04 juni 2015

Scoring systeem

Ik kopieer een artikel van Davis Straub over de werking van GAP, ons scoringssysteem. Het maakt aardig duidelijk hoe het werkt:

Hang gliding (and paragliding) scoring is a mess.  It is a mix of distance and time points as well as a variety of other points that have various explanations for their inclusion.  Those explanations are often at odds with the underlining equations and algorithms that actually implement them.
Most pilots just ignore all this and just fly the competition and attempt to do their best.  A few pilots, including some top scoring pilots attempt to understand (or think that they do understand) the scoring systems so that they can maximize their scoring.
Our competition format is "Race to Goal." So why don't we just score the competition so that the pilot who gets to goal first wins the day?  If you don't get to goal, well, too bad, you just don't get any points.  This is a race to goal after all.  This would in fact be real points.
So why distance and time?  In hang gliding and paragliding it is often the case that pilots don't make it to goal.  In fact most of the time most of the pilots don't make it to goal.  One of the reasons is that the goals are often set at a distance that makes it difficult to get to. Only the most skilled pilots are able to make it to the goal.  With a wide spectrum of skills represented among the pilots it is difficult to call a task that actually makes it a race to a goal.  Meet directors (and pilots) feel that the task should be difficult to accomplish and long enough to represent a real challenge to all the pilots, including the best pilots.
Therefore, points are split between the pilots who do make it to goal, they get distance and time points, and pilots who don't make goal, they just get distance points.  The arbitrary formulas (created to replicate "good looking" curves) that determine this split, and determine the distribution of time points and distance points weigh toward more time points as more pilots make goal.
Of course, task committees can misunderstand what is a makeable goal and conditions can be such (or change to be such) that no pilots can make goal.  You don't necessarily wish to throw out a day just because no one can make it to the goal.
So time and distance are at least measureable for each pilot and they are real enough.  Maybe the points awarded for your distance and time are arbitrarily distributed, but you do know your actual speed and distance.
If you are going to have a race to goal, perhaps you should all start the race at the same time.  It doesn't make much sense to start later.  In paragliding competitions race starts (everyone starting at the same time) are the rule.  With hill launches everyone has the opportunity to get off the hill and up in the start cylinder, so that they can all start together.
Hang glider competitions, whether aerotowed or hill launch, most often have multiple start gates.  Pilots can start later than other pilots.  It's like they are in two or three or four different races.  Multiple start times are used because meet organizers can't be sure that they can get all pilots off and up in the start cylinder in time for a single start time.  Of course, if they extended the time between the launch window open and the start gate time this would at least be possible, if not guaranteed.  But, the trade off is that the first pilots to launch have to spend a lot of time in the air waiting for the start gate.
So, wait a minute, we are talking about devaluing our race to goal, just for the convenience of pilots.  Just because we don't want them to over exert themselves waiting for a start gate.  This came about when pilots weren't forced to launch (an ordered launch with an implied push) right when the launch window opened, and we used the promise of multiple start times (and a shorter time between the launch open and the first start time) to encourage pilots to get going.
With multiple start gates we run into the problem of one race interfering with another race.  The pilots that take a later start gate sometimes have the opportunity to use the pilots who have taken an earlier start gate to mark the thermals for the following pilots, thereby allowing them to fly faster than they would have otherwise.  This seems so unfair and an active discouragement to those pilots who took the earlier start time.  Our race to goal is now getting very confusing.
So "leading points" and "arrival position points" (or their equivalent) are included in our scoring system, and they are meant to reward pilots who get out in front and get to goal early.  They are meant to offset the advantage that the pilots behind them may have if they use the earlier pilots as thermal markers.
These "funny" points also encourage pilots to launch earlier (when there wasn't a forced launch) which helped the meet organizers get the task going.
So how are "leading points" calculated?  The documentation for the FS scoring program (the "official scoring program") is found here: http://www.fai.org/downloads/civl/SC7_GAP on page 33. Leading points are calculated using the area under the pilot's flight curve.  They are more accurately labeled as "flight curve area" points.
Top pilots (well, Jonny Durand, Dustin Martin, and Zac Majors) have advocated to me to be sure to use a scoring system that uses these "leading points or "flight curve area" points because they contend that they reward pilots who pull the leading gaggle.  What does the FS documentation say about this?
The documentation states:

Flying "in" the leading group (perhaps because they left at the first start time).
The documentation does not say that the "flight curve area" points reward a pilot who pulls the leading gaggle.  It is difficult for me after reading the documentation to make that determination.  Perhaps you can.  These pilots contend that they have seen that result.  They have not provided me with any documentation of that claim.

Geen opmerkingen:

Een reactie posten