Is the 200ft below Min Finish Height Rule Working?

John,

I was included on an email discussion thread which included the email which I posted last night. I thought it was important and decided to post here and invite broader discussion. I did not have "permission." I think the ID of the author is unimportant. I take responsibility for the post but also want to make sure that everyone knows that the (in my opinion excellent) email itself was not written by me. I cannot take credit for that.

The intent of posting is absolutely NOT to call the RC into question. I think the RC is doing an excellent job of managing the various "special interests" and trying to address safety issues and manage the rules with an eye towards fun, growth & attendance. I have deep respect for the job the RC does and has done even though I have clearly complained about some rules from time to time.

I agree, the 700 margin that is currently provided by the rule is probably to low. Here is my logic...

I think there are two basic options: 1) raise the height 2) lower the penalty (option 3 is to remove the rule entirely...but I am deeply concerned about that idea)

The problem with lowering the penalty is that the average contest pilot is (whether he/she admits it or not) going to fight to avoid a 25 points penalty, let alone 400. Getting the penalty formula "just right" so that the penalty is high enough to produce the desired behavior (encouraging the contest pilot to build in a greater risk buffer height before embarking on final glide) yet low enough to encourage the pilot to safely finish straight ahead if final glide degrades into the penalty zone is going to be VERY DIFFICULT for the RC and US contest pilots to "negotiate". ;-)

I think some pilots would circle outside of the finish cylinder even if the finish penalty altitude was 200 ft and the penalty was 5 points. Its just the nature of competition and the fact that most of us get away with it most of the time. What's one circle going to hurt? Was that a bump? Etc, etc.

Therefore we need enough margin to allow for safe circling outside the finish cylinder for this rule to not have "side effects" which are creating unintended risks. If there is a penalty, pilots are going to fight to avoid it. It's just that simple.

Sean


On Monday, January 20, 2014 11:24:53 AM UTC-5, wrote:
If you read the actual original post, rather than just say "finishes again, let's blast the durn rules committee" it is quite interesting.



It documents pilots doing crazy things -- thermaling at low altitudes -- in return for a few points to get over the finish height.





I am often told, "pilots can make their own decisions, they won't do stupid things just because a few points are on the table." This post and associated data (if we see it) document the opposite.



So, if you think about it, these observations make a strong case for raising the height. OK, if when "close" they're going to do nutty things, we had better move the ground down another 500 feet, so with a finish at 1000 feet, even these dumbbells will have a cushion. If anything, these observations call for a hard deck, or at least a hard deck in the last 5 miles, to remove the temptation these pilots are obviously falling prey to, to do silly things.



It is mighty, mighty hard to go from these observations to the conclusion that moving everything down 500 feet, to putting the same cliff in points at 1 inch above the barbed wire fence at the edge of the airport, rather than 500 feet over the ground, makes it more safe. Then the same pilots thermaling at 550 feet, 1 mile from finish will thermal at 50 feet, 1 mile from finish. Like they did in the good old days, producing the good old days accident reports.



A minor ethical quibble. I saw this post by its original author, who asked me for comment, which I did, privately. Sean, did the author give you permission to pass it on to RAS, anonymously?



John Cochrane

Why not use the total height?

So height (AGL) plus potential height (speed²/(2*a)) That makes ballistic pull-ups useless, allows, actually favors smooth finishes.

Then set the total height rather high and substract one point per feet too low.

On Monday, January 20, 2014 3:03:41 PM UTC-5, J. Nieuwenhuize wrote:
Why not use the total height? So height (AGL) plus potential height (speed²/(2*a)) That makes ballistic pull-ups useless, allows, actually favors smooth finishes. Then set the total height rather high and substract one point per feet too low.

Simply put- because this becomes a pilot and scoring nightmare. Note that each glider converts kinetic energy to potential energy differently.
UH

Rules can't fix stupid. For each reiteration of the rule set we pilots
will find a way to do something stupid with them, and so we spiral down
the rabbit hole of never ending rule changes that become more complex
every year. In my opinion the rules should be as simple as possible.
They should not promote dangerous behaviour, but their primary objective
is to provide a fair way to determine the best pilot. The simpler the
rules the more time the pilot has to worry about flying the aircraft.


Luke Szczepaniak

On Monday, January 20, 2014 2:21:11 PM UTC-6, wrote:
On Monday, January 20, 2014 3:03:41 PM UTC-5, J. Nieuwenhuize wrote:

Why not use the total height? So height (AGL) plus potential height (speed²/(2*a)) That makes ballistic pull-ups useless, allows, actually favors smooth finishes. Then set the total height rather high and substract one point per feet too low.



Simply put- because this becomes a pilot and scoring nightmare. Note that each glider converts kinetic energy to potential energy differently.

UH

UH answer is deep -- pay attention.

RC gets two constant demands. One, as in the original post, is to add carefully constructed point carpentry around the finish, with 10 points for this and 20 points for that. The other is to simplify the rules, and especially to make sure pilots don't need to do lots of strategizing and in-air calculations. As you come up with alternatives, make sure they satisfy simplicity and clarity too!

John Cochrane

On Monday, January 20, 2014 1:39:11 PM UTC-8, wrote:

RC gets two constant demands. One, as in the original post, is to add carefully constructed point carpentry around the finish, with 10 points for this and 20 points for that. The other is to simplify the rules, and especially to make sure pilots don't need to do lots of strategizing and in-air calculations. As you come up with alternatives, make sure they satisfy simplicity and clarity too!

John Cochrane

This particular issue has received scores of hours of thought, debate and analysis, including just about every conceivable scenario from the beginning of a final glide (below, at, and above optimal Mc final glide OR best L/D glide to the finish) and every major decision scenario as glides go bad (or don't get better - including able to make the cylinder but not the airport, vice-versa and under different penalty structures) ALSO various lift scenarios (none, less than current Mc, and climb rates all they way down to climbs so slow you are losing speed points faster than you lose penalty points). Zoomies at the edge, terrain in the last 10, 5 and 1 miles to the airport, proximity to a ridge, number of runways, configuration of the approach versus the finish, trees at the end of the runway and number of competitors trying to land at once with how much energy. It ALL gets assessed and debated, including the bizarre potential choices pilots might make (although pilots can be very creative in coming up with bizarre things - the analysis did include low circling to get up to finish height even right up to the edge of the cylinder). Then what gets discussed is which are likely versus unlikely scenarios and which ones are pilot decision issues versus places where the rules beg the pilot to take a chance in order to score more points. LASTLY it all gets put into the filter of don't change anything and make it simple (against the tide of requests for specific exceptions to handle odd cases).

In this case the higher order issues boiled down to: 1) The rules should not be set up to award points to pilots who cross the finish cylinder at an altitude from which it is unlikely that (s)he can safely reach the airport (including scenarios with the runway not lined up, into the wind and with trees), 2) Assess a modest penalty for most common glide gone wrong errors, such that a pilot would not ignore a reasonable-looking climb along the way on a marginal glide to MFH.

If you do the math what you find is if you allow (as we do) different finish heights and different sized finish circles you can end up with not much room between the bottom of the mild penalty and "can't get to the airport" height. A penalty structure that varies the steepness of the penalty depending on the cylinder radius and MFH is possible, but complex and was set aside as was restricting the finish height to 1000' or above as some sites with ridges like the flexibility to finish right off the ridge.

It's mostly documented in the RC notes. I'd be happy to take anyone through the "all the scenarios" analysis offline - there's a lot to think about before you boil it down to a set of simple rules and it's easy to fix one thing while braking something else.

9B

On Monday, January 20, 2014 6:37:18 PM UTC-5, wrote:
On Monday, January 20, 2014 1:39:11 PM UTC-8, wrote:



RC gets two constant demands. One, as in the original post, is to add carefully constructed point carpentry around the finish, with 10 points for this and 20 points for that. The other is to simplify the rules, and especially to make sure pilots don't need to do lots of strategizing and in-air calculations. As you come up with alternatives, make sure they satisfy simplicity and clarity too!



John Cochrane



This particular issue has received scores of hours of thought, debate and analysis, including just about every conceivable scenario from the beginning of a final glide (below, at, and above optimal Mc final glide OR best L/D glide to the finish) and every major decision scenario as glides go bad (or don't get better - including able to make the cylinder but not the airport, vice-versa and under different penalty structures) ALSO various lift scenarios (none, less than current Mc, and climb rates all they way down to climbs so slow you are losing speed points faster than you lose penalty points). Zoomies at the edge, terrain in the last 10, 5 and 1 miles to the airport, proximity to a ridge, number of runways, configuration of the approach versus the finish, trees at the end of the runway and number of competitors trying to land at once with how much energy. It ALL gets assessed and debated, including the bizarre potential choices pilots might make (although pilots can be very creative in coming up with bizarre things - the analysis did include low circling to get up to finish height even right up to the edge of the cylinder). Then what gets discussed is which are likely versus unlikely scenarios and which ones are pilot decision issues versus places where the rules beg the pilot to take a chance in order to score more points. LASTLY it all gets put into the filter of don't change anything and make it simple (against the tide of requests for specific exceptions to handle odd cases).



In this case the higher order issues boiled down to: 1) The rules should not be set up to award points to pilots who cross the finish cylinder at an altitude from which it is unlikely that (s)he can safely reach the airport (including scenarios with the runway not lined up, into the wind and with trees), 2) Assess a modest penalty for most common glide gone wrong errors, such that a pilot would not ignore a reasonable-looking climb along the way on a marginal glide to MFH.



If you do the math what you find is if you allow (as we do) different finish heights and different sized finish circles you can end up with not much room between the bottom of the mild penalty and "can't get to the airport" height. A penalty structure that varies the steepness of the penalty depending on the cylinder radius and MFH is possible, but complex and was set aside as was restricting the finish height to 1000' or above as some sites with ridges like the flexibility to finish right off the ridge.



It's mostly documented in the RC notes. I'd be happy to take anyone through the "all the scenarios" analysis offline - there's a lot to think about before you boil it down to a set of simple rules and it's easy to fix one thing while braking something else.



9B

Clearly there was a big departure from gradual penalty to the land out penalty. Big change that moved the dangerous flying outside the finish cylinder.. Why the land out penalty. I don't think the land out penalty was well thought through.

The land out penalty should be rolled back and gradual penalty should come back if any.

On Monday, January 20, 2014 5:12:21 PM UTC-8, wrote:

Clearly there was a big departure from gradual penalty to the land out penalty. Big change that moved the dangerous flying outside the finish cylinder. Why the land out penalty. I don't think the land out penalty was well thought through.

The land out penalty should be rolled back and gradual penalty should come back if any.

It mostly boils down to whether you believe there should be a significant points benefit for making a finish at an altitude from which it is impossible to reach the airport. That's a powerful incentive to roll the dice. Should winning hinge on betting your glider (and your life) in exchange for points in that way?

The rest is just math - and how low you set MFH.

9B

On Monday, January 20, 2014 9:04:44 PM UTC-5, wrote:
On Monday, January 20, 2014 5:12:21 PM UTC-8, wrote:



Clearly there was a big departure from gradual penalty to the land out penalty. Big change that moved the dangerous flying outside the finish cylinder. Why the land out penalty. I don't think the land out penalty was well thought through.



The land out penalty should be rolled back and gradual penalty should come back if any.



It mostly boils down to whether you believe there should be a significant points benefit for making a finish at an altitude from which it is impossible to reach the airport. That's a powerful incentive to roll the dice. Should winning hinge on betting your glider (and your life) in exchange for points in that way?



The rest is just math - and how low you set MFH.



9B

You did not provide a logical reason why this big change (scoring as land out) was made rather then just raising the finish height and leaving gradual penalty.

You did not address any problem you just moved the problem somewhere else.

You did not provide a logical reason why this big change (scoring as land out) was made rather then just raising the finish height and leaving gradual penalty.

The point of scoring low finishes as a landout is real simple. When you're deciding "shall I land in the last good field or press on" at MacCready 0 plus 50 feet, it needs to be crystal clear that you will gain nothing by pressing on. This is not safety legislation -- points are off the table, make a good decision, points are the same either way. With a graduated penalty there is always some benefit to pressing on. And complexity. Didn't you guys want simple rules? Try figuring out the points to finish 397 feet low.

It just moves the hard ground down. You used to be scored as a landout -- with none of this mollycoddling graduated penalties -- if you missed the fence by a foot.

Don't think of it as a "penalty." The task is to start below (say) 5000', get inside three turnpoints, and finish no less than (say) 700'. If you didn't do that, you didn't fly the same race as everyone else. In what other sport can you miss the finish line by 200 feet and still get a "finish?" And want more?

John Cochrane