.
Seasonally, they'll be able to generate PRS reports
to email/snail mail to each player showing all the
factors that effect their PRS.
A couple screens to allow them to edit players
histories (to easily fix problems), enter drops,
track position transfers, etc.
If we focus on getting the concept right, I'll
handle the code and make sure it's a piece of cake to use.
Mike Mulka
Darrell Shimel