Although I know little about air marshal mode it sounds pretty interesting for us MP folks. The great thing about MP is the adrenaline rush of fighting real humans with unknown and highly variable skills. I agree with Jason that it should stimulate MP participation and bring out what really makes MP fun-teamwork.
I've been thinking that merging this with functionality similar to what we find in persistent dynamic campaign servers we would bring what SP mission players love-a personal contribution to a strategic objective.
If Air marshal mode could be expanded to enable the air marshals to quickly and easily command their side's AI and ground resources to be placed or moved by taking direct control of what coding and algorithms do now in dynamic persistent campaign servers this would open up a whole new dimension to the sim. This would mean making it easy to command the movement of convoys, direct tanks to attack certain enemy defenses when the air marshal thinks force balance is in his favor, to place his limited resources such as flak to points of his choice like around airfields, critical bridges, factories that have vital production etc. Maybe even deciding which factories produce what. Probably allowing changes on the fly during map action would be tough from a coding perspective but a time slot could be set aside during the map roll to let a commander quickly assess what's going on and, working within rules-based limits. direct things to move to achieve a strategic objective. Of course, guessing what the guy on the other side is going to do will be a big part of success.
To do this some work needs to be done in multiplayer like improving the net code to allow more objects that interact including AI and players. Its hard now to get a player limit up to just 32 with very modest amount of AI moving about and interacting in the server. Maxing out at 16 players will leave you hunting for a fight on a full size map. I would love to see lots of AI activity in a server that could handle 100+ players. Also, better and easier accounting of losses and production and more attention paid to the resiliency of objects and make sure updates don't wreck the behavior of objects (like bridges). Also AI behavior could use some work especially in ground-based combat.
I loved playing the Coco server (rest its soul) because you could make a contribution to a multi-week-long campaign. Want to bomb factories or storage?- that reduces the enemy's available resources, want to dogfight?- your kills reduce the enemy's plane availability and sends pilots back to a rear field, want to hit tanks engaged in a battle? this helps your side reduce resources and enables taking control of the enemy's territory and local assets.-All of which factor in to the opening state of the next map. Eventually, everything that has happened during that multi-week period serves to influence the final outcome when a side can no longer defend its last base. It also allows one side to turn the tide.
Its so much better than a points type server where winners just need to shoot down a few more planes or knock out a set number of targets that repeat every couple of maps so it gets uninteresting pretty quickly. You kind of have two competitions going on at once, one inside the other. Individual players and their skill level and numbers factoring at one level and the air marshal and his ability to think strategically in another.
I think Coco gave up his beloved experiment because the updates kept causing fatal issues that too often prevented the server algorithms from doing their job. He had to do all kinds of algorithmic gymnastics to make it work within the game's limited ability to handle AI (1 tank had to equal 5) and to adequately calculate the results at the end of each map roll. All this takes a lot of time and effort even someone with Coco's exceptional talents.
It was a great idea though and it was a lot of fun.