| Author |
 Topic  |
|
aeajr
477 Posts |
 Posted - 10/13/2010 : 12:53:10 PM
|
The official minutes will reflect the details but it looks like we are going to have a lot more dates open on the 2011 calendar. HL and TD can now run independent calendars.
The HL geography now covers all states East of the Mississippi.
So get the word out so we can see if there is interest in more contests for 2011.
|
Best regards,
Ed Anderson
Long Island Silent Flyers
|
 |
|
|
josh_glaab
50 Posts |
Posted - 10/13/2010 : 10:33:26 PM
|
Constant-Delta Normalized Landing Method
There are two main methods for scoring landings. One tacks-on the landing scores after the flight times are normalized and the other adds the flight times to the landing points before normalizing scores. The benefit of the non-normalized landing method is that landings are always worth the same round score no matter how long the flight was. But there are two drawbacks to the non-normalized landing method. One drawback is that the landings do not account for large changes in landing conditions, due to ground-thermals, or wind-changes, etc. While it is true that all have to land a landing each round, that round could span 60 minutes and conditions can change significantly during that time. We normalize flight times to accommodate for changes in thermal activity (due to changes in the atmosphere), it would seem reasonable to do the same for landings.
Another drawback of the non-normalized landings is that you have to hit a landing to maximize your round score. No matter how many minutes a flyer may beat the rest of his flight group, if the winning pilot does not score well in the landing circle, his score will not be near the max round score. Another scenario is when nobody in a group can get back to the landing due to scratching out time in poor conditions. Even though the pilot that flew the longest, and landed on the field, was the best pilot for that round, his round score would still be penalized by not having any landing points. As a result, he will lose points to other pilots that do hit the landing in other flight groups that have good thermal conditions and easy landing approaches.
Non-normalized landings also preclude a scenario where a pilot could decide to continue to work very light lift, and sacrifice a landing attempt, to have a maximum round score. In this sense, a pilot could out-soar the competition and overcome other pilot's landing points. This would be more in-line with a "soaring" event and make it less of a "landing" contest.
One drawback of the normalized landing task is that the landings have a variable effect on your round score. Note that your round score is what results from whatever normalization/addition process is being used for each round. For example, let's consider two pilots. Pilot A gets a perfect 10:00 flight and a 99 point measured landing (out of 100 points). Pilot B also gets a 10:00 flight, but completely misses his landing. Note here that the term "measured landing" is what the pilot reads on the spot landing tape. The result would be that Pilot A gets a 1000 point (max score) round, and Pilot B would get an 858 round. The difference between the two round scores would be 142 points! That is 42 points more than a 100 point customary non-normalized landing score. In the non-normalized scoring method, Pilot A would get a 1099 and Pilot B would get a 1000. It gets even more severe for short flights. For this example, consider a 5-minute flight time for both Pilots. Pilot A would still get a 1,000 point score, but Pilot B would get a 752. The landing that Pilot A made would be worth a 248 point delta compared to Pilot B!
What is required is a landing task that would provide an order of magnitude improvement in the time-frame used for comparing landings. As stated previously, the current time-frame used to consider all landing conditions constant is the length of time that the round takes to complete (app 60 minutes). As a result, this requirement specifies that the prevailing landing conditions be reflected in the landing scores at least every 6 minutes, which would account for frontal-type wind-shifts. Two orders of magnitude would be desirable (every .6 minutes) to attempt to account for ground thermals. In addition, another requirement is to enable a pilot to maximize his round score while not getting ANY landing points and out-soar the competition.
My proposal would change the normalized landing tasks to effectively replicate a non-normalized landing result for pilots who have identical flight times. It accomplishes this by scaling the landing score by the flight time by a specified amount. Pilots would still read the landing tape (measured landing), as they do now and report that score, but the resulting landing points would be calculated based on the flight time and the amount of round score that a CD wants to allocate for landings. The resulting landing points would then be added to the flight time for the normalization process to calculate round scores.
To explain this more, consider a case where a CD wants to have the maximum round scores due to landings be worth 100 round points. In this case, the difference in round scores for two pilots, one with a perfect landing and the other without any landing score, would be 100 round points. To accomplish this, the maximum landing score would be calculated by multiplying flight time by 6.666/minute (or .11111/sec). For a 1 minute flight, the landing would only be worth 6.6666 points. For a 10 minute flight, the landing would be worth 66.666 points. For another example, a 5 minute flight would subsequently have a max landing score of 33.333 points. The maximum landing score would then be multiplied by the ratio of the pilots’ measured landing to the maximum measured landing (ie an 80 pt landing would be 80% of a 100 pt max landing tape).
Consider the example pilots from above again. For a 10-minute flight, Pilot A would get 600 flight points, then he would get 66.666*(99/100)=65.999 landing points. His combined flight and landing score would be 665.999. Pilot B would get a combined flight and landing score of 600. The normalized round scores would be 1000 for Pilot A, and 900.90 for Pilot B. If Pilot A got a 100-pt measured landing, then Pilot B would get a 900 point round score. This preserves the current 100 point delta provided by the typical 100-pt non-normalized landing method. Proceeding further to consider the 5-minute flight example, Pilot A would get 300 flight points and only 5*6.666*99/100=32.9997 landing points for a combined score of 332.9997. Pilot B would get a combined flight and landing score of 300 (recall Pilot B did not hit the landing tape at all). The resulting round scores would be 1000 for Pilot A and 900.90 for Pilot B. Again, the maximum round score delta due to the landing would be 100 points.
Now consider a situation where Pilot A gets 8 minutes with a 90 point measured landing. In this situation, Pilot A would get 8*6.6666*90/100 = 47.9995 points. Pilot B could maximize his round score by flying for 8 minutes PLUS another 47.9995 (48) seconds. This would allow pilots to win the round AND maximize their round scores without getting any land points at all! Consider another situation, Pilot A begins his landing approach at the target time and is confronted by a severe ground thermal. Pilot A manages a 4 point (out of 100) measure landing. Everyone else in Pilot A's flight group miss the landing completely due to the turbulent landing conditions. Pilot A would still max his round score. If all of the Pilots in Pilot A's flight group got perfect 10 minute flights, then Pilot A would get a 1000. All the other pilots would 995.57 point round scores.
For another scoring example, consider a case where both pilots get the same landing score, but significantly different flight times. Pilot A gets a 8 minute flight with a 50 pt measured landing, Pilot B gets a 10 minute flight, also with a 50 pt measured landing. Pilot A would get 480 flight points plus 6.6666*8*(50/100)=26.666 landing points for a total of 506.666 points. Pilot B would get 600 flight points plus 6.6666*10*(50/100)=33.333 landing points for a total of 633.333 points. Pilot A's round score would be 799.999. Pilot B's round score would be 1000. In the non-normalized landing method, Pilot A's round score would be 850, Pilot B's round score would be 1050. It is interesting that the round score delta between the two flights is 200 points for both scoring methods.
Going further, one could consider the landing tape itself. Contest Directors (CDs) could make a landing tape with 50 increments and vary the size of the increments. If you only have 50 increments, the landing score would be computed slightly differently than above. The measure landing score (let's say 48 out of 50 points for a 10 minute flight) would be 10*6.6666*48/50=63.999 points for a 10-minute flight. CD's could also elect to have landings only worth 50 round score points. In this case, instead of multiplying the flight times by 6.6666 to get the max landing score, 3.3333 would be used instead. Varying the length of the increments could be used to adjust the complexion of the contest. Larger increments would favor precise touch-down times. Shorter increments, would favor landing precision.
One minor drawback to this method is that the resulting round score would not quite be based on 1 point per second flight time due to the effect flight time has on the landing score. However, considering a 10 minute task, a pilot would have to be more than 6 seconds off of his flight time to see his landing decreased by 1%.
For those of you who have your heads spinning with all this math, please relax and take deep breath. We haven’t even broke out the calculus yet (just kidding)! The pilots’ objective is still the same as with a non-normalized landing: max the flight, get the landing. But, keep an eye on the competition, if all are down (or almost down) you could maximize your round score without any landing points by flying longer. Since landings are scaled by flight time, the time needed to out-soar your competitors landing is less for shorter flights. It is also true that you don’t have to really stress over a landing for a short flight since again, the landings are scaled with flight time. If a CD calls for a small delta in round scores due to landings (say 50 points) the emphasis will be on precise timing and less on precision landings. If a sudden-wind shift/ground-thermal hits the group on approach and all have bad landings, recall that the landings are normalized and that your round score will be Ok and you are on an equal footing with those who have great landing conditions.
Before the advent of computers being used for scoring, this method would simply not be possible. Now, however, even the most minimum computer would do the calculations in microseconds. To implement effectively, however, the resulting round score sheet needs to be updated to show flight time, measured landing, calculated landing score, combined flight score, and resulting normalized round score to help all get familiar with the system.
The proposal defined herein satisfies the requirements previously stated to account for significant changes in landing conditions and moves the competition away from being a "landing contest" to be more of a "soaring event". I think this method is worth strong consideration as well as test drive, and I know just the contest to do that (HRSF/BRASS on 5/14 and 5/15/2011)!
Please send me your thoughts and comments!
|
Edited by - josh_glaab on 10/28/2010 9:19:42 PM |
|
|
|
ljb0001
37 Posts |
Posted - 10/14/2010 : 12:58:18 PM
|
That is sheer genius, Josh! I too prefer the elimination of the landing lottery but have struggled with the issue of the variable and very significant effect of the landing on the round score.
Accommodating your algorythm in the scoring software should be fairly straight forward. We could add the appropriate parameters to the contest set up screen and provide another scoring options for landings. Right now we can choose to include or exclude the landing points in the normalized score. We can add a third option which normalizes the weighted landing you are proposing.
I need to read your proposal again and wrap my head around your math but I see it as a very manageable solution to the problems associated with normalizing landings.
We should be able to support this new scoring method by the time the HRSS/BRASS contest comes around and we can give it a try there though I don't see any real issues with it.
Luis |
|
|
|
paul
13 Posts |
Posted - 10/15/2010 : 11:03:07 AM
|
If I understand Josh's proposal of 10/13, he would take 2/3 of the landing score and then perform an additional adjustment to that value, which is a ratio of the flight time over the task time and then add that value to the flight time. My interpretation of this is that my landing score will be penalized if i have a short flight.
In his first example, he shows that using his method yields the same relative difference in the scores. In another example he showed that pilot B could win the round without making a landing at all. This is the way things are now.
I believe that to be a good competitor you must understand the scoring system. This system seems a little complicated. I know that Josh is a great flier and a smart cookie to boot, so he can handle it. But i think that many of our competitors would have a hard time understanding the system and then trying figure out what they should be doing to maximize their scores. Isn't this the same problem that we have always had? Fly the task and hit the landing. It seems that Josh's proposal is aimed at addressing variability of the conditions in the landing circle. Maybe a simple solution would be to minimize the value of the landings. Instead of making a 100 point maximum, how about a 25 point max, or some other less influential value? Then we would be emphasizing soaring and de-emphasizing landings.
Anyway, my flying abilities have gone down the drain the last couple of seasons, so my comments are kind of academic.
I am going to apply for my entitlement next week. Paul Bell
|
|
|
|
soarkraut
24 Posts |
Posted - 10/20/2010 : 11:03:16 AM
|
Sounds like the proposed landing computation will require a lot of trust in the scoring crew. There would not be an easy and quick way to keep track of your flight and landing scores. Unless you had a similar program on you Blackberry. I don't want to get that intense about my weekends, at this time.
If the ESL insists on a precision landing task.........( I would prefer a more realistic, no lawn dart, landing)
A "risk and reward" landing may well be a lot more interesting. The one I have in mind would be an easy set-up with the Landing tapes we have now:
Using a semi circle,180 Deg., instead of the full circle. Draw a line between all landing tape centers(With a string). The semi circle away from the pilot would be measured as always. Anything past the line, on the pilot side, would be a zero.
Just my 2 cents......
Pete |
|
|
|
Topic |
|
|
|
ESL - Where to take the league
Forum Locked
