A Plea to CA for a new engine instead of a soon next Total War game. DX12 may deal the cards anew for Rome 2.

[alQamar]

By the way, as it turns out, most battle animations (as long as you don't count the 1v1), sound and unit models are perfectly fine. The problems are with the gameplay (among a lot of other things). If only CA didn't refuse to fix that, we would have a gem of a game right now. They can still do that, they already have the models and sounds and animations ready.

But I would say, they should create a new engine and patch the game with it, while importing the good bits of the game (especially models, ambient sounds and animations) into the new engine. But that's impossible now I think...

Very impossible to patch an existing game. I am very confident that the next TW will not use melee units but gunpowder units again so the flaws of the old engine aren't that obvious.

[Rittsy]

Patch the game with a new engine??? That would be an incredibly expensive and time-consuming patch for CA. Next Total War release maybe, but Rome II is and always will be a Warscape game.

[=Vastator=]

Patch the game with a new engine??? That would be an incredibly expensive and time-consuming patch for CA. Next Total War release maybe, but Rome II is and always will be a Warscape game.

And so will be the next (Warhammer:TW).
In the "Julian McadumbAI" video he was using a tool, and there was a folder for Shogun2, Rome II, and Warhammer...
Plus, the warscape engine seems a very good one for cheating people with awesome graphics and animations (only for promotional videos). Why change it?

[Fallen Phalangite]

I didn't vote, because if Rome 2 on this engine is garbage, then the next Total War on the next engine would be an outright disaster.

[alQamar]

I agree, but how do you think they will fix this game or release an unbroken game on this engine, which they failed to do since Empire? Where should the chance of a change be? The Warscape engine has several limitations, very well known, most in case of pathfinding and AI, that is why the heck we face an unfinished game every time, and Shogun 2 was no exception to that rule at launch.

I hope this could convince you that a startover may be the best chance to settle things for the better, or let the TW series behind all possibilities.

[Fallen Phalangite]

I agree, but how do you think they will fix this game or release an unbroken game on this engine, which they failed to do since Empire? Where should the chance of a change be? The Warscape engine has several limitations, very well known, most in case of pathfinding and AI, that is why the heck we face an unfinished game every time, and Shogun 2 was no exception to that rule at launch.

On the one hand, you may be right in saying it's the fault of the engine that's causing the game to have such dreadful AI. However, it's the programmers who have repeatedly failed to live up to both our expectations and the claims of their company.

How can you produce 5 major games (including FotS) and still not improve on the AI? Is it the engine? Is it the programmers? Who knows? It's easy to say "get rid of the programmers, and if the new batch can't do any better, scrap the system and get a new one", but it's much more complicated than that. When in doubt, though, I usually point the finger at the people behind it, rather than the system itself. At this point in time, there's really no excuse to still not being able to produce a half-decent AI despite the limitations of the engine.

[krisslanza]

On the one hand, you may be right in saying it's the fault of the engine that's causing the game to have such dreadful AI. However, it's the programmers who have repeatedly failed to live up to both our expectations and the claims of their company.

How can you produce 5 major games (including FotS) and still not improve on the AI? Is it the engine? Is it the programmers? Who knows? It's easy to say "get rid of the programmers, and if the new batch can't do any better, scrap the system and get a new one", but it's much more complicated than that. When in doubt, though, I usually point the finger at the people behind it, rather than the system itself. At this point in time, there's really no excuse to still not being able to produce a half-decent AI despite the limitations of the engine.

AI building is incredibly difficult. Even with all our technology we can still only have an AI compete with a human being in a game that is entirely logical. A game like chess, a computer can play a human to a standstill, if not victory because it can simply run every possible simulation.

But in a game as complicated and complex as Total War? Not a chance. AI technology isn't anywhere near the level of making it actually "think", it can only do pre-programmed tasks that can then be modified by any variables. You aren't going to see a very complex and engaging AI until we can actually create an AI that can think like a human being.

[Fallen Phalangite]

AI building is incredibly difficult. Even with all our technology we can still only have an AI compete with a human being in a game that is entirely logical. A game like chess, a computer can play a human to a standstill, if not victory because it can simply run every possible simulation.

But in a game as complicated and complex as Total War? Not a chance. AI technology isn't anywhere near the level of making it actually "think", it can only do pre-programmed tasks that can then be modified by any variables. You aren't going to see a very complex and engaging AI until we can actually create an AI that can think like a human being.

I never said I expected human-like intelligence. What I do expect is some degree of sound thought-process from the computer. Such as "Hmm... the defence is outnumbered. Maybe we could win this siege if we attacked from several directions at once, rather than charge for the nearest open gate."

[Yojimbo]

Warscape is a horrid unoptimized engine completely unsuited to large scale engagements. It needs to be dumped as soon as possible.

[Aenima]

I would gladly have waited another 6, 12 or even 18 months if it meant Rome 2 was released on an engine which the AI understood (basically that battles went smoothly and lived up to our expectations, instead of being blob-pits and siege battles where the AI doesn't understand what to do)

The sad thing is that I don't know if the game can be fixed through patches and mods to make it perform like we want. Are we always going to just have to put up with the AI standing around outside walled cities, and forever endure the super-tight quarters of the blob-pits? Is it possible to mod/patch these things out of the game, or are they inherent to the engine? If it's the latter, then I don't think I'll ever be able to enjoy this game.

[Davincee]

I dont even have a 64 bit, but im definitely willing to sacrifice something just to see my favorite genre of gaming coming close to my dreams again. Ill upgrade.

Signed.

[wangrin]

AI is "only" a series of rules or equations to decide, depending of parameters (entries), what to do.
For example, when to use siege equipments and what path to follow...

Of course, the more you have parameters, the more complex are the combination of rules... and the more chance to create weird issues you have.

So, it is possible that warscape engine architecture led to difficulties to handle some mechanisms such as siege for example.
But, we can only speculate here, as we don't know how, exactly, the game engine work or even is built.

Facts are :

• siege are a problem since the first game using warscape engine
• CA cancelled siege equipments up until TWR2 (sap_point, ladder and petardier seems to have been planned for ETW but never use, you can see it in battlefield_deployable_siege_items tables and you can find models.

But siege are not the sole problem...

[omzdog]

But, we can only speculate here, as we don't know how, exactly, the game engine work or even is built.

There is one perspective where we have concrete evidence of the capabilities of the engine where we do not need speculation.
That is the end result.

Much like in the idea that you know a car engine works because it makes the automobile move or works improperly due to inefficiency in long distance runs, Warscape can be judged based on what we see in real-time use of its scripts and logical versatilities through BAI.

I don't think anyone is saying that Rome2 is the pinnacle of Warscape, but fitting one of those old inline four engine 2L direct injection machines into a Ford Explorer (as evidence by taking out all the bulky siege stuff you were talking about as well as other features I won't bother depressing myself listing) isn't a good idea no matter where you want to go.

Not saying you judge an engine based purely on the car's performance, but its a great assumption to make if something went really wrong... or blew up... or Rome 2 was released.

Spoiler Alert, click show to read: 

Whenever someone talks about AI I'm always reminded of the realities of possibilities.
Everyone's pretty familiar with the fact that its impossible to map human behavior through bruteforce (scripting all possible parameters and coordinating their consequential reactions).
To call what we have in Rome2 intelligence - be it artificial - is an insult to sapient beings and inaccurate.
For the sake of building a possible game I'll ignore my impulses but if you read this entire post I want you to watch this youtube snip.

[VINC.XXIII]

Facts are :

• siege are a problem since the first game using warscape engine
• CA cancelled siege equipments up until TWR2 (sap_point, ladder and petardier seems to have been planned for ETW but never use, you can see it in battlefield_deployable_siege_items tables and you can find models.

Pretty interesting, do you have an idea what go wrong with the AI that don't use its towers, ram, ladders like in MTW2??

[General Maximus]

Warscape AI cannot learn to use siege engines - they are completely off the concept and working procedures of this trash engine.

In Empire and Napoleon, the usual cannon artillery did the job of siege engines to break the walls. To make the troops climb the walls, they gave us rope ladders, since the AI doesn't know what siege equipments like ladders are. It doesn't even know what those things are, let alone the talk of using it.

With Shogun II, they removed the rope ladders but allowed all the troops to climb like Ninjas. Castles are built in an entirely different way in Shogun II. And since walls in Shogun II's sieges mean nothing more than a cover from missiles, and walls-or-no-walls mean nothing since they are all climbable, there was no need for any wall-bashing siege equipment at all, cannons and onagers for example (even though they did got implemented and helped a lot). FoTS all the same, except that artillery is there not to bring down the walls but to kill the men standing inside the castle compounds.

Another thing they implemented was the burning of gates. This in Shogun II is realistic, since Japanese castles have wooden doors and gatehouses - earthquake shatters all arch-like structures (whether round or not) that have any stone structure above them, prompting the Japanese to build wooden structures that actually stood their ground - and wooden gates and gatehouses can be burnt easily.

All the problems they had hidden in previous Warscape games fell apart and got revealed in Rome II. They cannot give the ninja-climbing ability to the troops in this game, it would be too idiotic and unappealing to even the stupid 12 year old arcade crowd (apart from some totally uncaring stupid arcade kids). Nor is the Warscape AI able to learn to use the siege engines. The result? Only way the AI could find to get into a settlement was to burn the gates, so they retained that feature AND THIS IS WHY THEY REFUSE TO REMOVE IT. And since the AI was having problems at the walls, they decided to remove the walls for most settlements altogether.

CA went the lazy way, it's evident. Instead of actually implementing an AI that is capable of using those things, they just removed the walls and gave it the straight way through the gates.

In fact their strategy throughout this game is the same:

AI not advanced enough to handle the gameplay features? Oh yeah, too bad, let's dumb down the whole gameplay to it's level so that it is able to handle those features.

AI is too pisspoor to overcome obstacles? Too bad, let's remove the obstacles completely altogether.

*After saying that, CA employees do trollface smile*

[Humble Warrior]

Warscape AI cannot learn to use siege engines - they are completely off the concept and working procedures of this trash engine.

In Empire and Napoleon, the usual cannon artillery did the job of siege engines to break the walls. To make the troops climb the walls, they gave us rope ladders, since the AI doesn't know what siege equipments like ladders are. It doesn't even know what those things are, let alone the talk of using it.

With Shogun II, they removed the rope ladders but allowed all the troops to climb like Ninjas. Castles are built in an entirely different way in Shogun II. And since walls in Shogun II's sieges mean nothing more than a cover from missiles, and walls-or-no-walls mean nothing since they are all climbable, there was no need for any wall-bashing siege equipment at all, cannons and onagers for example (even though they did got implemented and helped a lot). FoTS all the same, except that artillery is there not to bring down the walls but to kill the men standing inside the castle compounds.

Another thing they implemented was the burning of gates. This in Shogun II is realistic, since Japanese castles have wooden doors and gatehouses - earthquake shatters all arch-like structures (whether round or not) that have any stone structure above them, prompting the Japanese to build wooden structures that actually stood their ground - and wooden gates and gatehouses can be burnt easily.

All the problems they had hidden in previous Warscape games fell apart and got revealed in Rome II. They cannot give the ninja-climbing ability to the troops in this game, it would be too idiotic and unappealing to even the stupid 12 year old arcade crowd (apart from some totally uncaring stupid arcade kids). Nor is the Warscape AI able to learn to use the siege engines. The result? Only way the AI could find to get into a settlement was to burn the gates, so they retained that feature AND THIS IS WHY THEY REFUSE TO REMOVE IT. And since the AI was having problems at the walls, they decided to remove the walls for most settlements altogether.

CA went the lazy way, it's evident. Instead of actually implementing an AI that is capable of using those things, they just removed the walls and gave it the straight way through the gates.

In fact their strategy throughout this game is the same:

AI not advanced enough to handle the gameplay features? Oh yeah, too bad, let's dumb down the whole gameplay to it's level so that it is able to handle those features.

AI is too pisspoor to overcome obstacles? Too bad, let's remove the obstacles completely altogether.

*After saying that, CA employees do trollface smile*

Absolutely right, on all points.

I`m wondering if our AI `programmer` is at least trying to fix the siege AI so it can use siege engines for Patch 6 and 7? I mean he got it to at least move the Towers and ladders to the walls, even if it completely forgets to carry things through.

The signs are bad of course: No word from anyone in CA like, "Were really working hard on the AI sieges. We`ll conquer this, we will."


I owe you and a few others about 5 Reps, but they`ll have to wait a bit!

[alQamar]

Absolutely right, on all points.

I`m wondering if our AI `programmer` is at least trying to fix the siege AI so it can use siege engines for Patch 6 and 7? I mean he got it to at least move the Towers and ladders to the walls, even if it completely forgets to carry things through.

The signs are bad of course: No word from anyone in CA like, "Were really working hard on the AI sieges. We`ll conquer this, we will."


I owe you and a few others about 5 Reps, but they`ll have to wait a bit!

Milkandcookies (Indypride) released a YouTube where he ask CA to fix the siege AI. He was runner-up in CA promotional tournament.

[alQamar]

Thanks for all your feedback so far, I hope we can keep this thread up.

[Bento]

Unfortunately, as many have pointed out, the fact that Rome2 sold and sold well means that CA have ZERO incentive to change their direction and/or build a new engine anytime soon.

New TW games will come out, it just wont be us buying them.

[General Maximus]

Considering that they are aiming for those stupid 12 year old arcade kids, who have no idea what words like 'fun' and 'immersion' mean and only care for quick and easy wins, this seems to be true. Now they don't even have the need to build anything.

But they will lose some sales, no doubt. Oh well, they don't care. They are going the COD way, just build and release arcade games in an endless tirade without much difference between all of them, that keeps their pockets full.