How does Winscore calculate finish altitude?

At 01:48 28 July 2007, Marc Ramsey wrote:
BB wrote:
In my opinion the clock should stop as soon as the
pilot enters the
cylinder. We shouldn't have pilots in the finish cylinder
still
racing.


You haven't met enough contest pilots. If the rules
change in this
way, pilots will aim to finish one mile out, 50 feet,
90 knots and
then float in to the landing, european-style.

Applying an appropriate penalty for finishing below
the Minimum Finish
Height would eliminate that behavior. I was actually
surprised to find
that the SSA competition rules provide no guidelines
as to how to
penalize pilots who don't make it into the finish cylinder.
Given the
difficulties of knowing precisely how high one is finishing,
missing by
50 or so feet shouldn't result in a huge penalty, but
it should also
never be beneficial to intentionally finish low...

Marc

DISCLAIMER: I understand under the new rules speed
points are no longer allocated pro-rata so as to create
bit more spread at the top of the scoresheet. My math
may, therefore, be a bit off.

The worst case scenario for making marginal final glide
decisions is on a short task where a pilot is climbing
slowly trying to make it up to final glide altitude.
The slow climb takes up lots of minutes per foot gained
and every minute drags down your speed relatively more
on shorter tasks.

So, say you are climbing at 2 knots. It will cost you
about 4 points for every hundred feet you climb, or
about 40 points to go from a white-knuckle 2-knot glide
to 0' at the finish up to a 2-knot glide to a 1000'
AGL arrival. For a 4-hour task the 4 points per 1000'
drops to 2 points per 1000'.

You could imagine a penalty structure that looks something
like: 8 points per hundred feet divided by the minimum
task time (or the winners time, or your time). This
eliminates most of the incentive to cut a last thermal
short since it is in the pilot's interest to keep climbing
if he thinks there is any chance he will be under the
minimum finish height and he is achieveing a climb
rate of 2 knots or more.

If you're climbing in your final thermal at less than
2 knots you are looking at a dicey glide no matter
what, and probably are contemplating a rolling finish.
It's not clear to me that a penalty structure built
around slower than 2 knot climb rate would do any good
- plus the penalties start to get kind of large (e.g.
16 points per 100 feet if you pick 1 knot as the climb
rate).

With the penalty structure I've described, if you finish
at 500' below the minimum finish height (so you are
at most 500' AGL) and actually fly to more or less
a full pattern it would take about 2.5-3 minutes to
get from the edge of a 1-mile cylinder to a full stop.
This is based on looking at a couple of my contest
finishes at Parowan where the runway is pretty long
and they were asking us to roll all the way to the
end. Guess what? The penalty as described above would
work out to the equivalent of an additional 2.5 minutes,
so the worst case scenario for a low flying finish,
would be no worse than taking the time to landing and
stopping. If you just barely miss the minimum height
you are a lot better off.

In terms of coming to a screeching halt in the middle
of the runway on a rolling finish - it's worth 2-5
points in my estimation. You need to weigh that against
all the other safety considerations and potential penalties
that might be imposed of you were really ver-the-top
about it. Plus the ill-will from your crew when they
have to schlep your glider halfway across the airport.

9B

At 01:48 28 July 2007, Marc Ramsey wrote:
BB wrote:
In my opinion the clock should stop as soon as the
pilot enters the
cylinder. We shouldn't have pilots in the finish cylinder
still
racing.


You haven't met enough contest pilots. If the rules
change in this
way, pilots will aim to finish one mile out, 50 feet,
90 knots and
then float in to the landing, european-style.

Applying an appropriate penalty for finishing below
the Minimum Finish
Height would eliminate that behavior. I was actually
surprised to find
that the SSA competition rules provide no guidelines
as to how to
penalize pilots who don't make it into the finish cylinder.
Given the
difficulties of knowing precisely how high one is finishing,
missing by
50 or so feet shouldn't result in a huge penalty, but
it should also
never be beneficial to intentionally finish low...

Marc

DISCLAIMER: I understand under the new rules speed
points are no longer allocated pro-rata so as to create
bit more spread at the top of the scoresheet. My math
may, therefore, be a bit off.

The worst case scenario for making marginal final glide
decisions is on a short task where a pilot is climbing
slowly trying to make it up to final glide altitude.
The slow climb takes up lots of minutes per foot gained
and every minute drags down your speed relatively more
on shorter tasks.

So, say you are climbing at 2 knots. It will cost you
about 4 points for every hundred feet you climb, or
about 40 points to go from a white-knuckle 2-knot glide
to 0' at the finish up to a 2-knot glide to a 1000'
AGL arrival. For a 4-hour task the 4 points per 1000'
drops to 2 points per 1000'.

You could imagine a penalty structure that looks something
like: 8 points per hundred feet divided by the minimum
task time (or the winners time, or your time). This
eliminates most of the incentive to cut a last thermal
short since it is in the pilot's interest to keep climbing
if he thinks there is any chance he will be under the
minimum finish height and he is achieveing a climb
rate of 2 knots or more.

If you're climbing in your final thermal at less than
2 knots you are looking at a dicey glide no matter
what, and probably are contemplating a rolling finish.
It's not clear to me that a penalty structure built
around slower than 2 knot climb rate would do any good
- plus the penalties start to get kind of large (e.g.
16 points per 100 feet if you pick 1 knot as the climb
rate).

With the penalty structure I've described, if you finish
at 500' below the minimum finish height (so you are
at most 500' AGL) and actually fly to more or less
a full pattern it would take about 2.5-3 minutes to
get from the edge of a 1-mile cylinder to a full stop.
This is based on looking at a couple of my contest
finishes at Parowan where the runway is pretty long
and they were asking us to roll all the way to the
end. Guess what? The penalty as described above would
work out to the equivalent of an additional 2.5 minutes,
so the worst case scenario for a low flying finish,
would be no worse than taking the time to landing and
stopping. If you just barely miss the minimum height
you are a lot better off.

In terms of coming to a screeching halt in the middle
of the runway on a rolling finish - it's worth 2-5
points in my estimation. You need to weigh that against
all the other safety considerations and potential penalties
that might be imposed of you were really ver-the-top
about it. Plus the ill-will from your crew when they
have to schlep your glider halfway across the airport.

9B

At 01:48 28 July 2007, Marc Ramsey wrote:
BB wrote:
In my opinion the clock should stop as soon as the
pilot enters the
cylinder. We shouldn't have pilots in the finish cylinder
still
racing.


You haven't met enough contest pilots. If the rules
change in this
way, pilots will aim to finish one mile out, 50 feet,
90 knots and
then float in to the landing, european-style.

Applying an appropriate penalty for finishing below
the Minimum Finish
Height would eliminate that behavior. I was actually
surprised to find
that the SSA competition rules provide no guidelines
as to how to
penalize pilots who don't make it into the finish cylinder.
Given the
difficulties of knowing precisely how high one is finishing,
missing by
50 or so feet shouldn't result in a huge penalty, but
it should also
never be beneficial to intentionally finish low...

Marc

DISCLAIMER: I understand under the new rules speed
points are no longer allocated pro-rata so as to create
bit more spread at the top of the scoresheet. My math
may, therefore, be a bit off.

The worst case scenario for making marginal final glide
decisions is on a short task where a pilot is climbing
slowly trying to make it up to final glide altitude.
The slow climb takes up lots of minutes per foot gained
and every minute drags down your speed relatively more
on shorter tasks.

So, say you are climbing at 2 knots. It will cost you
about 4 points for every hundred feet you climb, or
about 40 points to go from a white-knuckle 2-knot glide
to 0' at the finish up to a 2-knot glide to a 1000'
AGL arrival. For a 4-hour task the 4 points per 1000'
drops to 2 points per 1000'.

You could imagine a penalty structure that looks something
like: 8 points per hundred feet divided by the minimum
task time (or the winners time, or your time). This
eliminates most of the incentive to cut a last thermal
short since it is in the pilot's interest to keep climbing
if he thinks there is any chance he will be under the
minimum finish height and he is achieveing a climb
rate of 2 knots or more.

If you're climbing in your final thermal at less than
2 knots you are looking at a dicey glide no matter
what, and probably are contemplating a rolling finish.
It's not clear to me that a penalty structure built
around slower than 2 knot climb rate would do any good
- plus the penalties start to get kind of large (e.g.
16 points per 100 feet if you pick 1 knot as the climb
rate).

With the penalty structure I've described, if you finish
at 500' below the minimum finish height (so you are
at most 500' AGL) and actually fly to more or less
a full pattern it would take about 2.5-3 minutes to
get from the edge of a 1-mile cylinder to a full stop.
This is based on looking at a couple of my contest
finishes at Parowan where the runway is pretty long
and they were asking us to roll all the way to the
end. Guess what? The penalty as described above would
work out to the equivalent of an additional 2.5 minutes,
so the worst case scenario for a low flying finish,
would be no worse than taking the time to landing and
stopping. If you just barely miss the minimum height
you are a lot better off.

In terms of coming to a screeching halt in the middle
of the runway on a rolling finish - it's worth 2-5
points in my estimation. You need to weigh that against
all the other safety considerations and potential penalties
that might be imposed of you were really ver-the-top
about it. Plus the ill-will from your crew when they
have to schlep your glider halfway across the airport.

9B

At 20:42 28 July 2007, Andy Blackburn wrote:
At 01:48 28 July 2007, Marc Ramsey wrote:
BB wrote:
In my opinion the clock should stop as soon as the
pilot enters the
cylinder. We shouldn't have pilots in the finish cylinder
still
racing.


You haven't met enough contest pilots. If the rules
change in this
way, pilots will aim to finish one mile out, 50 feet,
90 knots and
then float in to the landing, european-style.

Applying an appropriate penalty for finishing below
the Minimum Finish
Height would eliminate that behavior. I was actually
surprised to find
that the SSA competition rules provide no guidelines
as to how to
penalize pilots who don't make it into the finish cylinder.
Given the
difficulties of knowing precisely how high one is finishing,
missing by
50 or so feet shouldn't result in a huge penalty, but
it should also
never be beneficial to intentionally finish low...

Marc

DISCLAIMER: I understand under the new rules speed
points are no longer allocated pro-rata so as to create
bit more spread at the top of the scoresheet. My math
may, therefore, be a bit off.

The worst case scenario for making marginal final glide
decisions is on a short task where a pilot is climbing
slowly trying to make it up to final glide altitude.
The slow climb takes up lots of minutes per foot gained
and every minute drags down your speed relatively more
on shorter tasks.

So, say you are climbing at 2 knots. It will cost you
about 4 points for every hundred feet you climb, or
about 40 points to go from a white-knuckle 2-knot glide
to 0' at the finish up to a 2-knot glide to a 1000'
AGL arrival. For a 4-hour task the 4 points per 1000'
drops to 2 points per 1000'.

You could imagine a penalty structure that looks something
like: 8 points per hundred feet divided by the minimum
task time (or the winners time, or your time). This
eliminates most of the incentive to cut a last thermal
short since it is in the pilot's interest to keep climbing
if he thinks there is any chance he will be under the
minimum finish height and he is achieveing a climb
rate of 2 knots or more.

If you're climbing in your final thermal at less than
2 knots you are looking at a dicey glide no matter
what, and probably are contemplating a rolling finish.
It's not clear to me that a penalty structure built
around slower than 2 knot climb rate would do any good
- plus the penalties start to get kind of large (e.g.
16 points per 100 feet if you pick 1 knot as the climb
rate).

With the penalty structure I've described, if you finish
at 500' below the minimum finish height (so you are
at most 500' AGL) and actually fly to more or less
a full pattern it would take about 2.5-3 minutes to
get from the edge of a 1-mile cylinder to a full stop.
This is based on looking at a couple of my contest
finishes at Parowan where the runway is pretty long
and they were asking us to roll all the way to the
end. Guess what? The penalty as described above would
work out to the equivalent of an additional 2.5 minutes,
so the worst case scenario for a low flying finish,
would be no worse than taking the time to landing and
stopping. If you just barely miss the minimum height
you are a lot better off.

In terms of coming to a screeching halt in the middle
of the runway on a rolling finish - it's worth 2-5
points in my estimation. You need to weigh that against
all the other safety considerations and potential penalties
that might be imposed of you were really ver-the-top
about it. Plus the ill-will from your crew when they
have to schlep your glider halfway across the airport.

9B

Typo:

For a 4-hour task the 4 points per 1000' drops to 2
points per 1000'.

Should read:

For a 4-hour task the 4 points per 100' drops to 2
points per 100'.

9B

In the good old days, there was a simple penalty structure. Finish one
foot low (one foot under the top of the airfiled fence) and you've
landed out for the day, hopefully in the last field before the airport
and not on the fence itself.

It seems a lot simpler to simply move the good old days up 500 feet.
Back then, each pilot set his own safety margin, and didn't expect
anyone except his beer buddies to listen to "but I was only 10 feet
below the finish height".

If we go to fancy altitude-based penalties like JJ wants, or Andy's
carefully figured penalties, then the race goes to guys like me and
Andy who are willing to spend all winter figuring out the scoring
formulas and how to game them. Even the current rolling finish plus x
minutes leads to some fine calculations about thermal strength, chance
of porpoising, and so on to optimally take advantage of the rolling
finish option. Fine with me in a way, I need all the help I can get
and I'm good at math. But simplicity also has its virtues.

John Cochrane

Good post Andy, but I believe we need a set penalty
to discourage deliberately doing a rolling finish on
a good day. I watched a well known pilot make a rolling
finish every day for 5 days in a row (1000 feet and
2 mile finish cylinder). I have recommended the rules
committee consider the following:

up to 100 feet low = 5 point penalty
up to 200 feet low = 10 point penalty
up to 300 feet low = 15 point penalty
up to 400 feet low = 20 point penalty
rolling finish = 25 point penalty
JJ

Should read:

For a 4-hour task the 4 points per 100' drops to 2
points per 100'.

9B

On Jul 27, 7:14 pm, BB wrote:
In my opinion the clock should stop as soon as the pilot enters the
cylinder. We shouldn't have pilots in the finish cylinder still
racing.

You haven't met enough contest pilots. If the rules change in this
way, pilots will aim to finish one mile out, 50 feet, 90 knots and
then float in to the landing, european-style.

If you don't think people racing inside the cylinder is a good idea,
then what you want is a "hard floor". 499 feet = distance points only.
Now, again, everybody inside 1 mile is done racing, but pilots aim for
500 feet, not for 50 feet.

John Cochrane

That is easily fixed with a penalty that should be in place of
finishing low. The way I have seen it done is to apply a 1 to 2 minute
penalty for each hundred feed low. The idea is to make it more
beneficial to climb than to finish low.

Brian

Lots of good points made.

I went flying yesterday and found that my SN10 has a beatiful digital
pressure altimeter readout, that is automatically calibrated before
takeoff to field elevation, and can be reset inflight to the latest
altimeter setting if desired.

I also found that my mechanical POS alitmeter lags about 100' during a
final glide, showing me that much higher that the SN10's no-friction
digital readout.

Guess what I'll be using from now on!

Back to the original subject (actually a spin off):

I still think the current hard cutoff at 500 ft is a poor setup, due
to the difficuty for the pilot to accurately judge his altitude at the
time of crossing the line. If the goal is to make pilots finish
higher (for whatever reason), then there needs to be a finish window
the pilot can aim for that if he accurately figures his final glide,
will not be penalized. Let's assume we can hit a 200' window - and
assume that 300' agl is the cutoff for a safe pattern. Setup the
scoring so anywhere in the 200 ft window (300'agl to 500'agl ) is
neutral - if below the nominal 500', then add the time it would have
taken to climb in (based on the climb rate in the last thermal). That
would remove any incentive to finish lower than 500', but give a
reasonable window to shoot for before a bigger penalty (automatic
rolling finish score) kicks in.

Comment? Obvious problems?

Kirk
66

At 22:18 29 July 2007, Kirk.Stant wrote:
Lots of good points made.

I went flying yesterday and found that my SN10 has
a beatiful digital
pressure altimeter readout, that is automatically calibrated
before
takeoff to field elevation, and can be reset inflight
to the latest
altimeter setting if desired.

I also found that my mechanical POS alitmeter lags
about 100' during a
final glide, showing me that much higher that the SN10's
no-friction
digital readout.

Guess what I'll be using from now on!

Back to the original subject (actually a spin off):

I still think the current hard cutoff at 500 ft is
a poor setup, due
to the difficuty for the pilot to accurately judge
his altitude at the
time of crossing the line. If the goal is to make
pilots finish
higher (for whatever reason), then there needs to be
a finish window
the pilot can aim for that if he accurately figures
his final glide,
will not be penalized. Let's assume we can hit a 200'
window - and
assume that 300' agl is the cutoff for a safe pattern.
Setup the
scoring so anywhere in the 200 ft window (300'agl to
500'agl ) is
neutral - if below the nominal 500', then add the time
it would have
taken to climb in (based on the climb rate in the last
thermal). That
would remove any incentive to finish lower than 500',
but give a
reasonable window to shoot for before a bigger penalty
(automatic
rolling finish score) kicks in.

Comment? Obvious problems?

Kirk
66



Hey Kirk,

I think if you have a 'zero penalty' band pilots will
tend to use it. I can't figure the difference between
and 700' finish with a 200' band and a 500' finish.

If you are going to try to ease up on the current 'all-or-nothing'
system adding a continuous penalty equal to some low,
but not minuscule, rate of climb. I think 30 to 60
seconds per 100' is reasonable. This would amount to
10-20 points on a long task and 20-40 points on a short
task if you finished 500' low - you could set 500'
under as the maximum penalty, or let it scale all the
way to worm-burner finishes (at a mile out!). The maximum
penalty could also apply to rolling finishes, or just
let the penalty for your actual finish height apply
irrespective of whether you roll or do a pattern. After
all, it's the finish height, not the shape of your
pattern that matters.

Or we could leave it to the CD's discretion. Then pilots
might try a little harder to not miss the finish height.

9B

kirk.stant wrote:
I still think the current hard cutoff at 500 ft is a poor setup, due
to the difficuty for the pilot to accurately judge his altitude at the
time of crossing the line. If the goal is to make pilots finish
higher (for whatever reason), then there needs to be a finish window
the pilot can aim for that if he accurately figures his final glide,
will not be penalized. Let's assume we can hit a 200' window - and
assume that 300' agl is the cutoff for a safe pattern. Setup the
scoring so anywhere in the 200 ft window (300'agl to 500'agl ) is
neutral - if below the nominal 500', then add the time it would have
taken to climb in (based on the climb rate in the last thermal). That
would remove any incentive to finish lower than 500', but give a
reasonable window to shoot for before a bigger penalty (automatic
rolling finish score) kicks in.

Comment? Obvious problems?

I'd suggest the opposite. I think I should be rewarded for every foot
that I have over the minimum finish height of, say, 500' AGL. So, if I
finish at 2000' AGL, I should get the actual time I spent climbing the
last 1500' deducted from my task time. It's more accurate, and it
favors my chosen strategy, what's not to like? 8^)

In reality, any halfway decent glide computer, or software with access
to pressure altitude, will prior to takeoff either automatically
determine the field elevation or let the pilot set it manually. The
same sort of problem exists with the start cylinder if one can climb to
the top. The glide software I use (which I wrote) automatically
determines field elevation just prior to takeoff. It monitors my
altitude in the start cylinder, signals me if I climb through the top
and does a countdown when I reenter, provides progressive warnings as I
approach the hard altitude limit (usually 17500' MSL out here), and
automatically adjusts my arrival altitude based on the minimum finish
height, all based on that initial field elevation measurement. I'm
confident that this will work within a margin of 10 or 20 feet, as it
using the pressure altitude that will ultimately show up in the IGC
file, and I don't have to pay much attention to any of it.

The SN10 also does a pretty good job at this (mine is better, of
course), what's the issue?

Marc