While you guys were playing AW and Warbirds, I was a sonar tech on a submarine 1989-1997. I didn't even have my own PC until I went back to USF to finish my engineering degree in 2000. Prior to that, I had played Jane's Fighters Anthology on a work laptop with only keyboard/mouse controls and my submarine had a PC with Chuck Yeager's air combat sim on it in the mid 90s, again keyboard/mouse. So, when I got my own PC around August of 2000 and made sure it was good enough for flight sims, I had downloaded Warbirds 2.7x and eventually stumbled on to Aces High. Initially, I liked Warbirds better because of the offline single player mode and huge plane set. But I eventually came to understand the difference between a 3d cockpit and 2d bitmaps and also found the flight modeling much better in Aces High.
My only real complaint about Aces High back then besides the lack of a P-40B/C was the accuracy of the guns. I could regularly hit with 0.50 cals at ranges of 800-1,200 yards, enough to shoot off tail surfaces. This also made guns exceptionally effective in a head on. I always set my sights to the max range adjustment to take advantage of this. But Aces High 2 changed something... seemingly the stability of the aircraft as a gun platform, so the guns have behaved much more realistically ever since. Setting the sights for more than 300 or 400 yards is not very effective, just as experienced by pilots in WW2. I can still get hits at 800 yards against AI flying in a steady circle, but not with laser precision. It now wastes a lot of ammo to do so, and against maneuvering targets, it is almost useless. In my opinion, it would be very difficult for the competition to produce a flight sim that had a comparably sized flyable plane set with equal or better physics. I love DCS World, but that level of modeling guarantees that the number of flyable aircraft doesn't grow very fast and they have yet to produce a useful number of historical terrains. The flight models have much greater potential for realistic behavior at the edges of the envelope and for engine performance/failures, but that potential is limited by how difficult it is to get the modeling correct even if you can get enough information to do so.
