[blml] Swiss Teams formats

[Steve Willner]

> From: Herman De Wael <hermandw at skynet.be>
>  Could you investigate if simply allowing IMP-difference 
> to determine the ranking would influence the outcome in a 
> positive/negative way?

I can't imagine it would make any significant change.  The WBF VP scales 
are almost linear already.

> We are currently looking at ameliorating the Final of the Flanders' 
> Cup, which is currently a Swiss over 9 matches of 6 boards, with 64 
> teams! The better players are starting to complain that it is becoming 
> a lottery.

I suspect the most effective single change one could make is to correct 
raw VPs for "strength of schedule."  That is, teams should be credited 
if they have played against stronger opponents.  Other than that, my 
guess is that small improvements are possible but not big ones.

> From: Robert Geller <geller at nifty.com>
> the inherent variability and randomness in bridge
> is sufficiently great that the present imp and VP scales
> are "good enough for government work."

I tend to agree about the IMP scales.  The WBF VP scale can undoubtedly 
be improved, though I doubt it will change outcomes by much in WBF play. 
(The near-linear scale is, however, wholly unsuitable for events that 
include a wide range of abilities.)  Other changes in conditions of 
contest might have a significant effect.  The reason for doing 
simulations is to test what matters and what doesn't.