by **Sascha** » Wed Jul 07, 2010 12:05 pm

Ok, the idea was something like that:

The current "final approach" leaderboard will be renamed in "landing".

Your landing scores will always go into this leaderboard, regardless of whether it was a full or a final approach.

The current "full approach" leaderboard will be replaced with something new. As mentioned above, the pure landing scores already go into the "landing" leaderboard even for a full approach. So there are two possibilities:

1. A single "full approach" leaderboard, with the sum of the landing and the approach scores

2. Two separate "full approach" leaderboards, one with just the approach scores (glideslope and groundtrack error) and one with the combined approach/landing score.

I see that 2) might be a bit confusing, so I think that I'll follow your suggestion...

About the scores:

* For the final approach I won't change much, but I'll count the "max distance to centerline" only between main- and nose-gear touchdown, to not punish pilots who wait until wheelsstop (which theoretically can make your score worse when you don't stay on centerline) - unless of course you roll off the runway, in which case the score is reset to 0.

* If the landing was "almost" OK (e.g. just slightly too high vertical speed, derotate too early, slightly too much weight on wheels, etc.) you'll still get a score (let's say between 0 and 100,000 points).

* The theoretical maximum for the landing score will still be 1,000,000 points.

* The maximum score for approach will also be 1,000,000 points, so the combined max. score would be 2,000,000 points.

* No points (not even for a perfect approach) if you crash at landing

* You won't get an approach score if the HAC turn angle is less than 180 degrees (in this case only the landing counts, just like in a final-approach scenario).

To compute the score for the approach I've had the following idea:

* At a fixed rate (e.g. 6 times per second) the distance to the glideslope and the optimal ground-track is computed.

* These numbers are squared

* After touchdown, the average of the error-squared is computed to derive the score.

This means that e.g. flying 100 ft off the glideslope for 16 seconds is equally "bad" as flying 400 ft off the glideslope for 1 second.

I think the square makes sense: If you're reasonably close to the glideslope, you won't get many "bad" points.

There should also be some kind of weight, depending on the distance to the runway (closer too the runway, 200ft off glideslope is much worse then 1000ft off glideslope when you're at 50,000 ft).

What do you think?