Open Source 3D Games

[David Nicol]

On 8/14/05, Justin Dugger <jldugger at gmail.com> wrote:

> Another solution I've long considered is a matching service. Just
> match players in terms of skill.  Cheaters, non-cheaters, it doesn't
> matter. In a way, you might consider it a handicap.  Cheaters ruin the
> game by being significantly better than their opponents, right?  Their
> unfair advantage significantly changes the otherwise fair outcome of
> the game.  What should end up happening is that cheaters play
> themselves (and perhaps amusingly accuse each other), with a certain
> amount of simply excellent players getting a most excruciating
> challenge.  Of course, this requires the active tracking and rationing
> of identities, which probably cannot be done without a fee.  The fee
> being the best way to moderate people from creating new accounts as
> workaround bans, and also serving to pay other needs.
> 
> Justin Dugger

The IGS (Internet Go Server) runs a rating and ranking service, and there
is an open protocol for clients.  Go is not solved, and IGS is practice for
meatspace games.  If I create an IGS identity that serves GnuGO running
on a powerful cluster, that IGS identity might get a pretty good rank, but it
won't help me at the go board.

I have consistently toyed with the idea of creating an IGS client that
is tightly
integrated with some software to alert me to dangers and situations that I
can't clearly see on the Go board, but I haven't done it yet.  When I finally
get around to it, my IGS rating would improve, while using my crutch.


Rating and ranking systems do not require fees.  I don't know how the
combinatorics of rating and ranking would work in a group endeavor. Certainly
the task will be more complex than the vast set of who won between the two
players that IGS uses.  Perhaps a system of peer ratings would make sense,
or simply who shot who could be plugged into the IGS algorithm, and players
with equivalent skill levels would be placed in the same arenas, with whatever
customizations they bring.

I'm actually startled to hear that this isn't being done already, since the IGS
algorithm approach

http://www.pandanet.co.jp/English/commands/term/TOC.html

has been well-documented for years.  Without handicaps, rating who-shot-who
would be simpler.  Please forward that link to people running big FPS servers
if they aren't aware of it.

An automated rating system would level the playing field, if the problem is
that cheaters have an advantage and spoil the fun for the non-cheaters.


-- 
David L Nicol