on topic. the long research times are indeed a balancer for longer games, I tried it from time to time and it really takes ages to reseach anything. on the cost side ... well, I guess higher costs will actually work, bc you also profit from the benefits a lot longer. if for example, a metal processing upgrade costs 800 instead of 400 and the according metal/ crystal adjustments, though with an unchanged effect, then it is likely equalled out by profiting from it for 8 hours instead of 3 or so.
so, research cost high - normal - low. why not?
also the other ideas with continuing research and extra levels and stuff. I'd say there is a few things you could do to make research better fit to very large games.
Bump.
I second this idea. I like VERY large long-lasting SP games and find the same problem. It makes it very easy to beat the AI, even with them all allied against me.
But then again, the AI in general needs much improving... especially for mods. It does the strangest things with new ships I make and has no idea what to do with the deployable mini-structures I made.
the ai does not handle it, because the ai was not made to handle mods. you will have to wait for skynet to get the functions you want. or you can try and fiddle with the ai yourself, which in all likliness is immensly complex work.
just to give you an example: if I made each race's battle capital ship 100 times as strong without any according changes to cost or other counter acting factors, my bet is that the ai will not change a bit in chosing what forces to employ. because it won't know this change has been made. at least that's what I think that the relative strengths assessments are more or less hardcoded and do not change dynamically with any mod induced change.
