I think it's worth analyzing the problems with quitting as a prelude to any solution. As I see it, there are basically two reasons you'd want to get people to not quit:
- When people quit the game in the first few minutes because they died once or twice, it wrecks the game for everyone else. The point of the game is to have fun playing it, and if people refuse to play it ruins everyone else's pleasure. This is similar to a kid who tips the board over because he doesn't like what happened.
- The current game mechanics mean that once someone on your team has quit, then unless you were ahead to start with, you can pretty much call it a game. (I have won games after someone dropped out, but only when I was ahead at that point -- the AI can play halfway decently if it gets to inherit good equipment and a level advantage)
I'm not sure if point 1 is as much of a problem as people think it is. It's annoying, but IMO the only reason it's really annoying is because of the connectivity problems we have right now. When it takes 10-15 minutes to get into a game, it's really frustrating when someone throws a temper tantrum and blows the game up right at the start; I don't have a lot of time to spend on this, and that can mean no games at all that evening.
If Stardock can get the connectivity to be better, then as long as this is something that happens in say less than 25% of games, I think I could live with it. e.g., maybe you can try to rig the system in Pantheon/Skirmish so that people who often quit in the first few minutes tend to play each other more than the general populace (but add a way for them to be "rehabilitated" if their habits change!). More importantly, there's basically nothing you can do to stop people from dropping out of a game that they don't want to play any more, so we really need to talk in terms of mitigating the damage to everyone else's enjoyment.
For point 2, I don't think that just dropping the Demigod and doing nothing else is a good solution (although it's probably better than swapping in an AI). If I don't have time to kill, I no longer bother playing 2v1; I've tried several times and unless someone on the other side is a total newbie, it's just pointless to play those games out. It's worth rehearsing why, though: the team with two Demigods has
- better map control: they can be in two places, you can only be in one.
- more damage output: if they're smart enough to stick together, they can win every fight by simply hitting you twice as hard as you hit them.
- twice the gold that you do: unless someone sucks and is dying a lot, all the gold income is from passive gold. The team with 2 players automatically gets twice your income, and probably more (since due to points 1 and 2, they probably hold the gold flags most of the time). This plays out in various ways...ususally that you're either behind in gear, in Citadel upgrades, or both, depending on how the two teams split expenditures.
I think that it would be interesting to see a system that tried to handicap the game after a drop to mitigate these advantages. It would also make unbalanced matches a more interesting prospect.
For 3 (gold advantage), the obvious solution is a flat-out buff to gold income (either passive only, or passive AND kill gold -- would make it even more important for the team with more manpower to not die). I think that less than 100% is good for the reasons stated earlier in the thread -- there are some gold thresholds that can be really overpowering, and letting one side cross them twice as quickly could be bad. But something in the range of 10-25% to start with, perhaps?
For 2 (melee combat), I would give the smaller side a boost to damage and/or durability. Again, I think 100% would be just too ridiculous, but maybe a 10-25% increase in HP and damage dealt? (i.e., like holding a really good Debilitating Flag and Fortitude Flag) You could also increase +HP and attack speed ... again, I don't know how much is enough here, just brainstorming. Hey, maybe 100% really would be reasonable -- we'd have to see how it played out.
I think 1 (map control) is the hardest to address. A boost to movement speed seems like it would be too overpowering while not really solving the problem anyway...maybe just give the team that's down players a decent-sized boost in the areas I called out above, so the game is about the larger team trying to use their mobility to win against an opponent who's overpowered in combat, rather than about them teaming up to harass and suppress a single player who can't possibly hold out against them. That seems like it could be kinda fun, even for the smaller team.
I don't have a great answer to any of this stuff (that's why you guys are the professionals and not me
) but I hope my rambling / brainstorming was useful, or at least amusing. The handicapping seems like something that could be modded in as an experiment if we had multiplayer modding support.