Gerolf Heinrichs <<email>> answers my questions regarding how altitude penalties work:

When we approached this altitude penalty rule of course we also debated the potential inaccuracies of the 3D-GPS altitude recording and how to counteract it.

What we do right now is, going by GPS-altitude, not barometric reading, because every pilot is able to produce this figure without having access to altering it. When switching on our 3D GPSes before take off we check the displayed altitude against the take off altitude. We set our barometric altimeters to take off height and we all note the respective differences between GPS height and actual height. That difference was found not more then about 20 meters so far. Ideally, at a later stage flight verification software would zero this difference, by simple shifting the GPS altitude to your take off height before taking off, but right now we don’t. We live with that error and the fact that this error generally seems not in the favor of the pilot as GPS heights tend to always read higher then the actual height.

The reason for setting the barometric height to take off height and not to the GPS reading is simply that you don’t want to screw up your final glide. It’s easier to remember the twenty something meters in your head and stay lower then the given altitude limit by that amount, plus a little margin in case you want to make sure to not even loose one single point.

As it turns out the procedure is rather workable, and not difficult to handle at all. Please bear in mind that until now we had to live with certain pilots often starting up too 200 meters higher in altitude depending on how they liked to see cloud base.

Now to your question why not policing altitude all along a pilot’s track log; why only at the one point, the start cylinder? The answer is feasibility. We simply haven't got the means here to check all data points on a track log – this would need to be programmed into the software and wasn’t available at this stage. If we wanted to police altitude limitation throughout a task, we would also need to come up with an appropriate penalty rule for it. Simply penalizing every one the same for having the same “too high reading” regardless of how long a pilot did so will not do. An appropriate penalty will probably have to be proportional to the altitude-vs-time area above that certain altitude limit of a given 3D-track log. Or as a simpler version of it, just counting the number of seconds one spends above the limit.

The idea to use the one start point at the start cylinder came from the Spanish pilots Francisco Vinas and Carlos Punet and was quickly accepted by the rest of the working group for its simplicity and practicality. It helps solve the problem of starting too high in an indirect but very elegant way.

Also the initial worries of pilots likely to speed up like sailplanes and pull up hard once passing the gate, did not materialize. Pilots came to quickly realize that is a lot easier to just fly along the cylinder line and simply swing into the start gate once time is up and altitude is sufficiently low.

Thinking back of Croatia 2006, we realize now how much hard ship and protesting we could have spared us had we pursued that idea already 2 years ago.