AI Plays GA more like conquest!

[EatYerGreens]

Just an observation, this.

In my current HRE/Early/GA/Hard campaign, I can't help noticing how badly some of the AI factions are doing at achieving some relatively simple GA goals, like holding onto their homelands. Last time I looked, the Byzantines had lost every single one of theirs and stand to score only about 5 points when the next tally comes (1204), from their conquests. They have the Novgorod backed into a tight corner (just Finland) but have themselves been pushed into Europe so that they border with me and now the Turks have arrived in Hungary, facing one of my borders. ~:eek:

Factions will decline offers of my princesses and may have died out, for want of heirs, as a consequence, zeroing their points completely. I can't get my princes married off either but the HRE has the get-out clause of electing new leaders anyway and that's less of a problem for me.

Basically, the AI control seems to be running other factions as if they're still aiming for total conquest. I don't get it. Anyone else with observations similar, or different, to this?

[OlafTheBrave]

Yes unfortunatly the AI seems not to focus on GA goals other than perhaps an attack against the player right before a count year. Quite often late in GA games I have seen sprawling AI empires that do not include their home provinces steadily advancing away from such said homelands. Also supposedly according to CA the only benefit to marrying foriegn princesses is the new king will start having offspring apon taking the throne. Its something I dont concern myself with much. The AI honors marriage backed alliances just like any other, ie it dont. I have had unmarried heirs assume the throne with ofspring already before they marry a noble's daughter and in my observation the in-bred vices seem to be a bit random. In fact the main trend I have noticed with V & Vs is that if you have one positive and one negative that the offspring will tend to have one positive and one negative.

[EatYerGreens]

So you're basically seeing the same kind of conquest activity I've been seeing? Interesting.

The thing is, it would take some extra coding to make the AI see a homeland province, which earns (say) 350 per year as being more worth holding onto than a province with 650, over a border on the other side of its domain. If, additionally, it sees the defenders there as 'conquerable' but the faction who stole its homelands are too strong for it to be taken back, then Sun-Tzu style coding will tell it to attack the weaker faction and forget about getting the homeland back. The net result, as the Byz did, in my campaign, is that they slowly creep across the map, in search of higher income or other factors (defensible terrain), all the while losing as much ground as they gain.

[yesdachi]

I think you guys are mostly correct but it seems that if left alone the AI does follow their goals, perhaps more so in easy. But the AI forgets them when things aren’t going their way. The Spanish, Egyptians, and Italians seem to push towards their goals more than others but that doesn’t say much. On hard or expert I see most of the point leader’s points coming from conquest rather than their actual goals points.:knight:

[EatYerGreens]

On hard or expert I see most of the point leader’s points coming from conquest rather than their actual goals points.

Yes, that's probably why I'm in the lead (just) at the moment :knight:
- I'm the only one bothering to meet the goals. ~D

Then again, I fully expect to get severely clobbered in the homelands (ouch!), in the next few turns.

It's a shame that the Crusade goals are now unachievable - 1192 is too late for an overland attempt, starting in Franconia. Somewhat academic, anyway, since the Pope declared war on me and now forbids me to launch one, even though I was re-communicated. Something very screwy about this particular campaign.

[Budwise]

Yeah, I know of a game where I was playing the Danes and took over almost all of Germany. They decided to ally with me later and expanded east but left me alone with all of their homelands and I stayed allied to them for over a century. Stupid AI sometimes.

[antisocialmunky]

Strategic AI is imfamously hard to program and thus is usually stupid. For example, the AI rarely guarding Constantinople resulting in my crusade landing on it and me getting 10k from selling off all the improvements.

[ajaxfetish]

Well, at least you can still use those crusades you've got ready for goals like drang nach osten (assuming the pope eventually comes to terms with you, of course).

[Weebeast]

The Turks just attacked me for having Greece. My empire was always in the same financial condition. I was furious and curious so I checked the GA panel and it turned out that The Turks had to conquer ex-Byzantine provinces. Although it's conquest related, Turks still aim for GA. LOL

You're right though. Most of the time the other factions just try expand all over the map.

I played on high, normal and as Egyptians.

[EatYerGreens]

The Turks just attacked me for having Greece. My empire was always in the same financial condition. I was furious and curious so I checked the GA panel and it turned out that The Turks had to conquer ex-Byzantine provinces. Although it's conquest related, Turks still aim for GA. LOL

Actually, I had gone and checked all the other factions' various goals and it's suprising how some of them have certain target provinces in common.

I posted elsewhere about the French having some of the Levant territories listed as 'Homelands' goals. Whilst this is true in terms of their starting position in High (or is it late?) era, I didn't realise this was the case, the first time I saw it. It looked somewhat unfair on them to appear to have 'lost' points right from turn one, with no opportunities open to compensate for this shortfall. If I'd have been playing the French myself, I'd have felt cheated - not to say confused - eg in that the HRE can score 13 points from Homelands and most other factions score a lot less than that, due to having that much less to be looked after.

Returning to the point, you get places like Edessa, for example, which are on the list of 'homelands' for both Turks and France. Anatolia is a Turk homeland goal but the game begins with it in Byzantine hands and so on. So 'starting provinces' and 'homelands' aren't quite the same thing.

For real oddness, a new goal recently appeared on the Almos' page, the one about over 30 buildings and improvements in Cordoba and Granada. It actually says this is regardless of who owns the province in the points year (just pray nothing gets smashed if it's invaded shortly beforehand).

[lugh]

In my experience, like yesdachi says, the AI kind of gives up on GA goals when it's put under pressure, but I often see the Egyptians speed building their mosque (usually because I ease the Turkish pressure on them by taking Anatolia and the Holy Lands).
France having Edessa and Tripli as a homeland goal always annoyed me. If I remember correctly the northern Frankish realm was much more Sicilo-Norman.