I'm not surprised that Dart leaves out what I think is the most important reason for the size of the flight sim community, as I know he disagrees with me about this - but I think it's absolutely vital.
Flight sims are written by simmers. So they concentrate on simulation. Flight model, damage model, AI, appearance.
But most people want to play games. And games are about story, experience. Pretending to be a gangster in an open world where you can go anywhere and expect to find people to interact with in a "free" way, while simultaneously having little stories that you can opt in and out of. In this respect the game experience of most flight sims tends to fall way, way short of the technical simulation.
You tend to either get the canned mission effect, which has several limitations. First it's difficult to do well - just look at the volume of Lara Croft like games that get panned for their level design, and you get a lot more opportunity to muck about with the physical environment in that kind of game compared to a flight sim where once you're on the air - well, there's clouds.
Second its impossible to stop the player going off mission in a sim. In a game like say the Force Unleashed the level designers can force you down a set path; but not in a sim, once you're in the air the mission designer is dependent on you choosing to go in the right direction. And in a sim the player knows that they are "meant" to be on the Western Front in WWI, so if you fly in the wrong direction there ought to be troops, aerodromes, etc. etc. to encounter and to react to you. It's very, very easy to lose the suspension of disbelief when you know you can beetle about on the wrong side of the lines shooting at things for ever without encountering any resistance.
Third even the Lara style games are getting old hat compared to the GTA ones that try and free you up to make your own path through the game rather than follow a mission/level designer by the nose.
The other option is the dynamic campaign - but here again three errors occur:
First there is the failure to maintain suspension of disbelief by making the user too aware of the mechanics of what was going on. You want to con the user into thinking it's real - that means a great UI, and as little as possible to jerk the user into thinking "this isn't really WWI, there's an algorithm running that's making all this up - wonder what the rules its applying are?". That was where DCG and whatever the other one fell down for me in Il-2.
Second there is the failure to present it to the user as if the user were a single pilot - the Falcon 4 / Rowan's BoB failure. A brilliant dynamic campaign, but presented as a strategy game (and not infrequently expecting you to double up as a pilot AND the head of the airforce if not the Commander in Chief). Sometimes these go as far as divorcing you from the notion of "being" an individual pilot in the conflict to the extent of letting you drop into any plane in any squadron at any time.
Third there is the tendency to create a dynamic campaign that's just too repetitive - RB3D headed down this route, yes it pulled the wool over your eyes for a bit but as Dart has mentioned before after a while you just knew what the (pretty simple) mechanics were and it became a bit same old, same old.
Personally I think that the campaign needs to be the first thing a sim developer looks at, because that's where you're going to get the majority of your uptake by all the people who frankly are never going to be able to appreciate whether your FM is brilliant or a bit ropey. My suggestion is that it needs to be focussed on making the player feel they are taking on the persona of a particular individual pilot in the conflict, and then make the game feel as free and realistic as possible. Handle the edge cases - if they ignore orders and head off into the wild blue yonder then 1) make sure they encounter appropriate opposition and 2) court martial them when they get back. But like GTA, try and provide a lot of mini stories, preferably ones that are infinitely adaptable to different locations, that can draw the player in. Knick them from all anecdotes about war flying.
Oh, and come up with ways to encourage the user to explore all your content. I observed with RB3D that I had a pig headed insistence on Dead is Dead, restarting in Sept 1916 whenever I died. So I rarely saw anything beyond April 1917. It occurred to me that if I had the option to rejoin the same squadron in the same virtual world as a new pilot on the death of my previous pilot, I'd have been a lot more inclined to continue with the war and see a lot more of the content RB had to offer.
Last edited by Mahoney; 05/14/09 06:35 PM.