(August 20th, 2024, 18:32)haphazard1 Wrote: Interesting. The free colony ships are one of the biggest "balancers" that the AIs get in original MOO, presumably because the AI would just be horribly bad at expanding otherwise.
From RefSteel's experience, I guess that rocks can just put their entire home world output (plus production bonus) into colony ships right from the start. No clean up spending needed, and that is a significant factor early.
I probably didn't describe the game situation well. It was a 24-star map (the same as a Small galaxy in Orion) with three opponents who turned out to be the bugs, rocks, and brains. The 'coids didn't get
everything - they just got the first colony out
fastest, and I'm sure they didn't build a colship from Turn 1. With a x 1.45 production bonus (I was playing on Hardest) Silicoid factories pay for themselves in about 5 turns, and it doesn't take them long to catch up with their homeworld population. The bugs' factories aren't quite as quick to pay for themselves, but their economy is even more ridiculous with the x 1.45 bonus. The MoO AI would be wasting its (even more immense) early production advantage on the likes of reserve tax, too-early missile base and ship construction, and inefficient research choices, and still always gets out to a big early lead until (and if) the player's far-superior strategic ability can overcome its far-superior productivity.
Xilmi's AI doesn't do that though; it goes full-throttle on expansion. In my test game, I managed to claim a rich minimal world for a third colony by prioritizing it extremely heavily, but the brains grabbed 'coids and bugs meanwhile were taking over every other habitable (i.e. non-hostile) world in the galaxy apart from one or two stolen by the brains, with the rocks of course poised to grab all the hostiles unless I stopped them militarily ... with a huge initial deficit and of course lower production per turn. Probably it wasn't winnable at that point, but it might have been interesting to see! The first test game I played was deceptive because I didn't look closely at the set-up and so wound up with the default one opponent for a "micro" galaxy, and since I was testing that one at "normal" difficulty, all I had to do was outplay the AI. On Hardest, you have to outplay the AI
by a huge margin, and the Xilmi AI is good enough that this is a huge challenge when it's doable at all. Fortunately, there are lots of difficulty levels in between. (And even on normal it's likely to be more interesting and challenging with more opponents in the game.)
