[Oryx]

Wow, I haven't posted about Il-2 for a very long time, but just cannot resist with a comment like the one below:

Originally Posted by K_Freddie

So, aerodynamic maths explain 100% of flight, a 100% of the time

Actually, it does. For the simple reason that flight is described "perfectly" by the laws of physics. There are no exceptions. At least, in many years of flight testing from inside the aircraft and from a telemetry station on the ground, I have yet to see a single case of an aircraft defying the laws of physics. Also, today the knowledge surrounding the various factors that contribute to the forces acting on the aircraft are very well understood and can be modelled very accurately - at least accurate enough that the prediction models get extremely close to the measured ones.

When it comes to flight simulation, however, the programmers must strike a balance between fidelity and practicality. The most difficult part is not to model the kinematics - the equations are quite simple - but to generate the data to populate the models. To generate this data is both time consuming and costly. I have worked with simulators where the fidelity was so high, that we would sometimes only use spot checks in actual flight to confirm the simulator predictions. However, to achieve that level of fidelity took wind tunnel tests, numerical predictions (such as CFD) and also in-flight systems identification. The costs are obviously staggering to create such a model. This is not feasible for a game, especially ones where more than one aircraft needs to be modelled, so developers have to make some decisions on how far to go in the modelling process. The result will always be a compromise. It doesn't mean there are some voodoo aerodynamic effects going on that engineers don't understand.

and the pilot's always wrong, according to the 'propellor head' on the ground.

I don't think it is one or the other: I often rely on test pilot comments during testing. However, what is very true and I think both you and Gaston seem to miss, is that the combat environment is probably the worst possible time to compare aircraft performance. There are just too many things going on to make any quantitative judgement, unless the differences between the aircraft are really large, such as a jet vs a piston-prop. For example, does the other guy really have a faster aircraft, or did he just start the fight slightly higher and was able to build up some energy into the merge? Or does he really turn better or is the fight just happening closer to his corner speed than yours? Half the time, you won't even remember your own configuration during the fight (speed, height, throttle settings, etc), let alone what the other guy was doing. Even if one aircraft type consistently outperformed another on a certain aspect during combat, the reason might very well still lie with better tactics rather than a true performance advantage.

I say the above with utmost respect to fighter pilots with whom I have also worked extensively. I have seen clearly inferior aircraft consistently beat superior aircraft in mock combat when the pilot in the inferior aircraft was experienced, especially when he was experienced in both types. An example that I have seen with my own eyes were fights between fighter trainers and front-line fighters, where instructors in the trainers could consistently give rookies in the front-line fighters a hard time. I bet some of those rookies were thinking to themselves that their mounts were not nearly as good as advertised, while in reality the instructors just understood better how to exploit the strengths and weaknesses of the two types.

Somewhere earlier in the thread I think Gaston referred to "canned tests". Yes, that is exactly what one has to do during flight tests to determine the true potential of the aircraft. The only way to really know is to isolate parameters one by one and then test them. Combat is not the time to measure what the aircraft can do - combat is the time to put that knowledge to use.

By the way, flight testing is about much more than performance - I have spent much more time on flying qualities and handling qualities testing than performance testing. Handling qualities are extremely important when the question comes up on whether you can consistently extract the maximum potential out of the aircraft. Yet, on gaming simulations the topic of handling qualities seldom come up as few people know how to measure and interpret them. Of course, these days even more time is spent on avionics and systems testing, but that is another topic.

A small final comment before I let you guys be. I honestly don't have the time or energy to comment on every point made by Gaston, but this one really stood out:

Originally Posted by Gaston

My note: horizontal combat was never considered outdated in all of WWII, except for the Allies in the Pacific: It covers about 95%+ of all Western air battle in 1944

This is simply not true. There were many dogfights in WWII and many did indeed end up in horizontal combat. In fact, I bet the natural reaction when bounced is to turn. However, attacking from superior height was the preferred method for just about everyone and it remains so today. Horizontal combat (in fact, dogfighting in general) is always a gamble - you may or may not win depending on the relative skill of your oponent. Attacking from superior height gives the attacker an "unfair" advantage, even if he has an inferior aircraft. Entering a fight with the plan to immediately enter horizontal combat is never a good idea. Only once the fight developes into a dogfight might a pilot with an aircraft with known good turning performance prefer to stay in the horizontal. The most consistent results always came from attacking using an energy advantage (height or speed) and then to get the victim on the first pass, ideally without him ever seeing you and without letting a "dogfight" develop - and this is exactly what accounted for the majority of aerial kills in WWII.