Espionage speculation

[KevinLancaster]

From the PAX demo, we now know that we will be able to warn other civs about an attack. Do you guys think it will be possible to give false information (to distract them while you prepare to attack) or plant evidence of an attack for another civ to discover?

 

[theadder]

I for one am not sure that the AI can cope with this sort of mechanic. There was a discussion about it in a previous thread some time ago.

My thought is that it would become too difficult for the AI to check if the player has lied, since it is possible for the third party AI to cancel its invasion; this might look like a player betrayal when it was not.

I expect that this will not be included, but we'll see.

 

[KevinLancaster]

Well, maybe they could have it so there's isn't a negative effect unless you actually decide to attack them. Also the "lie" could have a percent chance of succeeding depending on relations with that civ, their own espionage efforts, etc. Another thing is that maybe the positive modifier could be applied after the third-party civ attacks to prevent players from exploiting this lying mechanic.

Also, now that I think about it planting evidence would be a bit too convoluted and it would be too much based on coincidence to be worth it.

 

[theadder]

The problem in that case is that I can do it all the time so long as I do not take part in the attack itself; this would allow starting wars between the AIs very easily. Unless I am misunderstanding you.

It is probably just a bit difficult to put into the game well.

 

[KevinLancaster]

Well, there would be no point in doing it all the time as there would be no positive effect unless the AI you're lying about actually attacks (and maybe a negative effect if another AI attacks).

 

[theadder]

Well, there would be no point in doing it all the time as there would be no positive effect unless the AI you're lying about actually attacks (and maybe a negative effect if another AI attacks).

The benefit is that the AI that you are lying to is likely to do one of the following things:

1. Perceive soured relations with the AI that it believes to be preparing to attack it.
2. On that basis perhaps even launch any early attack on the other AI.

Those would be my potential motivations for lying to the AI; I am not sure that I can see any other reasons.

The AI that is being lied about does not need to take any action at all for either of those to be fruitful.

 

[KevinLancaster]

It could get less effective over time (for all AIs, since you need to keep up the web of lies) if you do it too often, and it could fail if the AI already has spies in the civ you are lying about (which eventually will happen if you lie too much).