Tobold's Blog
Sunday, May 14, 2023
 
Age of Wonders 4 - We can always dream

Age of Wonders 4 is a great game, but not a perfect one. And so one can't help but notice things that one would wish that they would be better. Now there are a few things where I am pretty sure that at some point in time they will be improved, either by a patch, or by one of the 4 already announced DLC. For example, minor inconvenience in the UI, on the hero screen you can't quickly switch from one hero to the next. That is one of those things that will most probably be patched at some point. But what about the improvements that are unlikely to happen?

I think the biggest improvement I would wish for is the artificial intelligence on the strategic level (the tactical AI is already very good). It is currently possible for the player to game the diplomacy system so that the AI opponents don't declare war on you. Setting difficulty higher just results in the AI opponents having a lot more troops, which is not very relevant if they don't attack you, and just tedious if you attack them. Tedious because only 3 stacks from each side can participate in a battle, so an enemy with 9 stacks instead of 3 just means you need to fight him three times.

The AI is also fundamentally incapable of thinking more than 1 turn ahead. If a distraction appears, the AI will chase after it. The AI is good at amassing troops for defense, but totally fails at moving troops forward to attack. After now 78 hours in, I still have exactly zero times an enemy besieges one of my cities; which incidentally makes the whole construction of defense structures useless.

I can imagine that some patch or DLC adds a parameter to the realm creation which makes the AI more aggressive, or diplomacy more difficult. Right now you can only set parameters making diplomacy with free cities more difficult, not with AI factions. Less likely, but already implemented in other 4X games like Civ6, would be a system in which factions have an overall strategy like "expansionist" or "aggressive", which you could choose when setting up your game. What I would really wish for would be an AI that simply would be better at playing the strategic game, but that is unlikely ever to be patched in. I'm not even sure this is something we'll see in Age of Wonders 5. For comparison, the Total War series of games now has 15 games, and the AI still sucks.

Then there is a list of improvements where I don't know if they will be improved or not. That concerns mostly the water and underground areas. In AoW4, curiously, both ship combat and the 4X gameplay in the underground half of the map is less good than in previous editions of the game. Naval combat in AoW4 looks like it was designed and programmed by an intern, offering just a water surface with no obstacles; the animations are comically bad, with all units being displayed as identical ships, repeatedly ramming each other to attack. The underground areas of AoW4 have comparatively few features and resources; you can't even conquer the whole underground without moving to the surface, as areas frequently aren't connected in any way underground. There is a system of digging through earth, but it is half-baked at best; you can find treasures or monsters, but no permanent resources for your underground cities. Having seen both of these areas designed better in previous games makes me think that this would be possible, but I am not sure that Paradox will want to put resources into these improvements.

Comments:
Yeah AI for 4X games must be monumental undertakings. Total War tries to compensate with buffs to the AI factions so that despite their ineptitude at the grand strategy part of the game they can still crank out armies.

I think 4X games are just far too complex and the tech just isn't there for a dev to make a realistic AI faction.
 
But if AI is going to take away all our jobs, while at the same time being unable to play our games, what does that say about the quality of our jobs?
 
I'd say there are a number of jobs with less decision making required and variables to deal with then a 4X game.

Although to be fair it would then question whether developing AI for those types of jobs would even be worth it. Why spending R&D money adapting some AI to take orders at McDonald's when a scripted app or website already does the task efficiently.
 
Post a Comment

<< Home
Newer›  ‹Older

  Powered by Blogger   Free Page Rank Tool