[blml] Swiss Teams formats [SEC=UNOFFICIAL]

[richard willey]

On 4/5/07, John Probst <john at asimere.com> wrote:

> The idea of playing shorter matches earlier and longer ones later is
> interesting; It says:- later rounds are more significant and teams must have
> more time to establish their true worth
>
> Say you play 5,6,7,8,9,10,11 board matches for your 7 rounds. You might just
> as well use different VP scales and play 8's throughout because you are
> saying that later rounds are more significant. Thus if the 8 board match
> scores 20-0 for a 30 imp win (per UK); you just create VP scales say 0-14 rd
> 1; 0-16 rd 2; 0-18 rd 3 etc etc with a max being round about 28-32 imps
> depending on the number of VPs. This makes the later rounds more
> significant, as indeed does playing matches of different lengths.
>
> John

Hi John

I don't think that holding board length constant and adjusting the
scoring system accomplishes the same thing.  When I originally
considered the idea of adjusting the length of the rounds, my primary
concern was to improve the accuracy of the sample process.

During the early rounds of a Swiss or Danish Teams event its highly
likely that teams with widely disparate skill levels will be playing
together.  Ideally, as the event progresses, the skill level between
contesting teams will converge towards one another.  I made an
assumption that the length of a match should be inversely proportional
to the skill difference between the two teams.  (The close the two
teams are in terms of skill level, the more boards you need to play to
accurately evaluate their relative skill level).

The initial simulations that we're running are (primarily) intended to
test this hypothesis.








-- 
Aristotle was not Belgian, the principle of Buddhism is not "every man
for himself," and the London Underground is not a political movement!
Those are all mistakes, Otto. I looked them up