AI encouragement events

[unmerged(16323)]

Yes I know the "AI cheats aren't the way to go" philosophy.
I've read the Hands-off thread in which the Daywalkers AI cheat events are tested. I'm not a fan of enforcing things just because they happened in history.

However we all know that the in-game Artificial Inteligence lacks.. inteligence. Since the game is playable only at the Coward level, the AI DoWs, but the wrong guy.

I've just got fed up with seeing a well developed Russia (having all of their historical European holdings) stop their expansion at Orenburg just because they are scared of DoWing the Uzbeks or any other steppe country. Or a well developed Ottoman Empire spreading from Croatia to.. Konya beacause they are scared of DoWing Egypt or just too dumb to take out Ramazan first. Or the Spanish. Sitting around doing nothing while they know perfectly well that Aztecs are still pagans.

I thought about it lately. Yes, events like "Russia conqueres the Golden Horde" (with inherit) are out of question. But what about events that encourage the AI to do ANYTHING about their nearby enemy.

But what about events that force teh AI to go to war if they have a stable and strong country? For example:

event = {
	id = <event>
	trigger = { 
		ai = yes 
		OR = {
			atwar = no
			war = { country = TUR country = MAM }
		}
		land = 5 # Technology superiority
		stability = 1
		neighbour = MAM
		
	}
	random = no
	country = TUR
	name = "Sultan Selim I prepers for war (AI)"
	desc = "AI Only event"
	style = 2
	date = { day = 1 month = january year = 1514 }
	offset = 30
	deathdate = { day = 1 month = december year = 1524 }

	action_a ={				#Ok#
		name = "They shall fall!"
		command = { type = war which = MAM }
		command = { type = treasury value = 200 }		
		command = { type = badboy value = -5 } # To prevent other contries from interfering
		command = { type = INF which = 476 value = 10000  }	# Angora
		command = { type = CAV which = 476 value = 7000  } 	# Angora
		command = { type = ART which = 476 value = 20  } 	# Angora
		command = { type = warships which = -2 value = 5 }
		command = { type = transports which = -2 value = 10 }		}
}

The number of troops and cash should be appropriate for the situation. In this case it's a major war, thus the aid is quite significant. But if the OT was to attack a 1 province minor, there would be no money or troops at all.

I made a bunch of events for the OT and they work just fine. They no longer sit around with thousands of troops doing just nothing. What do you think?

 

[Cagliostro]

it's "prepares". I don't disagree with the notion of helping the AI in this way, but why the money?

 

[unmerged(16323)]

it's "prepares". I don't disagree with the notion of helping the AI in this way, but why the money?

Eh... my bad english

There is no way of checking how much money there is in country's treasury. It would be unwise to force the AI when it's coffers are empty. Therefore a small amouth of money needs to be given, so it won't take loans which would only cripple it further.

And we all know that AI wastes money and wastes troops, so (in this case) the aid given to the Ottomans is nearly nothing.

 

[Khephren]

Ummm....this looks Daywalker-esque

 

[Archaalen]

Simply being similar to some of Daywalker's ideas doesn't make it a bad one. The main problem with Daywalker's for me is the AMOUNT of cheating the AI gets in these situations. I don't really see the use in giving out ships for this particular war, but that's a sidenote.

 

[Norrefeldt]

I think we should try to get new AI's first. We could make some important coutries switch AI's if they have low enough badboy and raise the warmonger value and make a specific hit list for them. The AI's we have now are in many cases sub-optimal and could certainly be made better.

 

[unmerged(16323)]

Simply being similar to some of Daywalker's ideas doesn't make it a bad one. The main problem with Daywalker's for me is the AMOUNT of cheating the AI gets in these situations. I don't really see the use in giving out ships for this particular war, but that's a sidenote.

Yes, mayby the warships are to mcuh, but transport ships would be good. The AI tends to gather troops in Smyrna and tries to ship them to Egypt instead of going all the way through Syria.

I think we should try to get new AI's first. We could make some important coutries switch AI's if they have low enough badboy and raise the warmonger value and make a specific hit list for them. The AI's we have now are in many cases sub-optimal and could certainly be made better.

Tweaking the AI files won't change much. The problee of just sitting around doing nothing also touches the Ottoman Empire in periods where it's warmonger setting is 80 or even more.

As I understand it (and tell me if I'm wrong), the problem is caused by the limited pool of DoWs the AI countries are allowed to make in the specified timeframe. This often causes that the most war-like countries don't fit in for their share of the "pool".

I don't opt for a "oh lets give them loads of troops so we're sure they succeed". It's more of "let's tell the Ai it's a good time to attack". That's all.

 

[Dawkins]

For me, this sort of event needs to be an absolute last resort.

Before this is resorted to, how about an event which simply triggers a war between the Ottomans and Egyptians if the Otts are above land 5 and at peace?. In my experience the main problem is that the Ottomans just dont get round to attacking the Egytians enough - if they did they could beat them quite easily so cheat events aren't required.

What this would do is simulate human intelligence, rather than give in game cheats. Humans know how important beating the Egyptians is, the AI doesn't.

 

[unmerged(16323)]

Before this is resorted to, how about an event which simply triggers a war between the Ottomans and Egyptians if the Otts are above land 5 and at peace?. In my experience the main problem is that the Ottomans just dont get round to attacking the Egytians enough - if they did they could beat them quite easily so cheat events aren't required.

What this would do is simulate human intelligence, rather than give in game cheats. Humans know how important beating the Egyptians is, the AI doesn't.

That's exactly what I'm proposing. The army and cash is there only to be sure the AI has resources to figh the war. As I said, simualating human inteligence was my primary idea, since, just like you, I got fed up with the OT letting Egypt live by it's own. (and not only the OT, other countries as well)

However, there is no way of checking how much money or troops the target (of the event) country has. So forcing the AI to war without it having money or troops at all wouldn't have much to do with "human inteligence", would it?

 

[Dawkins]

No, but as I said,

--In my experience the main problem is that the Ottomans just dont get round to attacking the Egytians enough - if they did they could beat them quite easily so cheat events aren't required--

Lost of people really don't like country specific cheat events like this. I would suggest that in this case they aren't even needed, since by 1500 the Ottomans can easily defeat Egypt - if only they would try and keep going till they win!

Even if the Ottomans start out with few troops, they will quickly build enough once at war.

 

[unmerged(24512)]

I'd love this... i usually help the AI by triggering their inheritance events manually(like when TUR inherits MAM), and triggering good random events for them. This is to increase the competition from the AI in the late game. I would love it if i didn't have to do this.

 

[unmerged(16323)]

No, but as I said,

--In my experience the main problem is that the Ottomans just dont get round to attacking the Egytians enough - if they did they could beat them quite easily so cheat events aren't required--

Lost of people really don't like country specific cheat events like this. I would suggest that in this case they aren't even needed, since by 1500 the Ottomans can easily defeat Egypt - if only they would try and keep going till they win!

Even if the Ottomans start out with few troops, they will quickly build enough once at war.

Allright, but it seems the main quarrel here is about the additional troops and money. They're just an addition, for the reason I stated above. If they don't fit in the AGCEEP philosophy, fine, let's drop them.

I didn't want to start another "AI cheat events" thread. The idea behind that event is to make the AI actually DoW at the best moment instead of just sitting around. My idea wasn't about "powering up" the AI country, it was about enhancing the Artificial Intelligence, because it just lacks it.

After stripping the "cheats", just as you said, Dawkins, this becomes "event which simply triggers a war between the Ottomans and Egyptians if the Otts are above land 5 and at peace" .

Could we actually talk about the idea not the additional troops and money?

PS. Yea, I know, you may have problems understanding me It's because I sometimes have trouble putting my thoughts and opinions into proper English

 

[ForzaA]

I see one (earlier on quite often occuring) possible block to this event already: an alliance between TUR and MAM (As I understand it, this has been fixed largely) I do not know about military access, but I suspect this might block the event from working correctly, aswell.
Also, this event does not check for truces or vasalisations, and could thus incur a significant hit on Ottoman stability (with all sorts of negative side effects)


EDIT: possibly/probably, the Ottomans just need a load of "declare war" events (with sufficiently short timespans and sufficient triggers so that it won't occur ALWAYS) to guide them to glory.. or atleast more glory than they usually get.

 

[doktarr]

I'd agree with others that the troops and reduced BB should be unnecessary.

Like others, I'd perfer to see AI-switching to make the AI focus on this region at the appropriate time. The other key thing to do is to put Egypt in a stable alliance at game start (add another nation in Arabia or North Africa) and reduce TUR/MAM relations to 50 or 25, so that the disasterous TUR/MAM alliance is even less likely.

I find that one of the biggest problems is that the sheep often end up interposing themselves between TUR and MAM, thereby preventing the conflict. In this case, we want the Ottomans to agressively acquire the Mediterranean coastline, so they get the land connection back. Can we focus the AI this way? Will the "region" entry in an AI file only affect colonization, or can it effect wartime expansion as well?

The event, if needed, should have a trigger of being a neighbor of Egypt as well. And if you're actually interested in making sure the Ottomans have superiority, the right tech level to wait for is tech 9, not tech 5. With a CRT edge and a land connection, victory in war is all but assured.

The much bigger problem is that the Ottomans will take provinces (or even money) in peace before they trigger the inheritance. So if I were going to script an AI help event, it would be to change event 3371 (AI Ottoman inheritance of Egypt) to trigger on controlling three or more Egyptian provinces. Something like this:

	id = 3371
	trigger = {
		OR = { 
			control = { province = 746 data = TUR }
			AND = {
				owned = {province = 483 data = MAM }
				owned = {province = 485 data = MAM }
				owned = {province = 491 data = MAM }
				control = { province = 483 data = TUR }
				control = { province = 485 data = TUR }
				control = { province = 491 data = TUR }
			}
			AND = {
				owned = {province = 483 data = MAM }
				owned = {province = 485 data = MAM }
				owned = {province = 490 data = MAM }
				control = { province = 483 data = TUR }
				control = { province = 485 data = TUR }
				control = { province = 490 data = TUR }
			}
			AND = {
				owned = {province = 485 data = MAM }
				owned = {province = 490 data = MAM }
				owned = {province = 491 data = MAM }
				control = { province = 485 data = TUR }
				control = { province = 490 data = TUR }
				control = { province = 491 data = TUR }
			}
			AND = {
				owned = {province = 485 data = MAM }
				owned = {province = 491 data = MAM }
				owned = {province = 492 data = MAM }
				control = { province = 485 data = TUR }
				control = { province = 491 data = TUR }
				control = { province = 492 data = TUR }
			}
			AND = {
				owned = {province = 490 data = MAM }
				owned = {province = 491 data = MAM }
				owned = {province = 492 data = MAM }
				control = { province = 490 data = TUR }
				control = { province = 491 data = TUR }
				control = { province = 492 data = TUR }
			}
			AND = {
				owned = {province = 491 data = MAM }
				owned = {province = 492 data = MAM }
				owned = {province = 493 data = MAM }
				control = { province = 491 data = TUR }
				control = { province = 492 data = TUR }
				control = { province = 493 data = TUR }
			}
			AND = {
				owned = {province = 492 data = MAM }
				owned = {province = 493 data = MAM }
				owned = {province = 494 data = MAM }
				control = { province = 492 data = TUR }
				control = { province = 493 data = TUR }
				control = { province = 494 data = TUR }
			}
			AND = {
				owned = {province = 492 data = MAM }
				owned = {province = 493 data = MAM }
				owned = {province = 495 data = MAM }
				control = { province = 492 data = TUR }
				control = { province = 493 data = TUR }
				control = { province = 495 data = TUR }
			}
		}
		ai = yes
	}

This addresses a specific AI shortcoming - the failure to go for total victory even when it has the capability to do so.

 

[unmerged(16323)]

Of course I didn't mean only that specific case. It could be used for other countries as well. For example Russia, controlling everything west from Orenburg, just like they should, but not attacking SIB or UZB because they got scared of the wide and open spaces

As for the Ottomans, I also thought about changing that AI event... Don't know if it's historically justified.

Of course we could add events for OT-Venice wars, since OT rarely attacks and Venice often keeps their East Med possessions till late game.

 

[Dawkins]

After stripping the "cheats", just as you said, Dawkins, this becomes "event which simply triggers a war between the Ottomans and Egyptians if the Otts are above land 5 and at peace" .

Could we actually talk about the idea not the additional troops and money?

Sorry, I didn't want to come across as overly critical of the whole idea - quite the opposite

I think AI push events like this are an extremely good idea in lots of those frustrating situations where the AI just doesn't realise what is good for it!

 

[doktarr]

Of course I didn't mean only that specific case. It could be used for other countries as well. For example Russia, controlling everything west from Orenburg, just like they should, but not attacking SIB or UZB because they got scared of the wide and open spaces

Again, good AI design should help here. The new events should also help. But honestly, with the new trans-ural link being implemented it's no longer absolutely crucial for Russia to grab the Uzbek lands anyway.

As for the Ottomans, I also thought about changing that AI event... Don't know if it's historically justified.

Well, not really. The collapse of the Mameluke empire was caused by the capture of Cairo. As I said, this is more in the interest of addressing a specific AI shortcoming.

That said, the Ottomans did incorporate Syria into their empire while the war was still going on. So perhaps an event should be added that cedes all Egyptian territory north of Samaria as soon as The Ottomans capture it? This way, it shouldn't take more than two wars to trigger the event.

We could also add a "Ottoman/Mameluke tension" event late in the 1400s, which would be historical. These two events, combined with good AI-switching, may make coding AI-push events unnecessary.

Of course we could add events for OT-Venice wars, since OT rarely attacks and Venice often keeps their East Med possessions till late game.

I don't think it's necessary. Venice managed to hold onto Ionia, Corfu, Crete, and Cyprus for surprisingly long in real life as well. If they don't lose them until very late, I don't think that's distressingly ahistorical.

 

[Serk]

That said, the Ottomans did incorporate Syria into their empire while the war was still going on. So perhaps an event should be added that cedes all Egyptian territory north of Samaria as soon as The Ottomans capture it? This way, it shouldn't take more than two wars to trigger the event.

I agree, something like the french events for the hundred year war (Rouen is lost!, etc.) would be a good start.