AI and Diplomacy

[Hydro]

I’ve recently had a rather extraordinary run of AI diplomacy in a current multiplayer game. As Hive with PS/Planned I’ve maintained a Pact with Dee for ~70 years and the PKs for about 40 years. I’ve worked very hard (and haven’t used cheesy SE switching before calling the AI) to keep this Pact by:

* Granting requests of tech

* Trading any tech when offered, and offering to trade techs

* Giving tech when my diplomatic standing deteriorates (this happens a lot with the PKs, who for some reason don’t like my PS)

* Making sure we are all at war with the Evil Spartans (the other MP player) so we have a common enemy

* Gladly declaring war on my pactmate's enemy (Pirates), and then prosecuting that war

* I only call once per ten years or so, or when they have a tech I can trade for (eg – they’re done with a SP and are willing to trade)

This has been interesting since it seems this is what Vel commented on in his MetaGame discussion in his strategy guide. Although I’ve given away a lot of tech I’ve also gained a lot via tech trading. The commerce income is a huge benefit for the Hive since I couldn’t otherwise support my infrastructure while running a 10ec/turn deficit (with a 60/40 lab/eco split).

The benefits go way beyond economics, however. By keeping the PKs and Gaian in ‘my’ camp I’ve isolated the Spartan MP, and forced him to ally with the Pirates, who are the second biggest empire but have a pathetic lab output of 2.4/turn (that’s right – 2.4!! The idiot's running Fundy/Planned). My buddies the Gaians have 34, the FM and golden age Lal 105, and I have 135/turn. This alliance also forces the Spartans to think twice about attacking, denies them tech trading, and limits their expansion plans (they conquered the Drones ~2150). In short, without attacking anyone I’ve forced the Spartans into a corner with an aggressive AI diplomacy.

It won’t last forever, though. The PKs are #3, and if they ever get to #2 our alliance will be dust. They may decide to ditch The Hive if Pravin’s mood is poor when he calls me next. I have a better chance keeping my Pact with the Gaians, who are #5.

On the whole it has been very enlightening, and I’ll be the first to admit a lot of good luck. For instance, I got an immediate Pact with Dee ~ 2125 before I could go Planned, so have a lot of ‘good’ history brownie points. My war with the Spartans has aligned us to give more brownie points. But, the biggest bonus is the huge bump in attitude when I give tech away, or trade tech. This has been key.

In the end it will go south, likely with the PKs first. When that happens Pravin will have build a lot of cities with excellent infrastructure for the People of the Hive. After all, I’ve always thought Pravin would make an excellent drone…

[Snowflake]

I wouldn't object to bribe the AIs with money and techs, but I wouldn't trade techs from the AIs if I don't need them. I always try to keep my research cost as low as possible.

Also I've seen research output of 400/turn at 2200 (or approximately 3-4 turns per tech) for Hive I believe.

[Maniac]

Great game!

[Snowflake]

Maniac

[Hydro]

HongHu - we're not using directed research, so b-lines are almost (but not quite) meaningless. In order to get higher research you need energy restrictions lifted, tree farms, and probably hybrid forests with the juicy +2 energy/tile. You also need to do some pop booms w/GA (probably) to goose your populations, and that takes TF and HF, too to increase you psych and get a multiplier effect with a little allocation to psych from energy reaserves.

Using blind or semi-blind/directed research it is generally best to take any tech you can since it means you won't waste time researching it. Again, non-directed research is a completely different mindset. As a result the tech count may be higher, resulting in higher times between techs.

[CEO Aaron]

As a Morganite, much of my game pivots on successful diplomacy, so I've come up with lots of helpful guidelines on making the computer into the aggressive arm of your economic expansion.

1) You're going to have to screw any faction you share a landmass with, so don't waste too many resources keeping them friendly.

2) For factions you don't share territory with, always make the effort to make friends with factions whose aims are compatible with your SE goals. Sure, you can steal tech, but trading it is much easier, and you'll net yourself a good amount of trade income to boot.

3) Know who your enemies are going to be, and exploit them ruthlessly before you make them your enemies. If you can shake them down for tech or cash, don't be afraid to do so. It's easy to keep an honorable reputation while screwing the computer blind, if you're patient enough.

4) I give any compatible faction any tech they ask for, unless I'm planning their demise and they want armor or reactor technology.

[Hydro]

Very Machiavellian! I like it!

[Snowflake]

Originally posted by Hydro
HongHu - we're not using directed research, so b-lines are almost (but not quite) meaningless. In order to get higher research you need energy restrictions lifted, tree farms, and probably hybrid forests with the juicy +2 energy/tile. You also need to do some pop booms w/GA (probably) to goose your populations, and that takes TF and HF, too to increase you psych and get a multiplier effect with a little allocation to psych from energy reaserves.

Using blind or semi-blind/directed research it is generally best to take any tech you can since it means you won't waste time researching it. Again, non-directed research is a completely different mindset. As a result the tech count may be higher, resulting in higher times between techs.

Yes that make great sense. Sorry I wasn't aware that you were doing blind research.

[Commy]

Ah, okay, one thing about AI diplomacy. Why does the weaker factions always demand stuff from you? And why do some games I'll do a one city challenge, be in last place, and my city, that can't do crap, is the one that's nuked? Oh, and why is it that the AI always has to demand two techs instead of one on higher levels? That really ticks me of. One second, I'll be in the tech lead, then the next I'm in the dark ages.