On Sep 26, 8:10*am, Papa3 wrote:
On Sep 26, 10:30*am, Andy wrote:





On Sep 25, 6:09*pm, Andrzej Kobus wrote:

One thing we could think about is to increase the devaluation for days
with significant numbers of landouts - potentially based also on the
spread of handicaps in the contest. This would make dodgy days worth
even less than they are today. It also would probably require a modest
reduction in the handicaps for lower performing gliders to
compensate.

9B

How would we ever get this right?

We have plenty of contest history as input data. There are analytic
techniques to solve for that sort of thing. I expect you could trade
off one factor (handicap multiplier) against the other (devaluation
factor) to minimize the error between relative contest points in
sports class contests and relative seeding points.

I expect something like that is what we do today to establish glider
handicaps, just without the additional factor.

9B

Please no - *multi-variable calculus not allowed in the scoring
algorithms! * :-)

Seriously though, *I think if there's any "flaw" in the US Competition
Rules process, it's that we have too many engineers and mathematicians
looking for a perfect solution to complex problems (I count myself in
that category by the way). * As a management consultant, I'm sure
you've counseled clients in the beauty of KISS - Keep It Simple
Stupid. *This strikes me as a KISS moment.

Handicaps on gliders are probably good enough to get an indicative
level *of comparison. * Obviously though, they're only as good as the
model they're based on. *While the polar is (more or less) known/
knowable, *the full range of conditions in a contest are, if not
infinite, at least pretty complex. * While basic models account for
lift strength and height, *I don't believe they can incorporate all of
the other things that go into a competition in anything but homogenous
conditions. * *Wind, unfriendly terrain, ridge flying, *thermal
spacing, *and a hundred other things affect the outcome of a
contest. * While a group of gliders flying in "roughly" the same
performance bucket will be affected equally, ships at the outlier end
of the spectrum will be disproportionately impacted by any contests
where there are larger deviations from the norm in any of these
variables.

To give a simplistic/extreme example. * We have a guy in our club who
flies a 1-26. *For a while, he also owned an ASW-20. * When he owned
the 20, he was always in the top 5 in both handicapped and
unhandicapped contests. *He also routinely wins the 1-26 Nationals.
Now, put him in a 1-26 (with it's 1.6 handicap) flying against an
ASG-29, and he's likely to finish near the bottom of that contest if
there is even one weak day. * Given that many of our nationals are
decided with only 4 or 5 days of flying, that makes it pretty unlikely
that a great pilot would be rewarded by flying the 1-26.

While that's probably the extreme example, it's useful for
illustrating the point. * You have to draw a line somewhere in terms
of "bunching" ships into a handicapped class that's close enough to
eliminate the majority of the luck factor involved in having the right
ship for the conditions. * For better or worse, the IGC has already
drawn that line, so why reinvent it?

As long as you don't have to do differential equations in the
cockpit. :-)

I was suggesting that the handicaps be based off of actual contest
performance rather than trying to model actual glider performance in
the real world - which is too complex a task. Better to take an
empirical approach. No one really needs to have to understand the
methodology, just the result - their handicap.

9B