I can see that, but I can also see why Stardock did what they did. Because a) you don't want to turn Galciv into a Civilization wannabe, and it's space. Space doesn't particularly have borders. I think I heard someone (Brad, maybe?) say a long time ago that the planets are the borders. In other words, that's just your sphere of influence. The political borders are the planets.

Here's a couple of suggestions (off the top of my head) that could somewhat ease the whole "AI can't handle surprise attacks" problem that wouldn't require excessive AI coding and so forth:

1) Moving next to an AI planet with an armed ship or transport is considered a declaration of war (unless you have an alliance).

2) Moving next to an AI planet uses all remaining movement points for a given ship.

This would at least give the AI a turn to respond to a sneak attack.

Cheers

h

Wow, just did a test game and your right it's true. Very suprising. Did Stardock intend this I had thought previously it just affected the planet it was built on.

On larger obscene and suicidal maps I guess evil may be the right option purely because of this wonder. On obscene and suicidal your going to be playing catchup a lot and 100% eco bonus will certainly help to catchup although it will take time to get. On most maso maps I find it easy to outcolonise the AI so am never really playing catchup so neutral is prob the best choice then as its benefits come much quicker.

couple inspirations taken from this thread, thanks everyone!

as someone suggested the ai should detect which weapon and defense techs you are researching. to make it interesting there should be a diplomatic option to limit what kind of techs you research or implement in your ship designs, like an embargo: no missiles on your ships else lowering diplomatic relations or worse.

then something like a skirmish: for example if you attack and destroy someones ship in your space it does not become a war directly but just lowers relations and you are not seen as agressor. of course if relations are already low it will not be tolerated and war is declared. even more sneaky would be to have a diplomatic option saying you accidentally destroyed someones ship anywhere in space or close to some other conflict - reaction depends then on diplomatic relations.

the border could be a close limit around your planets (just a ships range or less) but you choose whether to enforce it or not, having for example trade penalties if you set up a very restrictive limit but bonus in terms of restricting someones influence. or you could even choose to have it as border starbases or starbase modules. something like a border or customs module or base which allows you to set up borders and claim space and then decide to do levy customs on freighters, limit influence, etc, all with realistic malusses associated of course. or it could have possibilities to control other ships access to that space, with possibilities to warn ships/races against infringement, or even being able to force them to move out. or having it as a ship module where a ship can make another move out of the space and if the infringing ship does not follow it becomes an attack -> diplomatic situation -> war?

probably all not easy to implement well but imo additions to better represent cold wars, diplomacy, politics in general and to give more options - taken from real life

ps.
dirty weapons? massively destructive but with diplomatic consequences, embargos? conventions against them? rogue civs?
add more shades to it all!

And how, exactly, would I ask permission? They would have to do something like add an option on the trading screen that says "let me move through your space". Of course, once that was available, a human would just find a way to get the AI to agree to let him go through its territory, move in his transports, and then sneak attack anyways, putting us right back at square one. That's a whole lotta work on the devs' part to slow down an exploit by three or four clicks.

Ok, let me give you a scenario:


You have the Drengin on one side of you, and the Korath on the other. The Drengin and the Korath go to war with each other. Drengin ships cross your border headed for Korath space, and vice versa. Do you, as a player, immediately declare war on both of them, despite the fact that they don't care one whit about you and would probably beat you senseless if you fought both at once?

By your statements, you're saying that yes, you would declare war on both the Drengin and the Korath, immediately and automatically. Doing otherwise would be stupid!


Here's another one. You have the Torians on one side of you, and the overwhelmingly powerful Yor on the other. You're allied with the Torians. The Yor attack you, triggering the alliance and dragging the Torians into the war. In order to help fight the obscenely strong Yor military, the Torians send their ships across your border. Do you, as a player, immediately declare war on the Torians?

By your statements, you're saying that yes, you would declare war on the Torians, immediately and automatically. Doing otherwise would be stupid!


Sarcasm aside, I hope you can see why programming the AI to be paranoid nutjobs with a hair-trigger "declare war" button would probably just make them even more stupid than they are now.

I have not said that, thats your interpretation. I meant that the AI is being stupid when it waits for all your transports to position yourself just outside their planets, then declares war and then after you already invaded all his planets he starts to react. So its not that the AI should declare war the moment one player ship is in his space.

Also your comparison to what a player should do in the same case is invalid because we know how the AI works and we know that it doesnt do surprise attacks as the AI is very predictable by first declaring war and then starts sending his ships to attack you. Since the AI will not "know" what a human is doing there should be coded in the AI to look for signs that an attack is coming.

So the AI shouldnt be coded to respect other empires space since its already predictable enough to first declare war before starting to invade so there is no need for such mechanism.
But to protect the AI from cheap sneak attacks from a player, if the AI sees a great number of player ships in its space then he should first warn, and then if they are still there it should declare war and react. Or atleast react by having ships ready to defend his planets (besides the sitting ducks in orbit without orbital fleet management). This should give him a couple of turns to prepare for a possible invasion. Just as a human player would do if he was playing against another human player.

In your second example, as Ive mentioned, since you are allied with Torians then this violation in your space should be allowed. There should be some kind of thrust between two allied parties.

You have to see it in the terms of "turns to get to your planets" rather than "millions of light-years". When speed is extremely high (as in this game) then extremely long ranges really means nothing since an enemy can pass through that in one or two turns anyway.

And if the player gets dragged into a war by an ally, and the only way to help is to go through that ally's territory? "Why thank you for helping me, but I'm afraid I have to declare war on you as well since your ships are close enough to help me."

What if the player's empire is split into pieces by an AI's territory (i.e. a "pocket" surrounded by the AI and possibly map edges), and the Korath declare war on the player? Gee, guess that "pocket" is Spore-fodder since the player isn't allowed to move ships to protect it without their AI neighbor declaring war over it.

And even if you DO give the AI a hair trigger, a human player will find a way around it. That's the big problem. No matter what you do, the AI will be unable to defend itself from a sneak attack because the human player will find a way to either get around the restriction or, even worse, use it to sneak attack anyways.


This pretty much admits that there's nothing a player - human or AI - can do to stop a sneak attack without being a paranoid psychopath.


A perfect exploit, then! Build up an alliance, then sneak-attack when they least expect it!

ANY leeway the AI EVER gives a player is an invitation to exploit it. Every measure you implement to do away with false positives just opens even more doors to get around the safeguards and do what you were going to do anyways.



The ONLY way to make the AI able to react "in time" to a sneak attack is to declare war the instant one of the player's ships - ANY ship, even unarmed ones - gets within two turns' move of the AI's planets even if they're still in the player's territory. Anything less, and the player WILL find a loophole, period. But frankly, if the choice is between a bit of naive trust vs. being a foaming, raving psychopath, I'll take the naive version of stupid over the blood-crazed moron version of stupid any day of the week.



Edit: Just for the record, I'm not arguing that they shouldn't make the AI better. In my opinion, the better the AI is, the more I like it. Good AI = Good Thing. What I'm arguing is that it is NOT as easy as "just make the AI do X"

And I agree with Legacy, Yamota. You're suggestions simply aren't at all practical. There is no AI that would be able to out tactic a human player without some sort of rule change.

There have been a number of real-life scenarios where countries have backstabbed allies without morale penalties. Nazi Germany on Russia, Israel on Great Britain, Italy on Nazi Germany (well okay maybe that one had morale penalties), Russia on China....

Generally I think the AI plays well. The only exception to this is when I am invading his worlds. Without fail there will be a ton of ships stacked at his planet that, were they combined into a fleet and launched, would repel my invading ships. Yet every time the ships stay in orbit for me to pick off one by one. Not the brightest strategy really.