I still think you are somewhat overestimating the abilities of human players to adapt.
Of course you could potentially figure out some ways to circumvent the new costs, but the thing you seem to forget is that most humans play for fun and human players get bored. I'd rather doubt that many players would change their play style too much, since the fully aggressive (and costly) way just seems to be the most interesting. So a change would, excluding those "I just want to win and will use every wired, cheaty and boring trick" players, would rather benefit the AI, since it has the ability to restrain itself.
1. We are not talking about just one human, we are talking about the aggregate adaptability of the forums, reddit, livestreams, Google, etc. Consider how things have changed for PUs. It used to be that PUs were easy for everyone to get with minimal prereqs. So strategies developed to make the most of it. Rules changed and one or two players figured out how to manage game of thrones style play. Everyone who wants to can find a decent shot at playing the PU game without needing to figure everything out themselves. The AIs will only engage in PU seeking behavior if the devs dump a decent bit of man-hours into it.
2. Don't be silly about mitigating the costs. People will do all sorts of stuff to rack up bonuses. DDRJake created a OPM China, bankrupted it, and then had crazy mad bonuses. No one can predict how humans will adapt. The obvious trick is to just merc it up, but you also have options with Allies, Rebels, and HRE mechanisms.
3. It is a strategy game, this isn't a button masher where people just fire and forget, a LOT of people look for the most optimal way to manage things. I mean heck after every patch there is some thread showing that somebody has gone through the trouble of optimizing an Ottoman build for WC. Changing the incentive structure (e.g. rewarding manpower conservation) will change player behavior - that is what we call
strategy. The AI will only change its behavior if we program it to.
Of course players change, we do it all the time. It used to be that people would make a mad dash to release all the big deceased states (Persia, Ukraine, Guyenne, Shan, etc.) as their vassals then let rebels top off their cores). We do not do this now. Instead we pick one or two good targets to vassal feed and stock up on those. It used to be that people would chase fleeing, broken stacks across empires to shatter them, then carpet siege the place, that has changed as well. Of course people will change their play, they have done so after
every patch.
I also think that it probably wouldn't be too difficult to adapt the AI to that system. Making it more willing to peace out (removing the stupid length of war modifier, as well as some of the pointless -1000s) would probably do.
You are wrong. As a professional strategist I can assure you that the system you propose is among the harder types of things for AI programming. You have created a positive feedback loop where going from efficiently winning a war to inefficiently winning a war will rapidly slide into ineffeciently losing a war and rebel/vulture AI problems. It is all a downward AI spiral.
Say the AI starts losing a war, now it has less money, it can afford fewer mercs, so it uses more manpower, so now it has less money -> fewer troops. But that is not all, suppose the AI starts losing control of territory, it then gets increased WE. This in turn drops revenues and increases manpower expenditures (via revolts) as well as gold expenditures. Are we done yet? Nope. The AI is coded to trade stab away when it is heavily gold constrained so it is more likely to drop stab. How does it recover? All the other stuff you nerged - WE reduction, stability increases. Say you have AI Russia with Time of Troubles going and the Polish AI sensibly declares war. Manpower tanks, gold follows, that leads to even more manpower burn, with resulting higher WE and higher rebel counts. When does the AI pull out?
Positive feedback loops require heavy pattern recognition skills - that even though things have barely begun to head south, we need to very quickly stop the reinforcement cycle.
As far as making the AI easier to peace out, Wiz has pretty directly said he does not intend to do that. And getting rid of some of those -1000 will bring back the AI problems that got them put there in the first place. After all, your whole premise is that the more resources the player is burning, the more the AI benefits. A lot of those -1000 are to keep the player from getting low cost wars.
.png){kind=avatar}
