OHgamer said:
Agree with Mowers the model was totally wrong.
If Poland colonizes the New World before Spain the model is wrong.
AI-controlled nations and ahistorical potentials
- Thread starter OHgamer
- Start date
-
We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
Agree with Mowers the model was totally wrong.
If Poland colonizes the New World before Spain the model is wrong.
das said:
Amen!
And please people, the answer to EVERY SINGLE QUESTION is not, "I want that to be an option!".
I swear, just turning the game on would lead to 453 option toggles if Johan followed the advice of the forum. Each of which he'd need to code support for.
Exactly.das said:
I also like what Hive said about being a mix between (unmodded) EU2 and Victoria. As much as I like the AGCEEP and VIP mods and appreciate the incredible amount of work that has gone into their creation, they are simply too deterministic for my tastes. One of the most interesting things for me is to see history unfold differently each time I play. Now that doesn't mean I want to see the same ahistoric things happen each time either, like the 16th Century Chinese Superpower, but seeing odd behavior like Ottoman India in one game and Brandenburg replacing the Austrians as the Central European monster in the next keeps things interesting.
First, I think that the AI should definitly respond ahistorically to ahistorical circumstances that impact its choices.
As for ahistoricallness in general, I don't mind ahistorical behavior so long as the median case remains historical. Russia colonizing africa every so often is fine, its only when Russia predictably colonizes Africa that its problematic. Likewise, a chinese superpower arrising every so often isn't problematic, so long as it doesn't happen most of the time. It would be nice if any given nation ended up in a place roughly resembling where it was historically after 150 years every 2 out of 3 times I'd say, with nations affected by the player diverging more.
As for ahistoricallness in general, I don't mind ahistorical behavior so long as the median case remains historical. Russia colonizing africa every so often is fine, its only when Russia predictably colonizes Africa that its problematic. Likewise, a chinese superpower arrising every so often isn't problematic, so long as it doesn't happen most of the time. It would be nice if any given nation ended up in a place roughly resembling where it was historically after 150 years every 2 out of 3 times I'd say, with nations affected by the player diverging more.
Smirfy said:Agree with Mowers the model was totally wrong.
If Poland colonizes the New World before Spain the model is wrong.
Aye. But if Poland manages to get a few colonies in the New World, the model is interesting, not wrong. If that happens in about 1% of your games that is.
hjarg said:
That example is plausible, obviously not very but its plausible all the same. Obviously Poland taking the whole new world by 1580 is not plausible.
Thats what is at stake here, what is plausible or not and how to model the relative scale of plausible actions. This all comes down to the base model at the end of the day.
Just like its plausable for Portugal to ignore Brazil for Say Africa or India or somewhere else.Mowers said:
It's probably better to try and recreate the factors that explain why certain nations did what they did that rely overly on scripting outcomes, so the factors are linked up with eachother, in such a way but still allowing for those factors to be changed, either by the AI deciding on a different solution or on human player intervention. I would make the game less warlike though except for some particularly warlike civilizations, like the Ottomons and there should be more opportunities for peaceful building and trading and these should be more evenly weighted against warfare in victory point terms.
hjarg said:
Unless, of course, Poland dominates the Baltic coast and pushes west as far as Bremen or Hamburg. With access to the Atlantic, all bets are off.
Then, I'd be happy to see them colonize. Although British Isles and Gaul/Iberia should be primary colonizers simply due to access to the ocean and possibility of focus.
Poland shouldn't be colonizing unless it is "secure" in its borders, which is not at all certain based on their neighbors.
What you guys are saying sounds to me like having a few AI files types. That is a number of them with roles, or manifest destiny (you say that in english?), like for instance the "naval monster", the "colonizator", the "CoTs collectionist", the "african ruler", the "italy focused", the "Anti-heretic"...and then an AI for the historical path of each nation, which will mostly apply but it might happen that on starting a game an AI role could be randomly chosen instead of the historical one. How much random? Well, I guess that was the point of the first question in the thread, so, we are like in the beginning again .
No, what you think of this idea? This would work only for big countries, of course. Or another idea that is coming to my mind for small countries, define a country-size limit and if it's reached a random AI is asigned at that time. I mean, we are talking about 400 years. That looks to me like a long time to be doing the same thing, ain't it? The AI could change sometimes during the game. Like for instance, Hesse getting big enough at mid-XVII-century, getting to the coast so a AI is chosen to her keeping in mind (or RAM if you want) that Italy is too far, or unskilled king for trade (doesn't like CoT's targets), but a coastal province open new unhistorical oportunities.
sounds fun to me. I think though that historical games should be an available option anyway.
.No, what you think of this idea? This would work only for big countries, of course. Or another idea that is coming to my mind for small countries, define a country-size limit and if it's reached a random AI is asigned at that time. I mean, we are talking about 400 years. That looks to me like a long time to be doing the same thing, ain't it? The AI could change sometimes during the game. Like for instance, Hesse getting big enough at mid-XVII-century, getting to the coast so a AI is chosen to her keeping in mind (or RAM if you want) that Italy is too far, or unskilled king for trade (doesn't like CoT's targets), but a coastal province open new unhistorical oportunities.
sounds fun to me. I think though that historical games should be an available option anyway.
I believe that the AI should ALWAYS be historicaly.
Except when yourself do something ahistorical which should make the AI act ahistoricaly to counter your actions.
Except when yourself do something ahistorical which should make the AI act ahistoricaly to counter your actions.
alvaro, i think this is a wonderful idea. Say 40 preset AI strategies that it randomlly selects every 30-50 years but avoiding those that are unplausible with a few basic checks - do i have a port? Do I have lots of ports? Is the a very large aggressive country next door with a large land force etc chuck in a random element youve a great game
Skarion said:
I think some randomness helps the game to not to be the same game, always. But only if it's a logic randomness. mmmm,these too words don't fit together very well, do they?
come on, if you have the chance to choose (I also prefer historical behaviours), sometimes you feel like trying crazy things just for the fun. or you never had a look of the "weird EU2 screenies" thread? come on, confess!!
Skarion said:
The question is how do you design an AI to act ahistorically to a players ahistorical behavior while remaining within relatively historical frameworks. Or does having the AI behaving within a historical framework make it much more difficult to react effectively to ahistorical behavior by players when it happens?
jorian said:
And Brandenburg did have tradeposts in Africa and IIRC also in South Asia in the eighteenth century, at least for a period.
Also a brief attempt at a trading company to China, not a huge success.OHgamer said: