As we approach the release of the beta many of the game's systems and mechanics are being wrapped up. What has been nagging me the last couple of days however is the way the AI will turn out. Having read the AARs posted by Brad I am concerned about the ability of the AI to handle the growing complexity of the game.
Some spells, like raise/lower land have left me wondering whether the AI will be able to use them effectively. A spell like raise land is "multidimensional". A player can use it to attack or defend (to block or funnel enemy troops), deny or secure resources or simply erect a barrier to block off a part of the map for the rest of the game. IMHO I don't think any amount of scripted or quasi-scripted AI work will give the AI the tools it needs to use this spell effectively. The spell simply has too many variables each of which can drastically affect the game. I assume this problem isn't unique to raise/lower land. There are many spells which, in the team's quest to rid FE of WOM's unimaginative and boring spells, will simply be out of the AI's reach when the game is finally released.
I have a feeling the AI will play a large role in FE's success. How many times have we seen a game with awesome game systems and never-before-seen features fall flat on its face because of a crappy AI? I really don't want FE to share this fate. What makes me worry even more is that, apparently, Brad is the only one working on the AI. I do not doubt Brad's programming skills, but I think AI is one of those things that needs a team to turn out well. FE has a lot of promise but the strength of the AI will ultimately determine whether the game will receive an enthusiastic or lukewarm reception.
I think it's safe to say at this point we are ALL worried about AI. But, I think you are wrong to worry about how the AI will be judged from a beta. Imho, anyone who judges a game based on a beta isn't worth listening to.
My point is this: we all worry about the AI, but what really matters is the AI at release, not AI at beta.
AI is the one aspect of the game that can be easily improved even after release. Brad has often said that the GalCiv AI went through a lot of improvements after the release of the game. It is much easier to improve the AI after the game is in open beta or released, as you get tons of more feedback on AI behaviour from players.
I'm not too worried about the AI expect for exactly the Raise / Lower Earth spells. I don't even like the gameplay concept of modifying the world map, so I would hope to get rid of these spells altogether. I am fairly certain that they are the most complex thing that the AI has to handle in FE and the part of the game that is most open for abuse.
The AI currently sends all units to 0,0 each turn. But the animation is smooth. What more do you want?
There's gonna be the greatest battle in 0,0. We call it The Battle of 0,0 or just The Battle. (I'm from the future.)
I'm worried about the AI given the ultra-complex nature of fantasy strat games.
Making an AI for GalCiv is probably a lot easier then an AI for a game such as FE. That's not a knock but the variables are very different and more numerous in FE because it's a more complex game militarily with all the spells and interacting statistics, not to mention tactical battles, which I'm suspecting will be pure hell to do sufficiently well enough.
I think Brad's going to be working the hardest out of everyone once this hits beta and we learn how to exploit stuff.
Remember the world, apart from the starting position, is a lot more dangerous now. So that should help things a bit, as you can't just focus all your energies in wiping out the AI before it can get started like you can in WoM.
im hopeing the world is filled with fantasy content. an alive dynamic world with events and nice graphics... with that in mind i agree that the ai can always be moddified and improved over time. what i might want to see if a player assaults a wildland area and fails that wildland area will start periodically sending out units to attack said player.
While I think setting the AI to counter a raise land spell is easily within FB's grasp of AI writing (If Human casts Raise land to block off a capital, then Computer Player counters by casting the opposite spell on the same tile.), I am thinking of trying to lower land my way through a wildlands to get to the gooey center.
Oh the exploits we will come up with...
Confetti. Lots of confetti.
That kind of a counter is called scripted AI, and it's something that Frogboy doesn't want in the game. That's the whole root of this problem; modifying the world map is such a multi-dimensional ability that it is extremely difficult to calculate optimal gameplay decisions with it. Even us humans are not able to predict the outcome of using such spell well, as it affects so many things wherever you cast it. What chances do you think the AI has?
That's why the more I think about it the less I want the spells in the game.
Lol, I like to worry about things that haven't happened yet. It's productive!
Agreed. I never liked how these spells were presented as low level utility spells, it's just wrong from both a gameplay and an immersion perspective. Spells like lower mountain might be fun if they are both expensive and high level, so that when it is used it is something impressive.
Yes. I agree that if the Raise and Lower, Destroy Land etc. spells are in the game they should be high level and expensive. Terraforming should be tough work!
might be neat & add to the feel of the game if lets say two or more sovs are trying to raise/lower/destroy land & a counter is cast on it for example, there could be a small chance that the magiks will 'combust' causing one of those great chasms that dot the landscape from the last epic magic battle, a smaller chance to unleash some kind of demon or it resurrects one the great beings that have fallen in tha past (or creats a new one), or a node is formed?
Well, if my terrible memory serves me right, in the older versions of WoM, you had to finish a quest to get the raise mountain spell. Any spell that is possible OP should have some degree of difficulty both obtaining and using. I'm sure Kael is way ahead of this.
The AI may also be capable of using the lower land spell to clear a path to any object of value. That would be how I would solve it with this kind of non scripted AI. The only thing I don't think the AI can or should do is try to wall in a human player. It would just be a waste of time. Although, it would be pretty cool to have a terraforming war that leaves the earth shattered and broken.
The point that these spells should be high up there and costly is absolutely right. I would say higher than forestwalk, but lower than teleport.
I would like to second this notion.
There are many great features available to you once you register, including:
Sign in or Create Account
