Tweet
I have CMNed a few games, and am playing in a few PBEM. While Darnan's fast start is not for me - the extra units, etc. distort faction balance IMHO (but that's like discussing female anatomy, nobody is right or wrong, we just have different priorities 
), he has come up with an excellent strategy to keep the AI in play and competitive with the players.
The three AI's are Miriam, the Hive, and the Spartans. They are all pacted together, which means (I used a scenario editor to test this) that they will vote as a block. This means that building the Empath Guild does NOT become quite so powerful, you still need a coalition of the other players to get elected. Between Miriam and the Hive some AI will always be up for election until well into the midgame.
Plus he gave Miriam the Planetary Datalinks. Evil, and combined with the pact, very effective. It means that the AI will never fall too far behind the players, as between trading or parallel research you will routinely get three players with the tech. It also makes sense, and could be argued to more effectively represent the network of fanatics supporting her cause. In the game I'm playing, combined with some superior starting locations and the de facto sharing of techs by pacted AI's, the AI's in this game will probably beat the humans to Air Power.
The only thing that would have strengthened this combo would have been better interpenetration of the AI pacted empires, so their units come in contact and they trade more with each other. I may combine that with my own strategies, which I will mention below. (By the way, I've seen links to an article by Ogie or Googlie on scenario editing, if anybody has the link please post it here).
I edit the AI tech foward and give them clean reactors and prototypes certain units, then take their techs back down. I also carefully pare the prototypes to what I want them to build. This can lead to the AI building things like Clean Formers and whatever mix of military units you think makes the most sense (i.e. 1-3t-1 clean garrisons). On larger sized maps this can lead to a much more competitive AI, as the overproduction of units doesn't screw things up as badly.
Giving the initial bases a Recreation Commons can help as the AI is really not that good at managing drones. However, in scenario building I discovered that you can give the AI too many base facilities, and they start grabbing the SP's all over again. Some free terraforming also helps, I like to give the AI a road net and sometimes a few boreholes, for whatever reason the AI seems very bad at creating those. Rain Forests - jungle forests - seem to be extremely potent for the AI, and I've had to be careful about them. The AI will grab alot of SP's if you put them in the jungle with a Rain Forest economy and clean units. Enough so that you need to be fairly careful.
The same applies to mineral resource mine squares, especially with crawlers. Let the AI have a crawler that begins at that kind of location, and it will build SP's almost from the start, and beat the humans due to the reduced costs for transcend level AI's. In fact combining Rain Forests with crawlered mineral resource mines means humans will get almost none of the first five SP's This is something which to me reduces the fun of the game, going after those early SP's is a critical part of the early game strategy.
I've found putting Energy Resources under bases gives the AI a much better chance to maintain research parity with humans, the AI is just so bad at managing research. Besides doing that, and the obvious jungle squares, extra mines and monoliths, roads, and base sensors, what other techniques have you folks come up with to keep the AI competitive, but not too powerful. This is for all the people new to CMNing due to the interest the rerelease of SMAX has produced.
), he has come up with an excellent strategy to keep the AI in play and competitive with the players.The three AI's are Miriam, the Hive, and the Spartans. They are all pacted together, which means (I used a scenario editor to test this) that they will vote as a block. This means that building the Empath Guild does NOT become quite so powerful, you still need a coalition of the other players to get elected. Between Miriam and the Hive some AI will always be up for election until well into the midgame.
Plus he gave Miriam the Planetary Datalinks. Evil, and combined with the pact, very effective. It means that the AI will never fall too far behind the players, as between trading or parallel research you will routinely get three players with the tech. It also makes sense, and could be argued to more effectively represent the network of fanatics supporting her cause. In the game I'm playing, combined with some superior starting locations and the de facto sharing of techs by pacted AI's, the AI's in this game will probably beat the humans to Air Power.
The only thing that would have strengthened this combo would have been better interpenetration of the AI pacted empires, so their units come in contact and they trade more with each other. I may combine that with my own strategies, which I will mention below. (By the way, I've seen links to an article by Ogie or Googlie on scenario editing, if anybody has the link please post it here).
I edit the AI tech foward and give them clean reactors and prototypes certain units, then take their techs back down. I also carefully pare the prototypes to what I want them to build. This can lead to the AI building things like Clean Formers and whatever mix of military units you think makes the most sense (i.e. 1-3t-1 clean garrisons). On larger sized maps this can lead to a much more competitive AI, as the overproduction of units doesn't screw things up as badly.
Giving the initial bases a Recreation Commons can help as the AI is really not that good at managing drones. However, in scenario building I discovered that you can give the AI too many base facilities, and they start grabbing the SP's all over again. Some free terraforming also helps, I like to give the AI a road net and sometimes a few boreholes, for whatever reason the AI seems very bad at creating those. Rain Forests - jungle forests - seem to be extremely potent for the AI, and I've had to be careful about them. The AI will grab alot of SP's if you put them in the jungle with a Rain Forest economy and clean units. Enough so that you need to be fairly careful.
The same applies to mineral resource mine squares, especially with crawlers. Let the AI have a crawler that begins at that kind of location, and it will build SP's almost from the start, and beat the humans due to the reduced costs for transcend level AI's. In fact combining Rain Forests with crawlered mineral resource mines means humans will get almost none of the first five SP's This is something which to me reduces the fun of the game, going after those early SP's is a critical part of the early game strategy.
I've found putting Energy Resources under bases gives the AI a much better chance to maintain research parity with humans, the AI is just so bad at managing research. Besides doing that, and the obvious jungle squares, extra mines and monoliths, roads, and base sensors, what other techniques have you folks come up with to keep the AI competitive, but not too powerful. This is for all the people new to CMNing due to the interest the rerelease of SMAX has produced.
. His defensive layout is a little weak, though his tech and support makes up for that. Plus he got the bloody Command Nexus which is going to make squeezing him much harder. I will not mention which of the other AI's got what in the base layout and in units, I'm going to let the other players find out on their own 
Anyways, yeah, giving people speeder CP's instead of regular CP's does accelerate the early game, as this phase of the game does drag a little, IMO. Just be aware that you as a CMN need to give each Faction at least one regular CP, or the game will crash, as there is some hardcoding here that says that if the game doesn't detect a regular CP on the opening turn for each Faction, the game will terminate. FYI. 
However, just be aware that only the #UNITS Section of the alpha(x).txt file is imprinted onto a PBEM game. FYI in case other people are using this tool for PBEM's.
I'll probably be starting a picture thread over @ CGN with some camping pics from this summer in the near future. FYI.
The old avatar was from one year, this new one is less than a week old - she's a little over 18 months. Still no TV, still wonderfully behaved within the context of being 18 months old. At least that's what EVERYONE tells us, and I do not exaggerate. OK, UP - we love to sit on daddy's lap when he's on the machine - she has her own old IBM keyboard she sits on, pounds on, and pries the key covers off of. That is if she cannot get on daddy's lap.
Comment