I'm glad you enjoyed it.

Story time, hopefully you haven't heard too much of this before. I remember playing MicroProse's F-19 and there were times when I'd be RTB'ing with some MiGs on my tail. Often I'd buzz some of our cruisers hoping they'd somehow engage the bandits making my life a lot easier but no joy. I got so frustrated that I created a new pilot who went rogue and blew up a bunch of friendly assets and landed at an enemy airfield. I discovered you couldn't keep a negative score and you couldn't defect. I vowed that if I ever had a chance to make flight sims that if possible stuff like this wouldn't happen to anyone else.

So a few years later I got into game development and after a few projects I had the chance to work on the AI for Fleet Defender. This was a dream opportunity for me. FD was based off of the code for F-15 III but the AI code had been around a lot longer than that and they figured it was a good time for an upgrade, especially with the change in focus of the game. In the previous games the enemy planes would be spawned when the Player was detected and there would only be one group at a time. Of course when that AI was initially created they had to get around some pretty limited hardware so I don't fault them at all for this. I set it up so that so that any planes in the world would be there from the start unless they were something like replacements for the Ready 5 aircraft and so would be "spawned" below decks or on an airfield. I also really didn't like the idea of encounters feeling canned so instead of telling them specifically what to do I gave them behaviors/jobs. AIs had waypoints to follow at the beginning and upon reaching a waypoint that might change the behavior for the flight, like "start attack run" (with details on what the targets were, profile for the run and such), but there was a Command AI that could change the tasking for a group or just interactions with other elements would mean they'd have to adapt. This meant that the designers could just layout the basic mission but they didn't have to worry about the minutiae. It also helped with the problem of if you assume a Player will do things one way you can be sure they won't. The comms just came out from the AIs doing their jobs. People quickly realized in FD that it was no longer "You Against the World" but with the extra audio we could had for JF-15 and then expanded upon for JF/A-18 it was even more evident about how much other activity was going on in the world.

If we had done another game after JF/A-18 I probably would have been doing a dynamic campaign. We already had persistent damage in the game and my guys could already adapt to a changing situation so most of the work would have been developing an AI to generate the missions, probably between the Player's missions and in a mission format similar to what we already had, taking into account how the conflict was going, the priorities for the different sides and the resources that were available. With doing what I felt was a pretty good job with the full co-op multiplayer in JF/A-18 that would have been the last big bucket list thing I would have wanted to accomplish in a flight sim. Part of me was bummed that I didn't get the opportunity but we were so fried from the past few games that when EA decided it was time to move us to something else we were mostly OK with it. A few years later I was queried about my interest in doing a dynamic campaign for another sim but this would have been a contract position and once that was done I'd be having to find another job so I figured it was better to stick with a full time position.

Elf