Scoring Irregularities

Moderator: Comp Committee

Post Reply
User avatar
Chip
Site Admin
Posts: 645
Joined: Thu Apr 28, 2005 8:20 pm
Location: Sylmar, CA
Contact:

Scoring Irregularities

Post by Chip » Mon Jun 08, 2015 7:41 pm

After award presentations, I was part of a discussion about the scoring formulas that were used. Currently SeeYou is using OzGap 2005 where the scoring program by default DOES NOT use leading points, but utilizes arrival time points.

There is some serious debate on how that is implemented and which formula is more fair. The debate is pretty involved and take a lot of analysis, which I'm not prepared to discuss here. However, I do want to present the scores as computed by FS (the official FAI scoring program).

One of the results was completed with OzGap
The other result was completed with FS using GAP2015 parameters where leading and arrival position points are utilized.

Using OzGap 2005, the results basically mirror the results on SoaringSpot

Something else that was changed in this version of the program was that more detailed results were not being shown and the distribution of points. That is better served by FS.

So the links are above for your own analysis. Feel free to jump into the fray if you really are concerned. If you have an opinion, I'd like to hear it. Mainly I think leading points are needed for longer tasks, where pilots are camping on top of others and pimping, I don't really believe we have that issue here. It did appear to me that pilots that were in position to start, did just that.

User avatar
Chip
Site Admin
Posts: 645
Joined: Thu Apr 28, 2005 8:20 pm
Location: Sylmar, CA
Contact:

Post by Chip » Mon Jun 08, 2015 7:49 pm

When trying to enable leading points GAP 2000, SeeYou spits out errors, I notified the author (Davis).

User avatar
JD
Posts: 1696
Joined: Fri Apr 25, 2008 11:05 am

Post by JD » Mon Jun 08, 2015 9:56 pm

Yes Chip, I certainly agree that the results using FS with leading and arrival points is the fairer of all the methods. :wink:

User avatar
Chip
Site Admin
Posts: 645
Joined: Thu Apr 28, 2005 8:20 pm
Location: Sylmar, CA
Contact:

Post by Chip » Tue Jun 09, 2015 8:43 am

Then I'm not sure we are in agreement. I think leading points help encourage others to lead out. I don't think Rob needs that encouragement. However, the question is, should he be given points for doing so?

In this case, we had some pilots that did not really have an opportunity to get the 1st start. Some pilots didn't even launch until 1400, so should they be penalized?

User avatar
JD
Posts: 1696
Joined: Fri Apr 25, 2008 11:05 am

Post by JD » Tue Jun 09, 2015 10:16 am

Chip wrote:Then I'm not sure we are in agreement. I think leading points help encourage others to lead out. I don't think Rob needs that encouragement. However, the question is, should he be given points for doing so?

In this case, we had some pilots that did not really have an opportunity to get the 1st start. Some pilots didn't even launch until 1400, so should they be penalized?
I remarking from the limited context that by using FS with lead points I moved from 5th to 4th ahead of Ron. In this limited context, Ron had not only launched before me but also started the task at 2PM along with me then changed his mind and took the 220PM start.

This gave Ron the advantage over me of having any pilot who took the first gate on course marking lift. Furthermore, the original start was to be something like 130PM and not 2PM. I requested that we push it back to allow everyone time to get up to launch and for the conditions to fully develop. Rob agreed that 2PM would be a better start for everyone. So I'm not buying the late to launch argument. Nobody was making a mad dash to get up the hill other than Rob. I prodded everyone on Larry's track to get moving which we all did.

Now let's say that 10 minutes into my task I caught a boomer to 5K that was inside the exit cylinder and decided to restart at 220PM. Should I be allowed to exploit this advantage over any pilots who were stuck using their 2PM start without giving something in return? In this case shouldn't the pilots who committed to the 2PM start and were ahead of me marking the course be granted something to level the playing field?

Since we are flying in a confined area it's not difficult to see where a large number of pilots and thermals are located so once again pilots already on course are at a disadvantage over those using them as lift markers.

Since Rob was ahead of me even though we started on the same clock I was in fact using him to mark not only lift but poor conditions up ahead which I exploited to make decisions about whether to work the lift I was in or push on.

Based on the above information we should be using lead points unless pilots have no control over their launch times or are otherwise delayed beyond their control. Future events may have different circumstances and the scoring methodology should be announced in advance (if possible) so pilots may consider their strategy. I don't recall whether the scoring system was discussed in advance however I did make the assumption that we would be using lead points since there were multiple starts. I consider the two to go hand-in-hand. Our exit cylinder was huge with multiple house thermals equidistant from the first turn-point.

User avatar
Chip
Site Admin
Posts: 645
Joined: Thu Apr 28, 2005 8:20 pm
Location: Sylmar, CA
Contact:

Post by Chip » Wed Jun 10, 2015 10:33 am

I think pilots should help decide which scoring formula they want to use. In this case I was hampered by SeeYou because it only uses OzGap 2005 and Davis's scoring script.

If we want to continue to use SeeYou, we would need to create our own script in Pascal by the way.

Using FS gives us the most options in scoring, I like SeeYou for the analysis, so maybe that's just what we use it for. SoaringSpot is a nice feature, but requires SeeYou for the upload. I may just abandon it all together and use Airtribune or the SHGA forum for the Dahlsten Cup.

But at least I let you guys know about the issues and we can work them out for all future contests.

Here are the results scored in all three formulas. For everyone's education (if you care), CIVL/GAP 2015 is basically the same as CIVL 2014 except that LC is squared (meaning more leading coefficient points for those that take the first start). So the task scoring was alternatively accomplished in 2014, one with the squared value for LC and one without the squared value.

2014 Official CIVL Hang Gliding with squared leading coefficient value (GAP 2015)
2014 Official CIVL Hang Gliding without squared leading coefficient value (GAP 2014)
OzGap 2005

User avatar
JD
Posts: 1696
Joined: Fri Apr 25, 2008 11:05 am

Post by JD » Wed Jun 10, 2015 1:55 pm

I think the points spread using OzGap 2005 is the closest to what I would call a playing field leveler and help balance the decision making between starting now and waiting for a better start later or playing chase the pylons.

So let's talk about my handicapping points for using a 463-hour old obsolete sail and a draggy cocoon harness. :P :lol:

Post Reply