Terrain city bonus

[Gilgamensch]

Originally posted by MrBaggins


'Regrouping' will ultimately mean that the grouped units are ultimately closer to one point, than another.

Initially, the AI will have a generalized sense of defensive concentration. The closer the threat gets, the more definite it is.

Eventually, yes... the closer the threat the more exponentially threatening it is.

I think, I now get a feeling for how you want to implement it. If you get it working this way, would improve it a lot. In this case keep some inaccuracy in, like 15% percent +-. Otherwise, it would be hard to defeat the AI.

You can't describe the map, or individual threats to the strategic AI... You'd have to override the AI, using SLIC to a certain degree to achieve this.

Easy to do... you can easily tell what is being produced by an enemy... so you could play a little 'paper-scissors-stone' game... matching the 'threat' with an appropriate defense.

The AI should be well supplied with appropriate tech: it gets automatic upgrades, and coupled with some knowledge dissemination, it should never be so far behind as to be antiquated.

Matching threat with specific defense does complicate things a little (how do you split up troops when you have 'imperfect' resources?)

Even if you would leave as a minimum stack, like a combo of defender/long-range (1/1), it would be still better than now. You would always have imperfect resources (even as AI). The long-range should be upgraded according to the threat existing (introduction of plane => use air defence......)

You would need to include then the upgrade SLIC by 'can't remember'.

One thing about randomness, you could correlate it to the difficulty-level (i.e. Beginner 50%, Best 5% variation). Would really might high levels more difficult

[MrBaggins]

Originally posted by mapfi
Hey MrBaggins, this concept looks quite interesting. The whole thing should be worked out with Frenzy, producing a real complete overhaul of that script.

on a side note, I don't think we should give the AI knowledge of the enemies troups though.

Frenzy, at its heart assumes that the AI will invariably become disadvantaged as the ages pass, and hence needs to 'gang up on the human'... Perhaps thats true... but I'm looking to solve the root causes of this: developmental inadequacy... and strategic inadequacy (defense is especially important here- no defence= you lose cities= you get behind)

If AI civs stay relatively viable (or even superior) for the duration of the game, and there is a 'stalemate' of defensive capabilities... then diplomacy (and science/economic victory) should be possible.

If we introduce 'ethos' and encourage AI's (and perhaps the human) with similar ethos' to ally... then an attack on an allied member could certainly result on a 'ganging up'. It would be interesting to get changing ethos' too... like, perhaps Christianity/Islam... later changing to a more partisan form... such as NATO/Warsaw Pact.

MrBaggins

[Gilgamensch]

I believe the issue with the non-aggressiveness lies in two areas:

1) There are many more 'targets' of opportunity to pillage or 'HARASS'.

2) The AI is unaware of ablative attacks...
I.E. a single massed assault may not work, but multiple attacks could 'solve' the problem. All 'Free-Attackers' should be organized for efficiency, and then, if viable, attack.

For this:

The earlier might be fixable with dales suggestion (to be play-tested).

The latter: If you create the right combos and allow the AI a little peak behind the 'curtains (cheating)' it might be able to attack your right in your center.

But again would be needed to be playtested with the algorythm of MrBaggins. It could do a great job for this one as well.

You would only need to use it the other way round.
Spare units will get attack-combo, looking for the lowest defended city and go for it. Could even work for well-defended cities, just using the same algorythm.

Some people might consider it cheating, but it wouldn't be visible.

Also we might be able to include diplomatic features. Like only use it, when in war. And only after a certain threshold has been reached (units near/in the border), apply it.

[Tamerlin]

Originally posted by Gilgamensch
You would need to include then the upgrade SLIC by 'can't remember'.

I am currently testing Peter Trigg's Updater2 with the SAP2 and I can tell you this is another matter to attack and defend against an AI which is using up to date troops than attacking an AI with scores of outdated troops.

BTW, the Updater2 allows you to update your troops in the turns following the discovery of the advance triggering the updating process.

[Gilgamensch]

Originally posted by Tamerlin


I am currently testing Peter Trigg's Updater2 with the SAP2 and I can tell you this is another matter to attack and defend against an AI which is using up to date troops than attacking an AI with scores of outdated troops.

BTW, the Updater2 allows you to update your troops in the turns following the discovery of the advance triggering the updating process.

Tamerlin,

This is one of my next 'projects'. But if you like it, I think I'll gonna like it as well. the only thing I don't like so much, that you are limited to the next turn only (according to what I read..).

[Tamerlin]

Originally posted by Gilgamensch
Tamerlin,

This is one of my next 'projects'. But if you like it, I think I'll gonna like it as well. the only thing I don't like so much, that you are limited to the next turn only (according to what I read..).

With the Updater2 you can update your units when you want to and it is great. Peter and Pedrunn have worked seperately on two different updaters and Peter has finally borrowed some of Pedrunn's ideas to produce an outstanding mod.

I can send you the files (along with a french translation) though I am still testing them on a full game in order to spot any bug or design flaw. I am in 70 AD and it works fine.

[MrBaggins]

Originally posted by Tamerlin
I agree with you, this is typically the kind of cheat that can detract me from a game because it considerably reduces the strategical depht of a game (IMHO the Civ3 AI behavior is ridiculous). How can you build a strategy when you don't know the rules your opponent must respect? Though this is certainly coming from the fact I am playing tabletop wargames and boardgames I really don't like this.

Originally posted by Gilgamensch

I think, I now get a feeling for how you want to implement it. If you get it working this way, would improve it a lot. In this case keep some inaccuracy in, like 15% percent +-. Otherwise, it would be hard to defeat the AI.

*SNIP*

One thing about randomness, you could correlate it to the difficulty-level (i.e. Beginner 50%, Best 5% variation). Would really might high levels more difficult

Inaccuracy probably isn't the way to deal with this... instead, use 'visibility range': units must be within a (definable) minimum distance of one of the cities to be counted in the threat process.

The difficulty level would define the basic visibility range... low at Beginner, high at Impossible. Theoretically, you might want to increase the visibility range through the building of improvements which improve visibility.

I don't like introducing a cheat, if no cheat is necessary... but clearly... the AI is incapable of opposing a human. If anyone has any better idea of how to handle this... I'm more than willing to hear the suggestions.

MrBaggins

[MrBaggins]

Essentially, visibility range effects the AI's reaction time... If it can't see far, then it can only react to a certain degree: its troops can't teleport... so only a percentage of its plan is implemented, depending on the transit and movement capabilities involved...

[Gilgamensch]

Originally posted by MrBaggins

Inaccuracy probably isn't the way to deal with this... instead, use 'visibility range': units must be within a (definable) minimum distance of one of the cities to be counted in the threat process.

The difficulty level would define the basic visibility range... low at Beginner, high at Impossible. Theoretically, you might want to increase the visibility range through the building of improvements which improve visibility.

I don't like introducing a cheat, if no cheat is necessary... but clearly... the AI is incapable of opposing a human. If anyone has any better idea of how to handle this... I'm more than willing to hear the suggestions.

MrBaggins

For the range:

could be a nice idea, espacially if you could upgrade the visibilty of TI's and improvements. I never liked Globesat, but like local sat's, Watch-towers could be good. Visibility shouldn't go above 6-7 fields (large map+). This would also be a more realistic. So I even wouldn't consider it cheating. Even in ancient times, people reported what they have seen and the forces had been re-aranged. Would make sence. As there are still 'stealth'-units, surprise could still be there. So long we can tell the AI to use it as well........

You think it might be possible to include terrain as well? Like putting a 'watch-tower' on a mountain and increasing the range by 1, hill +0, anything else -1?

[MrBaggins]

yes

[Gilgamensch]

Will the AI, be able to handle it, that I think is the next question.

[MrBaggins]

Since the human doesn't get this strategic vision... and it applies only to the AI.... basically... yes

The AI gets a bonus when it builds vision imps...

Yes... we could give it a preference to build the improvements, in certain situations, but the more/most important thing is to have it develop its cities properly, first... including improvements.

[Gilgamensch]

But that would already be part of it.

No proper defence and everything would be gone......

[ahenobarb]

Originally posted by Tamerlin

*snip*

How can you build a strategy when you don't know the rules your opponent must respect? *snip*

That would be like real life! I couldn't resist.

Originally posted by MrBaggins

You can't describe the map, or individual threats to the strategic AI... You'd have to override the AI, using SLIC to a certain degree to achieve this.

Actually, MrBaggins put his finger on the main problem here even if you over come the hours of programming this will require, the AI is already programmed to behave in a certain fashion. I have noticed that it's preprogrammed behavior is extremely diffucult to override.

I have some code, where a specific type of unit is told to go to a specific location, but you cannot use a pthed move order, you have to tell the unit every turn to go to that spot, otherwise the AI takes over and does what it wants with the unit. And even setting the movement points to zero at the beggining of the turn, telling the unit where to move, adding the movement points back, only works about 90-95% of the time.

The AI code in CTP2, and probably most games, is incredibly stubborn. if only we could shut off a few of it's thoughts like Dave did to Hal, we'd be better off.

[Gilgamensch]

For the TI's,

one thing, which I was thinking about with your MOD:

As everyone will be 'euqal', shouldn't it be possible for the AI, via SCLIC, to tell:

if city inside [remember the circle]
and if counter (grass, desert) > 50%
then build science/farms?
elseif
build mines

Just as a thought? (OK, written in Pseudo-whatever)

this way the AI, could specialise some cities for science/trade and the rest would be production-houses.

For the impr's:

Similar for normal rushbuying:
For science-cities by after the 2 turns, or something like that......