Re: Re: Masters in Extended Contest

From: Alex Ferguson <abf_at_...>
Date: Thu, 11 Oct 2001 14:28:10 +0100 (BST)

Tim Ellis:
> I forgot the mastery cancelling rule for one fight, which meant that
> many of the contestants were fighting at around 2W - 7W rather than
> 2 - 7. It made the fight a little faster and generated more transfers
> than would have been the case otherwise, but didn't cause too much
> damage to the game. Mind you I'm not sure the same would be true if
> the scores had been 12W - 17W, 17W - 2W2 or 2W3 - 7W3...

Well, the first one is the nasty case: then you get into the "Pendragon syndrome". First one to neglect to crit's a sissy (and in KAP, dead to boot). Basically profilerating ties isn't something you want to do unless there's a clear narrative reason for it, as opposed to it happenoing or not due to the vaguarities of combinations of TNs.

> > (The 10W is a somewhat arbitrary breakpoint, but the idea is to
> > avoid both "everyone keeps losing" and "everyone keeps criticalling"
> > being the most common result, each of which is Icky.)
> I wondered about reducing by <20 (say by 10 or 15) until one
> contestant had "no masteries" , but the non-linearity of the table
> may end up making the results quite different depending on where you
> drew the line...

Yes, it would. It's also More Arithmetic. *warding gesture* One early version (whose author will remain nameless, except to say it begins with R... and ends in "obin Laws" <g>) had a rule that said reduce the higher person's skill to a fixed number (2W, I think it might have been) -- or something like that, I may have muddled the details. (I mention this one in relatively clear conscience, since it was in the semi-public "Convulsion version".) This is the most statistically elegant solution, in that it "normalises" all contests, with none of the drunken master syndrome or such stuff, but it'd be a nightmare to actually play... (Another contest, another series of base 20 subtractions and additions, lads...)

Powered by hypermail