Battle AI – By Mike Simpson
Just before the end of Empire the lead Battle AI programmer left CA to return to his family up north. Unfortunately, thanks to Mr Wilberforce’s efforts 200 years ago we couldn’t stop him. It left us with a battle AI, which at that stage, struggled to beat good players in a fair fight, and was pretty much at the mercy of great players, even with a level of handicap (I call it cheating) that is all too obvious.
Since then we’ve had our most talented programmers pick up where he left off, but becoming the master of a chunk of code that took almost three years to write is not an instant thing. In the updates so far we’ve got rid of some of the worst behaviours that are close to the surface, and have started to tackle deeper issues like unstable decision loops that cause the AI to mill around rather than hold its line and fire. We’re also starting to add new behaviours, for example taking better advantage of hilly terrain. These improvements take the code further than we’ve been able before and will be there for Napoleon but we’re not sure yet whether we’ll be able to reverse them back in to Empire in a future update – the code has moved on. If we can we will.
Our overall aim with the improvements is first to get rid of any erratic behaviour, second to improve general performance to the point where the obvious handicaps can be removed, and then to add a greater variety of ‘smart’ behaviours. None of these have a fixed finish line – it’s a process of continual improvement, and each game will get AI better than the last one. Including Napoleon.
Progress is frustratingly slow but thankfully rewriting the Rome: Total War codebase has left us with a clean codebase that is easier to work on, and an architecture that has way more potential than Rome’s. The main difference is the shift to a goal oriented planning system rather than a static system that has no long term plans. This has yet to fully pay off. But it will. When it does I’ll talk about it again.
Battle AI is not rocket science – its way, way harder than that. It’s so difficult that very few strategy games attempt it. Most use simple scripts of canned behaviour that fire when you bump in to them, and very simple swarming behaviours. They’re limited, and are “gamey” rather than real world. What I mean by that is that the tactics you use to beat them are something that you have to learn for each game or sometimes each scenario/level, and bear no relation to reality. What we strive for is a game where real world tactics actually work. It’s not the easiest path to take, but it is the most rewarding.
If you’re a genius C++ programmer, you understand exactly how difficult this problem is, but still think it’s the most interesting code problem in the world, apply for a job. We’ll find space for you.
Good luck lads!
Posted by Henrik Herlev on November 10th, 2009 at 3:35 pm
I want the AI to be an effective oppononent when I buy the game. Continous improvement may be a goal of the developer, but why as a consumer should I care?
Posted by Turbo on November 10th, 2009 at 4:04 pm
Im stunned. As much as I understand your problems and certainly appreciate your honesty I am stunned that under these conditions the game would be published. I know we have little or no say in this since we all ready paid for it, but from a moral point of view you owe us, man. You owe us to fix this. Perhaps not as good as you promised, but at the least way better then it is now.
Again thanks for your honesty, but you owe us.
Posted by Legion on November 10th, 2009 at 4:14 pm
Well, having been a rocket scientist myself, not sure if i agree with you. I’ll say codding an AI is a different area of expertise but it’s quite simpler sorry. You cannot simply compare the knowledge and technology required to reach a new conclusion in say, gravitational waves, to the one needed to develop a game’s AI system. It’s two worlds apart sorry.
Anyway, I sure hope you can fix the BAI of Empire total war because as it stands now, it’s hardly a game and once the “pretty” effect wears off it’s dull as hell.
Also, where oh where are promised the Mod tools??? Why the silence on that subject? And don’t tell us they’re coming “soon”.
Cheers…
Posted by mAIOR on November 10th, 2009 at 4:46 pm
[quote]but we’re not sure yet whether we’ll be able to reverse them back in to Empire in a future update – the code has moved on. If we can we will.[/quote]
I appreciate you honesty mike but this really leaves a sour taste in my mouth. What you are doing is acknowledging that the BAI is FUBAR, but you are not sure if you are going to fix it even though its probably the biggest aspect of the game?
Right now the game is only fun for me if I would only do autoresolves because the battles are not just not a challenge, but more a bore. AI that doesn’t seem to know what it wants to do, running in front of its own lines, then moving the back lines again it looks like a mess.
If you guys are not going to fix this up in empire and are just going to tell us to buy Napoleon I really see this as a stab in the back to your customers, and I am 100% sure it will take a big hit on Napoleons and any future CA sales if you screw over your loyal customers like that.
I was really looking forward to this game when it was announced and still haven’t been able to fully enjoy it. I hate to say it but when the game was released it was an abomination and not finished, I couldn’t even play the game without crashing untill about 3 months later. You guys have come a long way since then and I really hope you keep up the work for empire because a lot of people still want to play this game and are still judging wether or not they should just flick it off their HDDs. So far it’s been on my HDD since release, now please don’t let it have been a waste of space.
Posted by Munkey on November 10th, 2009 at 4:47 pm
Why doesn’t CA establish a branch in the United States, Silicon Valley perhaps? It’s fairly obvious that the best engineers are in the United States.
Too bad the yanks don’t know how to make a good game.
Posted by Grand Poobah on November 10th, 2009 at 5:47 pm
Mike,
I can’t say I envy you in your position right now. Some things that you’ve said in your blogs have raised big concerns, while at the same time have brought a level of honesty that is surprising and admirable… if it didn’t cause so much trouble.
I do appreciate your driving on through the hailstorm that each new blog seems to bring. That being said, I really believe that you all should to the utmost to make BAI fixes ETW compatible. I can only speak for myself, though have read others who’ve expressed similar sentiments, when I say that about a month ago I decided not to purchase any more titles from CA if ETW is really going to be left in its current state.
I suspect a lot of this will have to do with the great SEGA overlords, but it would go a loooooonnnggg way towards making amends with many of your loyal fans if BAI improvements made it into ETW. I understand it’s hard, but so many of us were sold on improvements that never came, and now to find out that the lead BAI programmer deserted… probably knowing full well that things were left in a shoddy state.
I do hope you’ll do your best. If ETW gets further support, despite having been told that no major patches are coming out, then I’ll reconsider my stance on future purchases. If not, then I’ll stand by my decision. I don’t mean offense by this. It’s just the principle of the thing.
Thanks, and good luck!
Posted by Rich on November 10th, 2009 at 6:34 pm
I have to agree with many of the comments above – CA really MUST fix this problem and apply it to Empire. They cannot expect people to fork out another 40 quid to see the improvements in Napoleon. I’m not saying that all the additonal behaviours need to be included in Empire as it is right to see devleopments in the next game, but at the very least the erratic behaviour must be addressed. I just don’t buy it that the BAI improvements can’t be applied to Empire. They simply must otherwise CA and SEGA’s reputation will be damaged permanently.
Posted by James Walter on November 10th, 2009 at 6:42 pm
I’ve got to say that even though there were some major bugs in the release of Empire:Total War(and believe me, I got to experience them) it was and is a great game! Sure it was quite boring from time to time when the game crashed in the middle of a session but between the crashes, I enjoyed every second.
Since this blog was about the Battle AI I have to say that it was one thing that fell under the expectation line, although I did enjoy the easy battles at a start I figured it was because of my low battle difficulty but as I learned the new strategies and increased the difficulty it got boring.
But nonetheless I am as exited for Napoleon as I was for Empire, if not more. Even if you guys does not fix all the crash issues and bugs and other problems I know I will enjoy that game more then I have enjoyed Empire. The unique thing with the Total War series is that it always keeps getting better. Even if the previous game did not meet the expatiations, the next one will.
Mike and other Total war producers: Keep up the good work, it keeps me from being bored!
Posted by SWJoe on November 10th, 2009 at 7:45 pm
“but we’re not sure yet whether we’ll be able to reverse them back in to Empire in a future update – the code has moved on. If we can we will.”
So…. you expect people to buy NTW after your already hinting at not repairing the AI in ETW?
Is this some kind of joke? Do you really think the fanbase wants to hear excuses? I have a job too man and excuses dont fly where I work.
Posted by True Warrior on November 10th, 2009 at 8:19 pm
“If you guys are not going to fix this up in empire and are just going to tell us to buy Napoleon I really see this as a stab in the back to your customers, and I am 100% sure it will take a big hit on Napoleons and any future CA sales if you screw over your loyal customers like that.” -Munkey, Comment#5
Very, very true. This is exactly the kind of thing you have to keep in mind with the current state of the Empire project, and how you’re planning to release Napolean before you finish the game we have now. From what I got from this blog post, you’re current strategy is to fix it in Napolean and make us other loyal customers pay twice for the game we wanted (To summarize, ‘We have great improvements planned… for Napolean. The game will be more playable in future additions… Including Napolean. The code has moved on… To Napolean.’) Please let me know I’m wrong, by doing the entirely possible thing (we know what you guys are capable of), that is fixing the AI (and other bugs) in Empire. If you won’t, then I’m done with the Total War series… And I don’t think I’ll be the only one. Ask yourselves this; How many Empire sales were made to players of the original games, and what would the profit reports of Napolean look like without those numbers and with the bad publicity the long-respected internet voices who are disappointed by the current state of Empire?
P.S. I liked the Mr. Wilberforce joke, that was classic
Posted by Jeremiah on November 10th, 2009 at 11:01 pm
Yes, I do understand what a problem that is and I completely support you.
And I would so very much like to take you up on offer for the job. :’( I hope offer will still stand in a few years time.
Posted by Nemo on November 10th, 2009 at 11:03 pm
Ok two words….
Mod tools.
give us those and we will all buy NTW without feeling like we just got backstabbed
Posted by Jesus on November 10th, 2009 at 11:40 pm
I saw the game great potential even before it was released … After I’ve played it I realize it’s huge universe and I like it a lot (and today I’m still playing it ) … But I also realized how an unfinished game looks like… (not that I’ve not seen others)
Next time (with Napoleon version) do your homework better ! Because “you owe us” like someone already said !
Cheers…
Posted by Michi on November 10th, 2009 at 11:50 pm
Tough customers or loyal fans? I don’t really know what to make of this fierce criticism. All I know is it makes me ill.
Empire is the most fun I’ve had so far in the TW franchise, and I’ve been there ever since Shogun.
You guys are spitting on the hand that feeds you. Let’s all cut CA a little slack.
Posted by Celcio on November 11th, 2009 at 3:17 am
Im seeing a lot of comments about “not believing that its rocket science”, and reviews talking bad about the AI. Geuss what, new and revolutionary things are rocket science. If you look back at historys greatest creations, you see that the first prototype though it works, is nowhere near perfect. The empire AI is the same way. im not going to point fingers, but if you look at games like *cough* Age of Empires and *cough* command and conquer, when do you actually use serious real life tactics? In those games, its all based on how many troops you have and what their skill is, if you dont have good amounts of good troops your practically an outomatic looser. When i learned to play “a certain strategy game”, i couldnt use any of my genious military skills(if i do say so myself). Rather, i had to learn how to mass produce, and gather resources, and that took place of military tactics. Wow.
Trust me, im not a big fan of the empire AI either. In fact, sometimes it bores me to death as they blindly repeat the same stupid mistakes like clumping troops over and over. It can ruin the expirience of campaigns at times. However its as good as its gonna get for our time period, because real life tactics require real life people. Its impossible to give a computer that only follows orders the free will to make mistakes and be creative, and smart for that matter. Thik about it, to make this game truly realistic they would actually have to program George Washington’s Brain or General Cornawalis’s brain for a real battle. face the facts, you cant programm a brain with C++, so rather than yelling at CA for not making a human like AI, accept the fact that they stomp over any strategic AI ever created.
If you can program a brain using C++, please introduce yourself to CA. If you can’t, shut up.
(by the way, a save in battle would be nice)
Posted by Sam on November 11th, 2009 at 4:09 am
I am trying to find an instruction manual on line for this game – any help? thanks
Posted by jackfolk on November 11th, 2009 at 6:27 am
I’m stunned, as well. Yo me this sounds like a quality issue inside CA first and foremost. I have no reason to doubt Mr Simpson’s word, and I won’t, but this excuse sounds like “the dog ate my homework”.
CA should have appointed an assistant BAI programmer, or a qualified verifier to the lead programmer’s work. Or if the lead programmer was such an irreplacaple genius in his work, they should have paid him doubl and tripple to make him stay. It would pay off, if CA trusted that he would deliver an amazing BAI.
The comparison with rocket science is unfortunate. Maybe rocket science seems easier because the quality systems applied in such fields are way way failproof than waht is practiced by a gaming comapny.
All that said, I am still a fan of the TW series and I sincerely wish CA all the best in their efforts to improve their games in all aspects.
Posted by Deme on November 11th, 2009 at 9:24 am
What i would like to know is where is the multiplayer campain mode we where promist. And if possible could we have a multiplayer campain mode that runs on dedicated servers? So in affect all players could take there turns at the same time and then when they have all finnished the server then playes the AI. This could then be able to save the Game when requested by the elected Admin, so games could be continued at a later date if required. This would transform the game into a supergame, in my eyes. And i am sure us at FPSGAME would be willing to help by allowing a beta service on our servers.
Posted by Kevin on November 11th, 2009 at 6:27 pm
I really appreciate your honesty in this Mike, and I have to say that, flawed as it is, I still really, really enjoy Empire. There is no other game on the market like it, and few that are as immersive.
I sincerely hope that the fixes for the BAI are patched into Empire, as that would do a great deal to restore faith in CA. It would also help if CA would unlock files and release tutorials so that the poor modders actually know what they are doing.
Posted by Erunion on November 11th, 2009 at 7:40 pm
Well I have posted several times on this and other blogs like this one, I guess you dont care what I have to say. Since my post dont get posted. Even if I been a fan since shogun. So I say this, you dont like my post. Well your never gonna get my money ever again. Good buy.
Posted by Leatherneck918 on November 12th, 2009 at 12:13 am
mske the time lines more appropriate.having to wait 10 turns=2.5 years is kind of ridiculous on an attack.make the real time more real time when you make it and the difficulty should go up more.
Posted by tim on November 12th, 2009 at 1:18 am
so . . . .what about that multiplayer campaign?
Posted by Dawoad on November 12th, 2009 at 1:55 am
hey sam u may be correct but ai is far from the only area lacking and as of yet not one single person has stated if the fighting,running,jumping animations will get fixed up to the standard of medieval II.
i could still go online and play a proper human and have a great game but oh woops i cant bear to zoom in and watch men stand idle or slide all over the place.
ps CA there are tons of us with mega ideas for the series and if we could simply point and say do this and you could code it the game would be a lot lot better. hell ill offer critical advice and feedback for free if you want and ive tons and tons of ideas.
and look to the community as well you hired jack lusted who i presume was only learning C++? well darth vader and others makde brilliant addons that5 should be considered for part time CA work.
Posted by rdonnell on November 12th, 2009 at 2:25 am
“These improvements take the code further than we’ve been able before and will be there for Napoleon but we’re not sure yet whether we’ll be able to reverse them back in to Empire in a future update – the code has moved on. If we can we will.”
You talk about goal oriented plans. Why was ‘release these improvements to Empire’ not one of the goals from the start? You claim you’re still supporting the game, and then you come out with a statement like this, which gives the appearance that Empire is to be left as is. Is empire playable? Yes. I haven’t had a crash to desktop (or crash-to-reboot) in months. Is it complete? From a player’s perspective, no, it is not. I’m still waiting for the game I paid for.
Posted by Deliverator47 on November 12th, 2009 at 7:58 am
I guess I’m glad that CA is now being honest about doing a half-assed job and releasing it like it was a completed game. Battle AI that may or may not be fixed for ETW, a multiplayer beta(one of the main reasons I bought the game) that’s in a constant state of “coming soon” and has been for 8 months now, campaign AI that never seems to get any better. As if it wasn’t enough to foist upon us a multiplayer BETA, it somehow takes 8 months and counting to deliver an unfinished, unpolished feature. It all really makes me want to go out and spend $50 on NTW while I’m still waiting for ETW to be advanced into a state resembling finished.
I’m sure it’s too late to do anything about this, but if you recall what id did with its Quake series was to break apart certain functionality into DLLs and release the source code for the DLLs to the mod communities. I guarantee you if you had opened up some of the source to the public, a lot of the features that we’re all still waiting for would have been completed at no cost to CA or SEGA by the modding community. As a consumer I don’t really care who does the work for what cost, I just want the features that were promised. You could even offer incentives to developers in the form of DLC or something, but what’s clear is having completely closed source code and insufficient programming talent isn’t working at CA.
Posted by Rick on November 12th, 2009 at 11:25 am
you guys dont know how nice it is to hear from ya. dev updates are always so uplifting.
Posted by Deac on November 12th, 2009 at 3:16 pm
Dear Mike
if u finally cant fix Empires,dont you guys lose your time trying it.
Learn from your mistakes and use this knowledge in order to make Napoleon an unforgettable game.
But,be honest and fair to us.And when Napoleon releases,give a discount for all the people who bought Empires (i gave 45 euros for it) and in the end didnt enjoy the game.
Good luck.
Posted by Domesticos on November 12th, 2009 at 5:24 pm
This is a difficult situation. On the one hand I really like the look of Napoleon and would love to play the game as the designers intended. But on the other hand one bitten twice shy. I just can’t give CA more money on the promise that this time it will be perfect.
They way I see it there are several options.
1) CA can prove they will support Napoleon by continuing to support Empire. That means AI improvements, muliplayer campaign and mod tools.
2) CA can offer a discount to everyone who already has ETW, something which will encourage those burned by ETW to take a chance on NTW.
I am very worried though that a software company could allow a major part of the game to be developed by a single person and not have a method of passing that knowledge across should he leave or be off sick. I work for a software company and everything has to be fully documented and code heavily annotated so that we can pick up each others work. Lead developers check code regularly and have meetings to ensure there isn’t a single point of failure. If the developer in question wasn’t following the rules it should have been picked up at the start not the end of the project. You can’t blame him for this, it’s his line managers fault for not monitoring compliance or the development directors fault for not having system in place to prevent it. Very sloppy business practice
Posted by Paul Kingtiger on November 12th, 2009 at 5:28 pm
Got to say as one who procures services from IT companies, you much prefer a developer to be honest about their issues. The TW series is good. The standards set in previous games have upped the bar. There is a commitment here to get better and correct mistakes. Lets be fair to CA and wish them the best.
CA you have a loyal fan base. If the work you put into Empire is worth it and the fans are worth it you certainly should do your best to improve this game and not just fix it for the next version. This blog seems to indicate that you will do your best and not just look for the quick buck. After all the continued success of the series depends on the trust of your fans.
Posted by Dom on November 12th, 2009 at 10:35 pm
If anything, I’d prefer a setting of historical reality for the player. Having troops completely obedient is somewhat unrealistic.
Many battles in history were won or lost because a particular general didn’t get along with his superior, or delayed attacking because he disagreed with the order.
It would actually be interesting (and somewhat frustrating) to watch a “Charge of the Light Brigade” unfold.
As a sailor, having an enemy fleet sail directly into the wind towards me is silly and really kills the reality of naval battles.
Posted by Rockcityfire on November 13th, 2009 at 4:17 pm
I have counted, that in this blog the relation to critics/approval = 50:50. It is much better, than earlier.
I have no complaints towards battle AI, because after updating of 1.5 it is practically impossible to win, if fighting armies are big and equally powerful. Let speak alone the game.
During the fight it is possible to correct division vanity, but if to correct the parameters of sectors of bombardment and the efficiency of bombardment.
Give the chance to moddings to check up these parameters and together we will correct this defect AI. Last year, in Moscow during Empire presentation, Mr. Brigden said, that СА will present to fanatics the powerful tool for modding. We are waiting-1 year.
I consider, that Empire-it is a serious progress in series TW. I would like to congratulate the developers of СА. Now we will take a poll in Runet and we will inform you about the results. They will be as it is expected on sales.
There is one serious defect at series TW. It is monopoly and absence of a competition from other strategies.
P.S. I understand the humor of МС, when it compares the working out artificial intellect of AI and the development of rockets. The humor is one of characteristics of intellect.
But if to take for a comparison the development of model of a rocket for the boy scout, such comparison is correct. МС is rights. AI – a difficult task.
Posted by Sargon-1 on November 13th, 2009 at 10:00 pm
The first time I play ETW there was a lot of bug. But for me the patch fix it for almsot 2 month not a single bug I met. Looking forward to ntw
Posted by oPatJRo on November 14th, 2009 at 12:39 pm
ok sorry guys but i think the wants of most total war players, ie the casual ones, dont want a ultra-realistic AI that adapts to your strategy in real time and has the brain of Alexander the Great, they just want an AI that is engaging and fun to play, i would know, im one of them. This to me, CA, means that the AI in Empire should, oh i dont know, just move forward in a straight line, not constantly running around in circles or sending cavalry units half way cross the map to kill your artillery every time, or maybe not glomming up in one spot and not shooting, rather just standing there. im sorry but it should not be that hard, during deployment you select your units position, and their there, in large blocks, why cant you make the large blocks of the enemy come forward towards my blocks and engage them. sorry but this age of warfare is without a doubt the easiest time period to tactically mimic, they truly did just line up in formation and move slowly forward! its so simple, i just want a good fight, no huge maps and no running around, and no units that retreat half way cross the map and then decide to come back!!!!!!! if your struggling with the AI in an 18th century game, how on earth are you going to tackle the mid 20th century? p.s., empire was a disaster and a disapointment and youve lost a future customer
Posted by taylor cary on November 14th, 2009 at 8:49 pm
I must say that i think you completly missunderstood the whole concept of fortification, specialy the one of “star fortification”. In fact, the fortification in the 18th century is, in your point of view, no better than the one from a 12th century wich is like saying that Military Engineers didn’t developed nothing in six centuries.
In my playings i can say that it is better to defend a certain province without fortifications than WITH fortifications. That’s bad don’t you think guys? I would be glad to help you on this subject because i feel it is crucial to the game (or better, to a next game). I should say that im a fan of yours for quite some time (some years) and in this particular aspect you reaaly made a step behind… Still, i wish you the best, in your next challenges.
Yours sincerely,
luís
Posted by luis on November 15th, 2009 at 9:29 am
Although Empire Total War hit the shelves with alot of unresolved issues there has been a steady flow of updates and fixes. As it stands post v.1.5 the actual Campaign Map AI runs nicely.
However, my interest in the actual single player in empire was short lived due to terrible battle AI. Theres no doubt that the the new campaign map, diplomacy, and economics is taking a turn in the right direction, but the battle Ai is so easy to defeat without sustaining any type of damage, siege battles have honestly always been an issue with all the total games, but the ai really struggles to take a fort in empire, its almost funny how pathetic siege battles are. I feel the best battle mechanics were found in the first Medieval total war. Rome and Medieval 2 without mods were too much like an arcade game and battles with 2000 plus troops on both sides could literally be won in 5-6 minutes. In fact I actually still play the first Medieval … its that good.
With all this said i’m still amazed at the amount of thought and strategy that this empire is capable of,but if there is one fix that would drastically improve the battle ai its just having the ai form a solid line, and not attack 1-2 units at a time. i’m definetly excited about NTW and the online play is awesome, definetly been a great experience online, i’m almost at the million kill mark!!!
Posted by David on November 15th, 2009 at 10:07 am
Hello, somethings that you can do with minor patches:
1)Fix BAI desicion making.
2)Make square formation unable to form when already in combat.
These are both good for ETW and NTW. I am not telling to further balance the game as NTW is the priority but at least make it work. Anyway could you elaborate as to the nature of your new AI feature a bit more?
Posted by eleftherios on November 15th, 2009 at 6:40 pm
To Mike
I distinctly remember that Empire was supposed to have a revolutionary AI that “focuses on long-term goals rather than short-term plans with no overall goal. This will provide a much richer gaming experience to casual and serious gamers alike.” Then after that I believe an update was supposed to do something similar. If Napoleon will actually accomplish this, great. If it still has the problems the Empire AI had I will quit Total War, probably along with tons of other fans who, like me, were very dissappointed with Empire but decided to give Total War another chance based on its past success, but will realize Total War has stopped making good games. And please stop using that excuse. I don’t know if Empire had long-term goals but screwed up on basics, or they just lied, but if the Napoleon AI will provide a challenge and behave realistically I don’t particularly care whether it has “long-term” goal planning or not. The only other real complaints I have with Empire are that cavalry charges stop right in front of the infantry without doing anything, and the mines that light infantry can lay blow up with the graphics of Age of Empires I. Those two should both be relatively easy to fix and would present a much higher gameplay quality image. Thanks.
Posted by John on November 15th, 2009 at 11:38 pm
Mike, in first place let me thank you for this blog, because it gives opportunity for the fans to reach out for someone within CA and not feel like we’re talking to a wall.
In second place, I think whatever happened with ETW is past now and although I’m sure you’ll do your best to save something out of it, the rules of the market say you have to move on and make your next title rebound.
As a great fan of your franchise and as a modder (EB team) I can only suggest what I think would be the best strategy to regain the fanbase:
1 – get back to the roots. The TW franchise is all about ancient warfare and strategy in large scale; not individual ships and bullets and cannonballs flying around. Further than Napolean it’s dangerous. I think Shogun II would close this cycle and restart a new one.
2 – bring the community in. The modders won’t abandon you, if you don’t abandon them. Give us the tools to keep your older games full of new interest and the fans around. RTW still lives because there are hundreds of excellent mods available. We can still keep the franchise “warm” and waiting for the next big hit, while helping pass over the (few) fumbles along the road. But we need the tools.
3 – take the time to make a great hit. It seems like you have the obligation to make one title per year. Perhaps SEGA wants that, but someone has to remind SEGA that it was Activison who created the most awesome game made until today: RTW. They have to put the resources (human and material) enough to make a big hit, without rushing it. A bad product next time and it’s very dangerous. I still remember what happened to Tomb Raider when the developers decided to make 1 new game every year…
I really wish the best of luck for everyone at CA and I look forward to your next blog.
Posted by JMRC on November 16th, 2009 at 12:35 am
The BAI is good if campaign and battle difficulty settings are very high. In spite of some unrealistic behavior, the aggressive AI can still provide interesting campaign and battle scenarios.
Unfortunately, as a campaign progresses (after year 1750+) the AI begins to hang up on a particular country during the end turn process. After a few turns, the game freezes and the only way out is to crash the computer. At that point, reloading with a prior save will not work because the save(s) are corrupted or the end turn still freezes. This has happened to me in three campaigns. I contacted SEGA tech support: they could not solve the problem.
My computer is capable of playing ETW.
I wish more effort had been made to stabilize the game to avoid this type of crash which completely ruins what should have been a lot of fun. In reading the forums it seems many other players are also experiencing crashes.
I am not going to waste any more time playing ETW. It’s impossible to enjoy the game when it crashes just when the campaign is getting interesting. I will not waste money on “Napoleon”.
JWN
Posted by JWN on November 17th, 2009 at 8:51 am
Hey
You know what life happens..! all in all i think ETW is a fantastic game! however i would like to know why the involvement of steam? no offense to anyone and maybe i am being greedy but steam is a real pain! I have been a fan of TW since the beginning and have been at the edge of my seat ever since but not the whole steam thing has really made things difficult.. will all the future TW’s be the same?
To be honest the TW team has created a series of eppic proportions and has made a dream come true!!!
I would love to apply to work for you guys but unfortunately i am hopeless at programming!
Look forward to your response.
Thanks again to the team!
Regards Aidan
Posted by Aidan on November 17th, 2009 at 2:51 pm
Having read the moans and groans I find myself feeling disgusted at some of the comments about the state of the AI.If any of the moaners can find a game that closely mimicks real life tactical command then I would love to hear about it .It appears to me that a lot of players have forgotten whats good about the whole TW experience , have dismissed out of hand why they even bother buying it .I most certainly didnt buy it because of its AI but rather for its scope and sheer size,thers nothing else out there that compares and I would like to say at this point well done to all the team at CA for constantly giving us somethng that grips the imagination day in and day out,yes we know you’ll get on top of the AI bugs and hopefully when you do then you’ll release patches to correct bugs in earlier games .I have played TW since its conception and I look forward to playing many more fine games and addons of the same genre, well done all at CA for giving us a unique gaming experience!!
Posted by Gray on November 18th, 2009 at 7:48 am
I’ll try to avoid repeating previous posts as much as possible. Suffice it to say that all have merit and imo they all should be taken under serious consideration (rocket science aside).
To me also it was obvious from the first time i played the game (literally) that the dev team had changed and with disappointing results. Bottom line is the community is dissatisfied and the game is not what it should be, especially in the light of the MTW2 and Kingdoms experience. Very good ideas overall, poor implementation though. The game needs fixing. You guys are top pros on the field, so fix it. The series has its fanatics and we will support the effort. But the effort needs to be made and it needs to be successful. If not, then all we can say is it’s been a great ride all these years. So either way, thank you guys. Try and keep it up.
Posted by Konstantinos Drakakis on November 18th, 2009 at 12:41 pm
Ive bought every single Total War product, Total War games are my favorite games, by far, ever, I enjoyed my Empire campaigns, but they were no challenge at all, even on vh/vh, it didnt matter how many stacks the campaign AI threw at me, the Battle AI just failed miserably. Fortunately, I love the time period so I dont feel like my purchase of empire was a waste.
That being said, I wont be buying Napolean unless more progress is made on Empire, I really feel like CA pulled one over on me with Empire, you sold me an unfinished product.
Fix Empire yourselves, or release the tools so the community can do it.
Posted by Aaron on November 18th, 2009 at 5:07 pm
Hi CA, to start with have been a fan of the series since shogun, the only 2 games in the world i play are Total war series & Football Manager. I think empires is a great game but i doesn’t work on my laptop as do not have 512 bit graphics card. Anyway i do agree that the AI is dodgy and does some really weired stuff and should be fixed. I agree also that i would rather that you take more time to make a game than bring on out every year. Also i hope that your next game is Rome total war or Medieval 3, some people find that boring but those games were for me off the charts espeicially when you hav stainless steel and take factions with one province. It gets really interesting. Empire is the most graphically great game and i like the new idea of how provinces work but the 2 previous games would have to be the best. I am a big fan & will probably remain a good fan but have been waiting for multiplayer since i heard about it? I am dreaming of the day whan i can play with friends on Campaign mode with the battles at least then AI does not become a problem. Please deliver this to the community.
Good luck guys.
Posted by Nathaniel on November 18th, 2009 at 7:00 pm
Good thing that you are (finally) working on it.
You see a too easy AI opponent as a problem.
But maybe, that’s not the problem.
The problem is that all gerenals you fight with, fight the same, with the same formations and tactics (and same number of troops).
If you fight against rebels, they should just rush at your lines without any formations.
If you fight against napoleon, he should be almost invincible.
The big problem here, is the difference between the armies you fight against.
And please, remove the stupid walls on the battlefield maps.
I hate them, they are useless and unrealistic.
The 2 big problems of the game are still diplomatics and battle AI, please fix them first, then work on better graphics.
I want Rome II : Total War , because RTW is still the best Total War of them all.
And sorry for my bad English I’m from Belgium
Posted by Ruben Heymans on November 18th, 2009 at 8:40 pm
I don’t understand the complaints about the AI for ETW, i hold no claim as a brilliant tactician but the TW games are their own genre, and on things such as the battle maps and the depth of everything, from the syncronization of soldiers fighting in a melle, to the micro management of regions, when compared to other games like civ, or generic RTS’s, it’s light night and day. While I am eagerly awaiting Napolean Total War, I must say that I am partial to Rome TW, and ferently hope for its sequel to find it way into the series. Good luck on the ne game.
Posted by Publius Scipio on November 19th, 2009 at 11:54 pm
Because i’m a fan of TW, i gave up on ETW and restarted to play MTW II…! Napoleon…? Depend’s …
Posted by JSilva on November 20th, 2009 at 1:45 pm
@35: A game with a better battle AI would be Rome Total War.
This is 2 games lates and the AI in RTW is still better, this is just weird.
Posted by Ruben Heymans on November 20th, 2009 at 7:58 pm
I am happy to be playing ETW again!
I want to apologize for some previous statements about the 1.5 patch screwing up the game for Nvidia 200 series cards.
It was not your patch fine sirs .. it was the latest Nvidia drivers (191) that have been destroying boards and cards (really .. temps into the 100′s).
Just google “nvidia 191 driver issues” and you will get the ugly story.
I was a vocal complainer and i was wrong .. just wanted to set the record straight.
Posted by Tony P on November 20th, 2009 at 9:40 pm
Well, i get what your saying and no doubt coding so a cp enemy thinks like a human is a crazy job, but I will tell you what i think.
I understand why fans would support your honesty, and other fans are angry with your answer from reading a few of these posts. Firstly, I do enjoy this game never the less, im still playing it, I personally am patiently awaiting this multiplayer campaign, some people mod tools, some BAI improvements. But I can’t really say I don’t applaud you for working so hard to fix the game, though it does make it look unfinished out of the package.
I really have no knowledge on where any of these things are in current state except the Battle AI that it seems your saying is unlikely to be solved. I don’t expect the details, but I’ll tell you what bugs me the most.
you promised BAI that wasn’t up to the standards u sold it advertised to be, and all the while, you advertised, 12 unit dlc, a warpath expansion, and special forces dlc, and I bought them all. Now that your done selling products to me for the game, you come out with a messege saying you have no interest in fixing it? And you keep Advertising Napoleon even in this blog.
I’m not rash, So I’ll make you a fair honest deal. I will buy napoleon 4 months after it’s release after reading the forums for positive reviews.
or you could get that multiplayer campaign out and I could play a human player and wouldn’t have to worry about the ai being as smart as him.
Posted by Ripa on November 22nd, 2009 at 8:09 pm
Thanks for your blog. I’m also a loyal fan, ever since Shogun, with just one thing to say:
…the only desire I’ve ever had was to play Multiplayer Campaigns.
If you allow human vs human campaigns, I will tolerate any number of problems with diplomacy or battle AI!
Posted by Matthew on November 23rd, 2009 at 4:51 pm
The thing you released in no way resembled a finished game, and you are perpetually asking for more cash for content that should of been in the original product.
Do you really think now is a good time to be looking into the battle AI?
“By failing to prepare, you are preparing to fail.”
— Benjamin Franklin
Maybe you should of listened to your own pontification.
Posted by angry customer on November 23rd, 2009 at 7:28 pm
If CA is having such a hard time with AI, why don’t you just open source it? The modding community is filled with brilliant programmers and people who have struggled to improve the AI of previous CA games. Use them.
Bad AI reduces sales. I have not yet bought Empire (even though I’m a fan of the period) because the AI was heavily criticized. I won’t buy a Napoleonic version if the AI isn’t improved. However, if you open sourced your AI and it could be modded back into the game, you’d increase sales and interest in your product.
As it is, I only play modded versions of Total War games. The games are not particularly great right out of the box, but become great when the community is let loose to improve them. Increase the number of things that can be modded and you increase sales. To play a mod, I still have to buy the original game.
Let us do the work for you and we’ll happily pay for the privilege.
Posted by Aurelian on November 23rd, 2009 at 8:45 pm
How come all the new monarchs don’t get trait such as Majestic or Fascination with Foreigners they just get traits like Born Vivant and Jug head,why would a Absolute Monarch have jug head as a trait!? I find this annoying and it makes monarchs seem more like regular ministers and not a monarch at all. I think Fixing this would make the game more realistic. Please fix this and by the way I am a big fan of the Total War series and best of luck in the future.
Posted by Pi Ryan on November 23rd, 2009 at 8:56 pm
It’s obvious that CA never intended to install a multilayer campaign, or host a good AI. I’m just upset that I read all those misleading articles before I purchased the game.
Posted by Scott on November 24th, 2009 at 6:26 pm
This game has lots of good ideas. Unfortunately it still has AI flaws that it had in MTW (as in the first one). Now MTW came out years ago, I would have expected the flaws in that system to have been at least addressed by now. But lo, my first ever battle against a superior force of Indians (as Britain).
I thought it would be tough. After all, these Indians were supposed to be tough. I was outnumbered, but I had guns against their archers. At first it looked promising. I was ambushed in forest and forced to go hand to hand with some warriors. All good so far. Then my footmen were able to get the advantage after I flanked the Indians with dragoons on horseback.
All good so far.
I was out of range at this point for my artillary. So I lined up my rifelmen and prepared to fire at what I thought were advancing warriors. I was concerned about the bowmen behind them. I started to fire thinking I would be able to take a few of the warriors down.
The warriors started running around randomly.
Maybe they are panicking?
Nope, they are just running around.
So I shoot them all, like goldfish in a tank. Reminds me of the extremely effective longbowmen from M2TW who just fire at whatever troops happen to be standing in front of them, without any concern that the enemy will charge. My only regret was that I could not hear the Indians saying “they’re firing at us”. I used to love that bit.
So I think I have an answer for creative assembly. Go independent once again. Ditch Sega, certainly ditch steam and become what you once were… a leader of the pack instead of a lame duck.
All I can say is, I’m glad I decided to wait until now to buy it, I got it for half price. Thankfully I didn’t buy Nepolean either.
I was sceptical about ETW. Simply because M2TW was left so unfinished (even after the expansions). It seems like, once again, the mod community will be picking up the programmers tab and finishing it for them. This seems to be happening right across the gaming industry. Sooner or later the community will get wise and (like myself and many others) refuse to buy a game on launch simply because they know there will be an unacceptable amount of bugs. I accept that there are many different PC systems and you can’t foresee every bug. But personally I believe that buggy gameplay and regular CTD’s should be an acceptable reason to take games back to the shops. I mean, if you were building cars you would have been shut down by health and safety by now.
Posted by seanys show on November 24th, 2009 at 9:26 pm
Sega/CA,
I expect an “essentially completed” product that I spent $70 for and ETW is NOT that.
The Battle A.I. IS atrocious and the Sieges are not real (i.e. Ropes for every unit?) as well as a host of other bugs which give us ETW at its current state, after 5 patches.
Sega/CA, Redeem yourselves and give us a patch/hotfix for ALL these bugs and give us the modding tools promised to us as well.
NTW being bought is contingent upon both those stipulations for me, a 10+ year customer of CA back in the Shogun: Warlord Edition days
Posted by Steve Kafkoulas on November 26th, 2009 at 2:25 am
I appreciate your honesty, like most other people. But I have to say as someone who has all Total wars except shogun (can’t find a copy) and who has been playing total war for all my gaming life (18) I have to say I am somewhat disapointed and hesitant to buy Napoelon. I admit I will probably eventualy buy it but I was very much looking forward to Empire. However Empire is not up to the standard that I feel the Total War series has established.
My biggest problems are Battle AI and Diplomatic AI, also what happened to family trees, and why can we only play 11 factions, why aren’t more of them unlockable (I know there is a mod, but that is not my point.)
Posted by Thomas on November 26th, 2009 at 1:04 pm
I wish there was a Shogun II: Total War.
Indeed, the diplomacy at least should be fixed, the ai is just silly with it’s demands.
Also, the battle AI is fine if I adjust my style of tactics to give the AI a chance. But, it still would be nice if i could play the battles as i want and the AI could respond to all of my tactics.
Posted by Aqnde on November 29th, 2009 at 2:36 am
I have loved the campaign in Empire since the first time I played it. But I have found the battles to be very boring!!!! There I said it!!! BORING!!!! I have played the game since Med:TW 1 and can say that the battles were better in that even!!! Rome: TW was by far the best game I have played on the PC and still to this day I play the multi-player where there are still at least 50-100 people online at all times. But with empire I have found the battles lack any of the excitement that Med 2 and especially Rome had.
It’s certainly wrong to say we will fix the AI if you spend another £30 but I think everyone is missing the point here, which is that the battles are boring, not because the AI is poor, but because IT’S NOT EXCITING!
Love CA and really hope you do the right thing by the fans and yourselves.
Posted by anthony on November 29th, 2009 at 9:52 pm
Cough. Well, the exact timing of Richard’s departure is neither here nor there, I suppose. I’d suggest clamming up, and demonstrating that CA is capable of writing good AI in the only really credible way – by doing it.
Posted by Jerome on November 30th, 2009 at 1:04 am
A noticeable improvement over previous titles, and wow, so much opportunity to improve, awesome!
Should of given me a job to play test it more, I could of helped the techies out some.
Posted by bod on November 30th, 2009 at 10:02 am
Oh right nearly forgot, please tell me why it is that CA continued to allow the ai to field unbalanced armies, consisting of almost entirely of cavalry or artillery for example?
Posted by bod on November 30th, 2009 at 10:10 am
Ah come on, give CA a break. the AI in battles since the latest patch isn’t that bad; I’ve certainly had a few eye openers in some of the smaller battles I have done where the enemy did something totally unexpected.
And I’ve noticed since the last patch that the AI’s armies are way more balanced as well.
Besides, “Every battle is won or lost before it is fought”…warfare during that period was all about strategic and tactical supremacy, bringing the right tech. to the field and choosing the right ground…when I play TW I count as a loss any battle where I lose >10% of my forces, if you plan it right and pick your fights with the right units it’s a pre-determined result. 80% of time spent is maneuver, 10% fight, 10% mop up.
Back then, battles were very rarely won or lost by the acts of individual units on their own initiative. It was all about the setup.So for me that’s what it’s all about (and has always been with the TW series).
Anyway, it’s a game – do you really expect Skynet-typeAI? Even the US Military hasn’t been able to get that right yet!!
Posted by Steve on November 30th, 2009 at 6:04 pm
If you really want a good programmer, you should reach out to Darth. I am sure someone mentioned it here somewhere, but I hope Napoleon TW brings some sanity to fortress battles. They had ladders and undermined walls in the 18th and 19th century too, you know — it puzzles me why the last team decided upon those ridiculous ropes.
Posted by B.C. Milligan on November 30th, 2009 at 6:46 pm
The only problem I have with this game is that the quality of animation and effects (such as the blood spraying when the soldier gets hit) is not as good as the animation in Medieval 2. The ship battles are great, but CA couldve spent more time on the land battles like they clearly did in Medieval 2 which is the best TW game in my opinion.
Posted by GRIM on November 30th, 2009 at 7:54 pm
you guys r so cool, to be open, receptive and candid about our complains…bcos the guys who bother to post really care very much about TW like me~
Posted by Jin on December 1st, 2009 at 1:35 pm
I loved the concept of turn-based campaigns along with realtime battles in Lords of the Realm II. I enjoyed wasting many hours on that game as well as on its logical successors, STW, MTW, RTW, and MTWII.
Now I’m an old guy, almost 50, and I still love these games but the problem is I have very little time to invest in one of these games. However, when I do have time, a mod of MTWII is the game I will play. and I am happy to spend months on a single campaign if I believe there is any chance it will be a challenge. Therefore, as soon as I determine that the AI is sorely lacking in the vanilla versions of these games, I start searching for a Mod that will mitigate some of the original’s shortcomings. It’s due to this expectation of a sorely lacking AI and the absence of a challenging mod that I have not purchased ETW.
I suspect many of your customers’ gaming histories are in part similar to mine. In your “Blog the Second” about the Campaign AI you wrote:
“Making a passive AI may have sold us lots more games in the US, but it wasn’t intentional. Maybe we’ll have a play style setting in the future, but for now our intention is to challenge the player with an AI that is as aggressive and varied as human players would.”
I am glad to know your intentions are in line with what most of us desire, and it gives me hope that you will continue to improve the series as more and more of your customers lose the distinction of being a casual gamer (initially we were all casual gamers). However, I don’t expect to purchase another TW series game unless it provides or allows at least one of the following:
1. Challenging Campaign & Battle AI
2. Multiplayer Campaign
3. Extensive Mod-ability
Instead, I suspect I will merely continue to search for an excellent mod for MTWII
Good luck, I know it’s extremely complicated but I believe if you are able to make the next giant, innovative leap into truly challenging gameplay in the strategy genre, you will have loyal customers for decades. Thanks for at least giving an old guy something to hope for!
Posted by Tristan on December 1st, 2009 at 8:03 pm
As so many have said before me, thank you for the honesty. Even though I don’t really like what I hear at least your not covering it up in sweet talk and BS =).
Please do your best to fix the AI in Empire, I really love it but I can’t get immersed in a game where the BAI Bunches all of its units together into a blob and lets itself get encircled and destroyed.
I can’t tell you the number of times in the campaign where I have beaten armies that I shouldn’t have been able to. For example I was at war with Spain, I had one full stack of completely intact units. He attacked me with, 8 full stacks of his. Two hours of repetition later I had destroyed all of his units while I hadn’t even had one of mine rout.
So please CA, fix Empire…you don’t have to make the AI perfect, but at least make it function. Thanks.
Posted by A_Parked_Car on December 3rd, 2009 at 5:33 pm
Imperial Edition must have a historic map poster. What is it with DLC units ? can u just add all extra units to Imperial Edition only, not like Steam (digital) offering exlusive unit not included in IE, but does not have 5 exlusive units as box edition has. From steam you get 10+1 special units, but from box Imperial Edition you’ll get 10+5 units (without Grey scottish cavalry unit, offered by steam). Why not make all exlusive units in one Imperial Edition box copy ?
Posted by Dan on December 4th, 2009 at 11:01 am
Please make NTW awesome because i dont want TW community to be gone forever
Posted by heh on December 4th, 2009 at 4:41 pm
This is a great business idea o… just think of the repeat revenue!!! WOW for history buffs ‘Quote’ we would want it for RTW as well. What i would like to know is where is the multiplayer campain mode we where promist. And if possible could we have a multiplayer campain mode that runs on dedicated servers? So in affect all players could take there turns at the same time and then when they have all finnished the server then playes the AI. This could then be able to save the Game when requested by the elected Admin, so games could be continued at a later date if required. This would transform the game into a supergame, in my eyes. And i am sure us at FPSGAME would be willing to help by allowing a beta service on our servers.
Posted by sean on December 4th, 2009 at 5:03 pm
I hope in twenty years when I’m in my old age like the poster in comment number #69 I’ll look back on these gaming days and think of that dude who posted about having hope for the future of AI and a game that not only delivers, but exceeds expectations. Cheers to you, I hope you’ll be 69 like your post #69 and still playing then too.
Lets be honest folks, the AI never really improved, did it? Rome: Total War will always be the best game, because that’s where they left off with the AI. Maybe the problem with the code is it’s the same code? I’m not programmer, but I know a little bit about foundations and like the good book says, through Christ, ya can’t put new wine into old wineskins.
Posted by Alex on December 6th, 2009 at 8:34 am
I wish you find the best AI programmer.
Total War games deserve better than all these bad posts from persistent flamers. One cannot blame them because you confessed the problem of the AI openly. Still, even with this problem, if there was a good balance around this code, all could be a lot better.
So, I do feel that there is also need for balance around the AI code except from the base code. These 2 must be handled well together or else one will exploit the other.
Posted by Nick Thomadis DARTH VADER from TWCenter.net on December 7th, 2009 at 2:31 pm
I think the AI is understandably shaky as this is CA’s first foray into latter periods but needs improvement. The Diplomacy is a more annoying one for me as i offered 6 techs for one and was resed by an ally who really liked me! Unfortunately i still suffer frequent crashes and have had to fight some battles up to seven or eight times so have found the AI does change occasionaly to minor things i do. Will i buy Napoleon? Not unless i see if for a fiver as i dislike steam and have not been able to ever play Empire with full satisfaction. Sorry Mike.
Posted by Stu on December 7th, 2009 at 6:40 pm
I’m nearly 50 as well, I’ve been a tabletop wargamer since I was 12, I have three periods that I game in Ancients, Napoleonic & WWII.
You guys have produced RTW it allows me to game alone or multiplay, after all the upgrades ect. it is a wonderful game.
Now, I purchased ETW because of my Napoleonic passion, I was disapointed that it stopped short of my era, but then, found out you were moving on to NTW, rippa! So I decided to play ETW to get the feel of the new system, not happy with it, but hey it’s not my period anyway. I’m glad you’ve chosen to go back and start with the RTW AI, at least it works.
I as a tabletop wargamer have come across many gamers in my time, some are not so simpathetic to the era they are playing add what we call the god view to this and you get unrealistic games, so to counter this back in the 70′s the rule sets included army lists to stop designer armies, to my point………
While trying to play ETW online I’ve come across a lot of young players with fast computers who play the “rule set” not the game, and I have seen bizzare armies that you can’t beat, so much so that I’ve noticed mutiplayer games that have to state “no more than 4 light inf” ect. The points system needs to be addressed as well as placing restrictions on non-common troop types. I still try to field armies that have a resemblance to a historical precident, alas to no avail.
We now play online by invite only to get a good historical game, hopefully you can put a simple system in place by way of either an army list or adjust the troop points to reflect rareity not just their combat ability.
One though that did come to mind is a sliding scale, for example if a troop type is rare it could become more expencive with each purchase above the historical precident. Therefore the designer armies would become more expencive and thus smaller and less common.
If you read this Mike I hope it provides food for thought.
Eagerly awaiting NTW.
Posted by Daggz on December 9th, 2009 at 2:04 am
i would like it if difrent generals used difrent tactics. i would like styles for every AI general. i want to be in for a suprise or a move that makes me have to think and counter move
Posted by mike on December 11th, 2009 at 7:10 am
Was just thinking, how about this for a story, where u can start the game enlisting as an actual soldier choosing from scout, gunner, infantry, cavalry, sailor or whatever, and fight in first person mode.
Actually go in first person mode and shoot a cannon or aim that musket yourself, wouldnt that be cool!
There is a first person vieuw allready, but u cant do anything with it, shame.
And after some battle,s ull get promoted etc, think about it, the possibilities are endless
Posted by LastMohican on December 11th, 2009 at 7:26 pm
Fix the game. It wasn’t cheap and we all had our hopes up about it.
Posted by Robert Bloodworth on December 12th, 2009 at 12:05 am
Thanks for your honety Mike, ut quite frankly it is the only I thank you for because this game isn’t worth a word in terms BAI. To make $400 dollars a week and spend literly 10% of your earnings on a new game froma series you love, then wait 4 hours for the game to install and get this content is so dishearening. I understand your MC score from the users of 67% caused concern to not get money for future games but if you guys can shp this product to retaillers you really don’t deserve more money. I know its not your intention bu for me to buy a game like this, that I can’t return or even trade in…feels like an infomercial scam.If this isn’t fixed I guarentee half the sales for Napolean are aleady out the door. Sorry if I’m harsh but I’m sure you made $400 a week once and losing $40 like that really huts…it really does…
Posted by Brandon on December 12th, 2009 at 8:54 am
What about this. You record each battle that everyone plays and upload the results to a central server (with player permission of course). Scan for patterns. When you recognise a successful human offensive pattern use that against human players and analyse their defensive response. Then use that defensive response against them. Keep doing this and the computer tactics should continuously evolve. Use the human players to train the AI. Easier said than done but maybe worth investigating. Btw, I’m a huge fan of the series and would love to work on this. You should have my cv on file, I applied online last friday.
Posted by Ilya on December 14th, 2009 at 6:42 pm
Hi, my name is Alex and I’m a long time CUSTOMER of yours (customers are getting rarer due to rampent piracy these days so I felt I should aknowledge that) I’ve bought and paid for Rome, Medieval 2 and Kingdoms and loved and enjoyed them all.
The great thing about Medieval 2 is that even though I really enjoyed the vanilla campaign, there have been at least 3 wonderful mods for it that totally breathed new life into the game and added very literally 100′s of hours of playtime for me.
Hearing fans lament about moddablility worries me, as does the frequently voiced concerns over battle AI. M2 had it’s ‘quirks’ but modding made it at the very least competent. I’m one of these aging (24 lol) fans who no longer believes a word of what the mainstream gaming press says. I tend to listen to what other fans of the series have to say. Quite frankly what I’ve read so far is troubling.
I’m in the process of upgrading my computer, primarily for the purposes of playing Empire (I’m that much of a fan of your past work that I’m willing to spend 3-4 hundred dollars to play a single game) it’s amazing what things like ‘reputation’ and ‘prior good will’ can get you.
I have to admit I was concerned with all the DLC and special unit stuff that went on at the release of Empire. Stuff like this is good at swindling a few extra dollars out of die hards and suckers but most fans can smell a lame cash grab a mile away.
It seems to be the way the business is leaning these days, puffed up mainstream reviews, portions of the main game cleaved off and sold on release day as DLC, and the alienation of core fans to meet the bottom line and rushed release dates that turn fans into paid Beta testers.
I’m planning on witholding my purchase of Empire and Napoleon until I see a bit more commitment to either polishing up Empire or releasing proper modding tools. Once this is done I will gladly hand over my 150 bucks. Thanks for the honesty though Mike, it is refreshing, don’t give up on Empire…
Posted by Alex Sellick on December 17th, 2009 at 7:58 am
Any longtime fan has heard this before. You guys have been promising for years “Our AI won’t be horrible this time around.” Unfortunately, the rest of the game is so incredible the horrible AI stands out even more. Get it together CA, your games are no longer must buys for me.
I really respect that fact you picked up Jack Lusted, but putting Richard Bull in charge of something this massive was your fault, not his. 3 people cannot make an advanced AI this game requires.
Posted by Greg on December 18th, 2009 at 3:31 am
I can repair damaged forts can you add a repair icon for the same ?
Posted by Andrew Rodgers on December 19th, 2009 at 5:32 pm
You guys need to get off CA’s back. You all keep saying mod tools will be the answer, but very few of you can make a real mod, let alone fix the player A.I. And don’t say buying Shogun is justification. “you’ll never get my money again, goodbye” comment 22. There no other RTSs like it, if you think this bad see Aoe3. Rush, rush, rush. No self defense. NO REALISM WHATSOEVER. So ya, go buy another, crappier RTS. As for me, i’m willing to fork over the 40 bucks. Innovation is never perfect you morons.
Posted by jimmy on December 21st, 2009 at 4:50 am
Why oh why do you have an aversion to realistic simulation? Without the facility to adjust game speed, units zoom around so that if one looks away for a moment, half the line is ambushed. Why all the tedious complexity of campaign parameters when the one absolutely essential thing (supplies and lines of communications) is completely ignored?
Add to that bugs upon release and an AI that appears to have really annoyed Adolf Hitler : ) and you really are shooting everyone in the foot, leaving it to the modders to try to save the day. This series could be SO much better.
Fog of war options would help a lot too. As I say, it seems deliberate that realism is avoided at all costs – a real shame when the graphics are so realistically state-of-the-art.
Posted by atd on December 21st, 2009 at 8:59 pm
I think one thing no one is mentioning about the ai is that if Ntw work that whilst playing your campaign your battle opposition can be controlled by a ‘drop in’ human player then the battle ai will be varied and interesting because it will be human! So if you enable this option then you will get the ‘ai’ level you all want. If this feature does appear in game at release then CA have made a very good strategic decision themselves, they cannot hope to provide the level of ai you all want, as someone said already if the US army can’t little ca won’t be able to! So instead they provide a simple option that most likely to be more effective and fun than any scripted ai. So really we need to hope they are putting their efforts into the campaign ai (and for me the diplomatic options) I really way to see more interactions between states on economic actions other than just trade or not trade.
Etw wasn’t that bad: The problem for me was if anything I think it was too big! And India could have been removed, America was bland and Europe lacked provinces. I’m looking forward to a France made of sections not just one large province! (who eva did that should be fired ;p )
I would like to see moveable solid borders in the future. Meaning france is again one province but an invading army has an ‘area effect’ so if peace is negotiated the area occupied is added to the territory of the occupior an the border is re-drawn on the map. I think it must be possible an would be far more realistic than the current umchangeable provinces maps.
Posted by Mike on December 21st, 2009 at 11:07 pm
All I want to know is…Will America be included in the NTW series. The U.S. was going through some pretty good wars during this time frame as well…War of 1812, Mexican American War, and the American Civil War, all have really good battles. If playing as the American Faction is not going to be included in NTW, will there be a series based on America?
BTW, not knocking CA in anyway by these questions, would really like to see a campaign map where the whole world is included complete and you can finally play as another important faction in history as well. NTW is going to be way better then ETW, can’t wait.
Posted by Jay on December 22nd, 2009 at 5:16 pm
I’m suprised myself because I’ve been playing Empire again, post updates, and it really has improved A LOT– now there’s still more to be done, but diplomacy I think is markedly better and battle AI really is improved too. Seiges still leave a lot to be desired animation wise- mostly because I feel like they just don’t make sense. They look like a line of ants fighting another line of ants, one ant at a time on the ramparts and the notion that you take a set circle in the middle of town and a timer starts is kind of silly.
STILL, I gotta say, it has gotten better, I am (finally) enjoying playing and appreciate the continued efforts. I think the bottom line here is your honesty, as others have said, is appreciated and the community would LOVE to continue to support your efforts- just gotta make sure the next one’s not a big bag of bugs like Empire was at launch- a couple of bugs, hey, it happens. Whole formations facing the wrong way, well… ah, but I’ve given you plenty of grief and would rather just say I appreciate your work for sure. Merry Christmas!
Posted by hayes on December 22nd, 2009 at 7:48 pm
Dear AI Development Manager,
I would like to share the following idea for future release:
To make the game in Multi-player mode more interesting, it would be great if you can borrow the units played by either the CPU or a Human player on your team.
The idea would be to request units from a friendly.
This feature would make the game more interesting because sometimes players drop from the game and we would like to take control of their units to prevent a slaughter.
Similarly, in practice mode, it would be interesting to get control of friendly CPU’s units to manage their fate.
Finally, in online battles, sometime you need help and you would be able to lend a unit to a player so they can manage it to save themselves.
Thank you for considering these ideas for a future release of the product.
Thank you,
Erik Boisvert, One of you greatest fan.
Posted by Erik B on December 25th, 2009 at 7:58 pm
I was disappointed that ETW did not allow you to put part or all of your force under AI control, as in the other TW games. I generally like putting most of my units under AI control, and then commit my reserve at a critical point. That way, I could enjoy the movie the game creates.
Posted by Roger Cooper on December 26th, 2009 at 12:23 am
Well I purchased ETW yesterday without reading these reviews, and maybe I wouldn’t have if I’d read them first. However, is it the case that it’s the people that are most unhappy are the most vocal? I can’t believe things are as bad as implied here. I’ll give the game a try and hopefully the BAI can beat me
Posted by Halesy on December 30th, 2009 at 2:14 pm
I think this is a common problem for the software industry. One solution is to cross-train people. So instead of having 1 person work on it 8 hours a day, you have 2 people working on it 4 hours a day. 1 person working on it full-time is faster than 2 people working on it party time, but more people are learning new stuff and gaining new skills so there are more reasons to keep them stay and work for the company. And when one of them have to leave, the new guy can learn it from the guy that is still here to pick it up quickly, especially if most of the code isn’t documented.
Just my 2 cents.
Posted by Zhi Wen on January 1st, 2010 at 2:43 am
Ok I have been reading some of the comments on the AI and frankly I have not seen one about the fact that we are not given the option for the AI to control our reinforements. This has me really shocked. I have been a long time customer starting from the Shogun days. The way reinforcements could be controlled in Rome and Medieval 2 rocked, and yet here in Empire your reinforcments count for almost nothing unless you auto it. I have heard some people say that most computers would not be able to handle it but I think that is crap because my computer has 3 times what the recommended hardware is for Empire. Please don’t get me wrong, I like the game, I really do; but right now I would have to say if the expansion is not better I will not be buying it for debut price.
Posted by Thomas on January 3rd, 2010 at 7:40 am
I develop and experiment with many tactics myself in the Napoleonic age and World War 2 age and I would love to be able to excercise them freely and effectively in Napoleon Total War. Being a Napoleon maniac, It would absoulutely make my year if you guys could really make something to real life. I wish I could help but I live half way across the globe…..at least over the Pacific =D
Posted by Martin Wu on January 5th, 2010 at 7:29 am
I am very disappointed. I don’t expect genius A.I. however I do expect A.I. to be functional. there are examples of acceptable battle A.I. in this genre. Sid Meier’s Gettysburg/Antietam and Waterloo Napoleons Last Battle, Austerlitz Napoleons Greatest Victory are good places to start.
I am frustrated by comments like “A.I. can never match a human.” It is this attitude that has allowed developers and publishers to put A.I. on the back burner and get away with assigning one person to designing what is arguably the single most important part of the game. Ask the people who play chess if designing good A.I. is impossible.
I am realistic. I don’t expect computer chess level A.I. in a game this complex. The amount of resources required to accomplish such a task “is” unrealistic, but good A.I. is not impossible nor is A.I. capable of challenging humans. Such rhetoric is a fallacy and unproductive. We as war gamers have come to accept the idea that a computer A.I. can “never” be as good as a human and it is just not true.
Humans have strengths that computer A.I. may never have in our lifetimes. Computers have advantages that humans can’t match. While humans offer unpredictability and the ability to think outside the box. Computers can think faster and make many decisions essentially simultaneously. Computers can quickly calculate the probability of success of unit “A” attacking unit “B” under any conditions and react accordingly to achieve the best possible result. Computers can calculate distance and speed and can decide what maneuvers are possible in an instant and randomizing certain actions can simulate unpredictability. There are many things A.I. does very well and can be used to make good A.I.
We as war gamers need to demand good A.I. If developers would devote the as much money and resources to A.I. that they devote to making water pretty or making sure that we can see individual cannons being loaded etc. this would not be an issue.
Aside from A.I. issues, the trend of publishers nickel and dime strategy of releasing DLC that should be included and not allowing the modders access so as not to screw up future revenue from DLC is backfiring. It is the community of grognards that make or break war games.
War games by nature are a niche market. Developers should stand their ground when publishers push for pretty graphics or dumb down game play to bring in more sales. This strategy pleases no one. Casual gamers will probably not be satisfied with any changes to Total War and grognards and longtime fans will be put off because their needs are not being met at the expense of meeting the unrealistic goals of publishers that want a game that appeals to the lowest common denominator.
I truly hope that Creative Assembly sees the light. I love the original vision the company brought to the genre. Sadly it appears that they have sold their soul and have lost control of that vision.
I wish you well Mike. I do not envy your position. I will probably get Napoleon Total War simply because I like and want to support Creative Assembly however I do not like the road the company is taking. I mean Steam required for install and play, limited mods and units released as DLC. I don’t mind paying for an expansion pack but charging extra for units is freaking ridicules and insulting to those of us who “gave” Creative Assembly it’s current success.
Posted by Joe Adams on January 10th, 2010 at 8:44 pm
I actually think the BAI of the standard fights is pretty good, however in every total war game I played(which means all of them). The siege fights leave room to beat the enemy with half their men, or even a quarter or less.
With the Rome and Medieval TW’s they all had considerable defaults concerning going up the wall, it was easy for the defender to send mass units to a breach, and then the units would clutter and bug up…completely running over the gap in seconds(which is still the case in empire btw).
With empire and it’s larger use of artillery the walls get smashed more often then they used to I feel, and now the main problem seems to be that unit formations tend to wait for all their men to arrive to the center of the fort(if your own units are stationed out of their way, else they go melee on you) and once all their men arrive they finally start to undertake more actions.
Posted by Ryan on January 12th, 2010 at 8:38 pm
Is it me or is bai in this game totaly sucky compared to total war. I think its cause of using guns as the main weapon of choice. Just to complicated to make intresting. Maybe i dont get it but i been playing tpotal war since the begining and loved every upgrade so far. Everything but the bai is so impressive maybe ill pretend its civilization and ignore bai.
Posted by todd on January 16th, 2010 at 2:34 am
Even though im happy that developers like ca/sega make this kind of historical games of great quality but i have to say that you have started to ruin your reputation lately. Im tired of poor excuses and false promisses. If the Battlle AI isnt improved anny time soon i will never buy a product from this company again. Like mentioned above: you owe this to your costumers. After all, we paid for the game right?
Posted by Chris on January 19th, 2010 at 6:37 pm
I have to agree with many of the above criticisms of ETW. I loved Rome and MTWII and was really excited about a more developed game set in the era of imperialism. Unfortunately even after a couple of patches had come out, the game was still massively flawed. Last time I played it your city fortifications still couldnt be repaired due to a bug in the code. Thats a pretty fundamental aspect of the game, and to keep releasing addons and new games whilst overlooking things like that is a reflection of the way modern games (like most modern media) seem to be developed i.e. with the company’s focus entirely on making a quick buck. If you really believe in creating a realistic and intelligent AI then I would suggest you simplify the game first and build from there, instead of ploughing forward from medieval strategy and warfare into imperialism, with the increased game size and detail inherent to that progression and without having the game engine to support it, you should really go right back to the beginning. Forget fortifications and artillery etc until you can simulate a prehistoric battle between two neanderthal tribal groups in a truly realistic way (something I currently don’t beleive you seem to be capable of). I know that no one would do this because it doesnt seem to be financially beneficial, but if you really are in for the long haul as you claim, then I beleive it would benefit you to do this. I hope to one day play a truly realistic, challenging and fun artificial opponent, so good luck with whatever you choose as your next project. I would suggest, however, that you make sure it works before releasing it.
Posted by jimmypops on January 21st, 2010 at 2:56 pm
actually thinking about it – its not just the battle ai that doesnt work. the map/strategy/diplomacy AI is also pretty ridiculous (especially diplomacy). Sometimes the decisions that computer opponents make are absolutely baffling.
Posted by jimmypops on January 22nd, 2010 at 2:22 pm
Can I just say that I am far more concerned about the diplomacy AI than the battle AI. In Empire the diplomacy AI has ruined several of my single player campaigns. Things like old allies, with whom you have traded and exchanged technology with for many years, suddenly declaring war on you for no reason really annoyed me.
The majority of the time they are also completely unreasonable, expecting two or three technologies plus some money in exchange for one of theirs. They won’t accept peace even when you haven’t fought them for 20+ years etc. I could go on.
Please can you do a quick blog explaining the improvements to diplomacy AI since the last game? For me this was the one stand-out area which needed improving.
Posted by Andrew on January 27th, 2010 at 1:11 pm
and there are some good points in this blog though because i do like the look and being that i live in the land of second chances i probably will get suckered into buying napolean with the hopes that it will be better but more than likely you will recieve yet another outlash from me 6 months from now, but just look at it this way at least i treat you nice unlike the manner i talk to those relic people.
Posted by Michel on January 28th, 2010 at 10:11 am
just a suggestion, would it be easier to create a few different attack and defense scenarios or doctrine for that matter to be copied and applied for the ai to follow as opposed to fixing something you say you are kind of in the dark, i mean i understand this dude wrote the program and split for personal reasons, im not just sounding off i am trying only to state what i would describe as greivences, for real, is this going to turn into another med2 issue where not until the entire community signs a pettition before you listen, because if so let me know and i will get one started.
Posted by Michel Ney on January 28th, 2010 at 10:23 am
I am a big fan, almost every nights I play around 2 hours with ETW, I have ordered NTW because I’m a big N fan too.
Multiplayers it’s a great idea but we need more powerful PC to use it .
I can see with my AMD quad-core that only one core is used by ETW,
please could you tell me if I miss something to use my full capacity quad-core,
my graphic card is based on SLI nvidia technology with 2×8800 GTX but sometimes during campains or battles my pc lag badly, I play ETW with XP without problem but with Vista the game crash, so I use it with XP
So can you confirm that NTW use SMP ?
Posted by cervicad on January 28th, 2010 at 4:47 pm
The problem at CA is they take more time with the Graphics, they dont care about the AI never had. If they did they would poor more resorces into making a better AI, but they dont. See all I have to do is download a mod with XAI and have a totaly different experiance with a better AI to know it is the truth… CA is taking us for a ride with the AI, they oviously dont give 2 cents about there fans, who would appreciate if they would sell what they advertised. Best AI ever, here is just a Quot from CA regarding the AI before release.
We were very aware early on that the AI would come under close scrutiny by our fans, and that’s why we’ve made this a major priority with Empire. We believe that [the game's AI opponents] should be the most real and challenging yet.
In terms of battles, Empire has expanded on the AI in two significant ways. The first is the introduction of a new plan-based approach where the AI can dynamically reassess and rebuild its tactics as the battle progresses. This gives it a more human-like feel, as it picks and probes with individual groups of units, as opposed to having a template of stock tactics for the whole army to run through, which quickly becomes predictable. The second is through the use of “special tactics” specific to certain nations and/or generals. It is also aware of the context of the battle within the campaign. Of course, the focus of battle gameplay has shifted, with ranged weapons becoming much more important, and the AI has undergone a major overhaul of the way it responds to and attacks with missile fire.
The campaign map AI has also been completely rewritten and acts to achieve longer-term goals and specific war aims.
We have had three talented programmers working full-time on nothing but AI throughout the project–a level of resource we haven’t had before.
but, only recently JACK COMES OUT AND SAYS this:
Just before the end of Empire the lead Battle AI programmer left CA to return to his family up north. Unfortunately, thanks to Mr Wilberforce’s efforts 200 years ago we couldn’t stop him. It left us with a battle AI, which at that stage, struggled to beat good players in a fair fight, and was pretty much at the mercy of great players, even with a level of handicap (I call it cheating) that is all too obvious.
Hmm one person quits and the AI goes out the Door. HMM what happened to the other two odiously they were not helping if they could not continue on with the AI..Plus why wait so long after the game was released before they come clean.
Posted by Leatherneck918 on January 30th, 2010 at 6:39 am
I started with Shogun, which was the best game ever, until Midieval which was.., until rome which was.., until midieval 2. I wasn’t thrilled about the switch to gunpowder based warfare but couldn’t wait because it would still be great. But I have not bought it because I worked for SEGA as a tester and they are really don’t seem to give an expletive about games, just sales. Thier products seldom are done well, they just publish on time with inferior products again and again and it is a huge disservice to such a great francise. Then again, I have to assume SEGA’s money and money making is why you chose them and continue with them. I upgrade computers just for your games, but will not this time. Figure something out please because I really love your games. PS when you dump SEGA, and get your stuff together, stick with hand to hand.
Posted by John on February 3rd, 2010 at 6:36 am
Support you all the way. One thing though: Mod Tools!
Posted by Maaf on February 4th, 2010 at 8:12 am
the ai isnt just bad it is outright lousy, first of all unless the computer cant deploy in a single line due to obstructions on screen that is pretty much its only tactic, and once that begins it usually takes a good 15 to 30 minutes to adjust that already straight single line another 6 or 7 times, once the advance has begun the next issue would be the out flanking,for instance when i present a line against the advancing ai and start a skirmish if i were to bring units on their rear, i.e. cav charging enemy cannons in the backfeild, the ai in all its wisdom decides it would be best to pull its entire line from the face of battle to attempt to chase my cav to the red line with their infantry, thats right chasing horses on foot.
while on the cav issue there are a few more topics i would like to adress, the first being the pathfinding. for whatever reason the smallest obstructions and sometimes no obstructions will cause the cav unit to take its sweet time to form a single file to pass it after i manually put it in a 3 rank formation, the second of these issues would be the charge bonus that you rarely recieve, there are a slew of other issues involving player and ai cav to go on about but i do not want to get to far from my main point here.
other issues with the ai and their single line battle tactic would have to be the annoying exploit of the instance of the player leaving cracks in his and or her line and the ai advancing straigt into mellee due to the fact that the entire ai army kept advancing for the crack of about 1 to 2 inches of screen space without adressing the formation of the other 19 units on screeen. a good example of this would be to set your infantry in a scattered line with guns intergrated to the line as opposed to concentrated at one spot, the ai will walk right up through the face of fire with out stopping to fire only running up like children because it seems as though the computer cannot recognize that the main line is fixed but a few inches foward, in otherwords to make this short my argument is why even equip the ai with guns if they arent going to fire them.
the next ai issue would most definitly have to deal with the actions it makes on the campaign map, thus being, if the ai were to iniate war with the player, as in the ai declaring it through diplomacy, instead of presenting your army or armies with a feild force, the ai will break up its army into little raid groups, unless the player holds a near by territory that is ill defended, for example if the ai has 20 units the only way it will attack is if it clearly outnumbers you otherwise it will divide the army into single units and just raid, which is ultimatley nothing more than nuissance value.
the second of the campaign issues would be the diplomacy, its as almost you the devolper are trying to pass of insults to the consumer via diplomacy by way of the ridicoulus offers it makes from its pathetic standpoint against the players usually indomitable army or armies. my point being the ai’s ridicoulus demands for outragous sums of money for peace or unbalanced tech trades or even teritories for that matter, and their unwillingness to accept reasonable terms of agrement.
the sad truth of this is that this is only the tip of the iceberg when it comes to all the problems that need be tended to, and that i could go on but i am to tired therfore tommorow i shall write yet another complaint until these issues have been properly adressed.
Posted by Michel Ney on February 8th, 2010 at 6:28 am
I appreciate that coding a realistic AI is incredibly difficult, but like many others I am suitably irate at the somewhat unfinished state that ETW was in on its release and still is now. That said I only experienced brief instability that sems to have plagued so many others, thou I have had too many corrupt saved games which really try your patience.
Rather than trying to hard code strategies directly have you considered using neural networks or genetic algorithms to ‘learn’ tactics from being trained on battle replays? Just a thought…
As for not updating ETW, I think you would lose a lot of dedicated fans of the Total War series if you weren’t to update the “worst behaviour” at the very least. I appreciate that the Demi-Gods at SEGA would rather you concentrate on the next commercial release, but the reason the Total War series and CA have been so successful is the quality of the games. ETW was possibly the worst quality release in the entire series and to leave it that way would be an affront to everything I had previously thought of the CA team.
Posted by mrwobbles on February 8th, 2010 at 5:37 pm
Take Command: 2nd Manassas, and Sid Meiers Gettysburg, are examples where the battle ai works well unlike ETW .
I know nothing how AI works sadly but i have some suggestions how i would like it to work. If defeated the AI will copy the moves the victorious human general used both tactically and army composition. At the moment when defending the AI will move its formation under fire so they get murdered, did the same in M2TW.
Or another idea is the AI will mirror the moves of the human general during the battle.
Posted by Grumpy on February 8th, 2010 at 9:06 pm
First of all I would like to congratulate the Creative Assembly for all their work in the field of historical Strategy games.
However, I have to admitt that the last version -ETW- was profoundly easy, not only during the battles but also in its Political aspect. How much sence does it make to be able to win the campaign in no time, playing even with the weakest Nation? I can appreciate the fact that creating a stronger Battle AI is difficult, but what about the political reactions of every nation in the game? For example, history teaches that whenever a nation starts becoming more and more powerfull, the other Nations tend to create alliances against that particular Nation, and the Napoleonic wars is just one example. This is called foreign policy of an elementary level! Not to mention other financial, religious, territorial, cultural issues that could ignite a war or an aggressive state or an Alliance of a GROUP OF NATIONS against others.
Therefore, not only the weak battle-AI of ETW, but also the way too simple MAP (very few towns,nations), the few choices the player has when NEGOTIATING with other nations, the PASSIVE status of opponent nations, the absence of an “unknown factor” during the gameplay, lead to a game promably designed for KIDS in the AGE OF 12 rather than over 16!!!
The example of EUROPA UNIVERSALIS, showes that a more complicated Campaign AI is possible, aiming the more experienced players, unless of course this is was NOT an ambition in the first place, and you ,the designers, just wanted to create a game for kids, based in “realistic” battles…which of course is yet to be achieved, as the maximum number of soldiers in the battlefield of ETW is approximately 2.ooo , which is less than 1% of the troops of a Campaign in Napoleonic Wars after 1805 (approximately 240.000 ) or -if we want to be more fair- a 4% of the troops in the average Napoleonic Battle ( no less than 50.000)
In conclusion, the game ETW doesnt resemble the wars of the 18th and 19th cent. NOT EVEN CLOSE, and its sad to admitt that the interest of the experienced player will be exhausted after a few hours of play.
However. there are ways to improve the game, by developing a more in depth Strategic analysis.
Posted by kostas on February 13th, 2010 at 12:43 pm
since the last update the “imporoved” the VH/VH ai, the game has been simply ridiculous. Recently when i was united provinces attacking the almighty maratha, i sent down 3 full fleets of heavy first rates to take out their ships (which they only had 1 full fleet of assorted brigs, 4th rates, and 6th rates) and after the battle, i had won but it cost me 24 ships which is unfathomable. its almost as if my own ships just became bored and started shooting at eachother. this oddity repeated itself in other instances of the same scale and setup. Also recently when I was attacking Bijapur i was seiging with two full armies, one consisted of line inft., grenadiers, mortars, and cavalry, and the other consisted of all hollad guard, blue guard, scots, and swiss line, and mortars; after the computer had played out the battle, i had lost half of the units out of both of my armies and lost the battle overall. and maratha didnt lose any units.
Posted by whisley on February 14th, 2010 at 5:30 am
hello empire community. In this blogi see almost everyone complaining about the AI and its being stupid. But I am here to complain about the reaction the the retarded AI. This update CA came up with has absolutely ruined my campaign expperiences with Empire Total and let me tell you, it is a darn shame. I want a reallistic experience, not seeing Marartha in the Caribean or taking Hannover like i have seen. I am posting this blog to see if anyone else is on the same page as me. PLEASE REPLY IF SO!!
Posted by whisley101 on February 14th, 2010 at 6:12 am
I really don’t understand the surprise at this situation.
It’s not new, it’s the same situation the Total War game has found itself in every time it’s re-released.
They release a game, it has significant major problems, but is still a commercial success (or at least it seems to be, since they keep making more without improvement). The significant major problems are given lip-service, some patches do a few things, then an “expansion pack” is released that claims to fix the problems…but doesn’t really. Then a patch or two to fix new problems, leaving original problems unresolved, and the programming power is moved to making the next game.
This has been the model followed since Shogun, and has been successful used to make money every time. It will continue to be used because, despite the consumers disgust and dissatisfaction with every installment of the game, we continue to give them our money to give us broken games.
I’m an old-time gamer. I remember when Shogun was first made, and the emphasis that was placed on touting it’s battle a/i which was supposed to have been programmed using Sun Tzu as it’s guide. Interestingly, the first game had the best (albiet still barely passable) A/I…here we are two decades later, and the same issues still exist.
This is not so much a game franchise as it is a marketing one.
Posted by b_roda on February 23rd, 2010 at 11:05 pm
Wow, iam not surprised to read exact the same phrases again…and again…They had problems with the AI since…ah..since the start of the series. For me the best AI was in :Rome, the worst in TW:M II. I dont think there is much of an improvement in the game from the last one…
Same thing with Multiplayer Campaign…this feature was first announced for the Rome Add On, but never saw the light….
But, this time, there will be no money from me.
Posted by Lol on February 24th, 2010 at 5:29 pm
Been pLaying from start, Shogun, like many of you guys. So far Rome was easly the best part. the dept of the game was so much better than ETW, even if BAI wasn t that great, you really got feeling of beeing part of conquering Rome. With ETW i always use same line infantery, (forget about art) it becomes boring very fast. Neverthelss this is still the only game i really love playing just hope NTW has more dept and a better BAI ofcourse.
Sorry about bad English.
Posted by peter on February 25th, 2010 at 11:13 pm
Hi,
I’d just like to say what a great job everyone at Total War is doing. I haven’t got Empire: Total War so I don’t have the most recent experience of TW gaming but Rome: Total War was a triumph. It was made in 2004 and I would still rank it as a top 20 game on PC. It’s far more adventurous than any other game of its time and has a stunning level of detail. I’m very much looking forward to purchasing Empire: Total War and am hoping that it will be even better than R:TW. There is a bit of a problem about what to do after N:TW, but maybe you guys could come up with a WWI/WWII/Victorian combination pack, but conscription-size armies and guerilla, aerial and trench warfare will be a huge challenge. Not to mention machine guns. I’m sure you guys will come up with something, though and it will be even better than N:TW. Anyway, thanks for making some of the best games of all time and I hope that you never stop making TW games.
Posted by Dafydd on February 27th, 2010 at 2:10 pm
OK, those people complaining on this forum about the BAI need to shut up. The finished product for NTW is not that intelligent–if you know what you’re doing, the only way for the AI to present any challenge is on VH/VH. And that’s how it’s always been with the TW series. Amen to improvements in this area, but I for one will continue to look to human opponents when I’m open to the possibility of getting my ass handed to me in a TW battle.
Posted by Reid on March 11th, 2010 at 6:09 am
Artificial Intelligence is not intelligent without the possibility to learn.
What is learning?
-Learning is a very large field of different behaviors and contains the observation and interpretation of organized and selforganized processes in order to understand which action is random and which has a goal, what is this goal and how can this goal be achieved and prevented. Learning is predicting the future out of information you can gather in a system with interacting with this system, identifying and remembering the constant and the changing parameters in your induced dynamics, it is the comparison and calibration of your established memory (called knowlegde) and the transfusion of that into a new and unknown situtiation with identifying similarities and differences. The important part of the learning ability is to judge a sitution right in order to know and evaluate all your possibilities, advantages and disadvantages to enforce one or many goals and to evaluate and remembering if the judging was right or a prejudice. Maybe the simplest learning process (and therefore the simplest ai base) is the trial-and-error principle with a very simple remembering function. This is a very very slow learning process with (first) totally random behaviour (the success depends on the remembering function).The idea is: something is better than nothing. Another very simple ai base is a behavior leaned from primitive animals: swarm and flock tactics. This strategy is a bit learning-resistent and comes from the principles: together is better than alone. You can try to enhance one of these strategies with moderate (micro)tactics (this unit beats that unit, concentrating fire and so on) and try to point out some stage goals et voilà this what programmers now call ai is born…
Maybe i had given a small glimpse at how complex the topic ai can be and i believe that in the long run the development of an ai that deserves the name will be more complex than to engeneering rockets, because it is near to the question ‘what makes man capable of engeneering rockets?’. Try to enjoy your superiority over the machine so long it lasts, i have a dream, that some day it will ‘deep blue’ you in categories you didn’t think of
Posted by Chaos13 on April 2nd, 2010 at 1:54 am
Dear Mike,
I don’t normally post but I couldn’t resist adding my two cents:
Posts 113 and 120 among others, had some points I would have to second. One had a discussion on scale. I can understand limitations on hardware, but I would suggest then making cities to match that scale. i.e. both Empire and Napoleon a town will have a population of over a million citizens, but then can only field an army of 2500 at most? Doesn’t make much sense. Rome was better at this, esp. with RS mod you had to watch your populations because if you got drawn into a heavy casualty conflict it could really effect your pop. When I play NTW I have to tell myself “well this regiment in reality would represent this many men; it’s scaled down.” Which is silly but I think it helps me communicate my point. When it’s the era of mass conscription, and you’re total forces are mb 5K men? That’s weak sauce, up the ante. Napoleon invaded Russia with over 600,000. I’m not saying match it but there’s modders that have unit multipliers. why not save them the trouble and make your own that works better? I’m just saying give us that epic feel.
The main thing I’d like to cover is the main topic of this thread: battle AI. I think we’re all in agreement that frankly, it sucks fat donkey ballz. In addition, I got no clue what you’re talking about NTW having improved AI. It’s just as bad as Empire! Artillery shooting into the side of a hills, inidividual units of infantry and calvary (as well the general!) doing suicide charges at random, and don’t get me started on seiges. It just stinks. There’s absolutely no cohesion whatsoever. That’s my biggest thing the enemy units don’t even attempt to work together. A quick remedy is to use human opponents but I find in my campaign battles that CAN be played online can NEVER locate a player unless I want to sit there for an hour til it finds someone. My patience never lasts that long.
I also second whoever came out and complained about the DLC. I believe they said “nickel and diming us” was how they put it. Why don’t you just give us what we paid for the first time and focus on the main product instead of all the expansions and new skins, and all that other junk that is just fluff and never addresses the main issue.
So I won’t list a slew of problems without suggesting a solution. I’d say listen to the feedback and input from the modding community not to mention you might find someone to fill that vacancy in battle a.i. programming. A word to the wise, you all at TW have some of the best and most loyal fans I’ve ever seen. The community and sharing and just general discussion on sites like twcenter.net and twheaven to name a few are unlike any fan community for a game series. So don’t take this or any of these post as us just complaining. Sounds cliche but we post cuz we love; we want to see the games succeed and become everything they’re intended to be and more. And with this great fanbase and mod community comes such a unique advantage. You should never hear the excuse at CA, “well we didn’t know what the fans wanted.” And with that said; when can we be expecting Rome II: TW? Food for thought. Keep up the good work.
Posted by TotalWarrior4Life on April 7th, 2010 at 9:07 pm
Much as I love the TW series in general I gotta say you people have got some hide the way you release unfinished games before they’re ready and then expect us to endlessly patch and mod to fix bugs mistakes that should have been ironed out before. As another observed, I expect to play the game out of the box and get a completed product.
As to the AI, I think its true that Shogun had the best AI and there has been no improvement since. the opposite in fact. Perhaps too much energy has been devoted to the games eye-candy, making an attractive looking but deeply flawed product.
TW Rome perhaps had the best balance of game play & looks.
Considering all the negative reports I’m unlikely though to fork out for the latest releases just now. Perhaps I’ll wait till they hit the bargain bin
Posted by Reverend Hellfire on April 11th, 2010 at 11:02 am
Hey guys great game first off, and while its good you guys addressed the AI and I love the new units but what i really want is some NEW MAPS! I would especially like some river maps like in Rome and Medeival Total War, as well as the ability to build fortifications on the play battle maps. Thanks and keep up the good work.
Posted by Ross Johndrow on April 14th, 2010 at 12:07 am
Hey first off awesome game, but i have a problem in the AI. I lag a lot in my game even though it is on low graphics so i am slow to read a battle and i lose almost every battle… I was wondering if there was a way to get a either lower AI or lower graphics. Its not fun when you love a game but lag so bad it hurts you so much that you almost cannot play the game… any advice?????
Posted by toddrapp on April 20th, 2010 at 6:02 pm
You know what would be great? If the people complaining about bugs and crashes ran a spellchecker before posting their rants. It’s nigh on impossible to read most of the comments around here.
You know what else would be great? If the developer could trust the so-called hardcore fans to buy the finished game after releasing a beta version. Since they can’t, we have to put up with the “release-now-patch-later” approach. With the infinite range of system configurations and servers currently in use (not to mention all the bugs, viruses, and design flaws in the typical home computer), how else are the developers supposed to detect problems before a wide release?
The TW series is epic in every sense of the word. Not just ambitious, epic. The historical research alone, translated into each game, is enough to scare off most developers. Many of the above comments are akin to complaining about a literary masterpiece because of the paper it was printed on.
Posted by SirDennisC on April 23rd, 2010 at 4:40 pm
Thanks you just free up space on my computer as I deleted Empire.Although patch 1.6 was a improvement,your messing again with the economics was the final straw.As it makes completing missions in your game a joke at best.Although money wise I had one of my better games,trying to reached your objectives now is a joke.Russia an Poland acts more like one country instead of two as neither really fight each other even when they’ll at war.
Then lord an behold if you don’t do something in Indian early the Maratha’s becomes a naval superpower that equal or pass England,making Spain navies looked second class.Heck by the 1770′s they even had a stronger an bigger navy,already building the best ships you can build in this game.Something I find laughable as try as any of the people which I know which has this game we never been able to steal the tech that fast as the game dictates we do other thing or our gentlemen are killed or unsuccessful.Only game I know where Spain navy took a back seat to anyone except maybe England.But then in this game I’ve seen Russia Sweeden,an the Maratha’s navies outclass Spain,France an sometimes England.Heck even when I play the Maratha’s I can’t get there that fast.
Seen Napoleon at friend,and as he says after a month the game became boring as it was to confining.Once you won once their wasn’t that urged to played again as you knew what to expect therefore it was easy next time.As he is like me and not a online player,as cheating or using cheat codes isn’t our idea of playing as some of the people he try it with online did as their armies were never ending and with what they own didn’t match the money they seem to be getting.20,000 or more and thats after paying for the cost of 6 full up to date armies stacks an then 3 to 4 stacks of up to date warship fleets not counting there trade fleets in each trade theater.
Posted by wrc on April 26th, 2010 at 6:01 pm
One change supposedly not so difficult to make in AI battle concerns the behavior of generals: please, avoid their suicide in charging artillery, by increasing their attitude to stay behind the lines, give them the chance to get off a horse to stay in a square, allow artillery fire against them to be effective only in the case of a direct hit or reduce the priority to choose generals as artillery targets.
Posted by gregorio on May 8th, 2010 at 10:31 am
Mr. Simpson, I see a lot of complaints regarding the AI. I also see that in your blog you mentioned that your key AI programmer had to leave after 3-years of work. I own Shogun, RTW and recently purchased NTW and have been emersed in it for over a month, just completing the SP Campaigns of Napoleon. I own a copy of Weapons and Tactics of the Napoleonic Wars and have been re-reading it as I’ve been playing the game. IT IS AMAZING how much accuracy and details of the weapons systems and tactics of this period that your team has built into this game. While there are suicide attacks and piecemeal units thrown into battle, occassionally, with a strong general, the AI can put together a strong fight. I’ve lost very few battles in SP mode, and usually because I didn’t follow Sun Tzu’s principles in Shogun, however, “the feel” of the game is outstanding. I will be playing this for a long time to come…and I’m still playing on the medium difficulty setting. Keep up the good work and the improvements.
Posted by DocWells on May 13th, 2010 at 2:52 pm
Although sometimes the AI has “interesting” moments, overall I’m fine with it. Sure it doesn’t match the skill and originality of a strong human opponent, but I think that would be expecting a little too much from AI. A huge amount has been achieved in Empire and I don’t think we should belittle that.
The flaws will be ironed out im sure as the series develops. One thing I think perhaps should’ve been included, which has now been announced for Shogun 2, is the effect of coastal areas on naval battles. I live in the Channel Islands, which were not represented in Empire at all, presumably because of their size, however they were a keypoint in the battles between Britain and France. Coastal fortifications should have made an appearance in Empire, as well as rocks and reefs which a cunning admiral could use to his advantage.
Anyway, keep up the good work. Im keeping a close eye on Shogun 2!
Posted by chris on June 15th, 2010 at 5:35 pm
Hey Guy i just noticed you had an update to the games AI, that well and all but does this update include a change in the number of men units carry? To tell you the truth I feel the GAME HAS BEEN RUINED!!!!!!!! Please get back to me i do not know what happened nothing ever asked me if i wanted some update that was only gona let me mannage like six ships per navy or 40 men per unit. WTF!!!!! Please get back to me i loved the game as it was before, the AI was ok, not perfect but not broke. Nothing to fix!!! just get back to me please i want my game back i FEEL AS IF I HAVE BEEN ROBBED.
Posted by Juan Paz on June 18th, 2010 at 6:16 pm
Hm not exactly my topic but still nice reading
Posted by No cd cracks on June 28th, 2010 at 3:36 pm
This thread simply confirms to me what I have long suspected with the TW franchise-there are two types of “fan”-those like me who are interested in the historical periods presented,and those I term “gamers” who sole aim is to be on some high scoring chart…..
I suspect the “gamers” are generally ones who have never read On War, The Prince,or the Art of War. If they had then I wonder if they would be such a hue and cry regarding “realism”. I suspect “gamers” load up armies that couldn’t ever reach even an approximate realistic depictions of their historical counterparts. “Gamers” have armies made of maximum experience and the best troop types to beat others so they can claim to be the next Alexander or Caeser .
Thats fine…..each to there own.
However “realism” in warfare involves friendly fire,troops marching in the wrong place, refusal to carry out orders ,cowardice. “Realism” is not neat and tidy, realism is chaos and bloody anarchy. Clausewitz called this friction-perhaps the mishap with the AI indirectly represents that.
So in light of that I would suspect that the howls of protest from some of the posters on here demanding “realism” actually want a game engine that is nor totally “realistic” but simply a game, which TW in all its guises is.If we were to introduce that level of “realism” I suspect that very few people outside West Point or Sandhurst would actually play the game because it would really be unplayable.
I think there is always way to improve things-but we should be careful what you wish for. That a Game Developer has put himself on offer says something about how the CA team values it customers. TW is a fine product, and is a intelligent thoughtful product that introduces people in historic periods-and hopefully encourages further understanding.
If those who feel as though they have been “robbed”-I suggest set up your own software house,employ the best in the industry,and produce a product that beats TW…..I can assure you, you will not have time to write poorly written posts….
Posted by Mr Savvy on August 8th, 2010 at 12:36 am
I think that the scale the team puts in to theses games is fantastic like with medieval 2 i have the DVD making and they put a all new bench mark for tactical game play GOOD JOB!
Posted by Harland on August 17th, 2010 at 7:38 pm
Shogun I was a good game. Battle AI was simple, an army made of regiments arranged in a variety of formations, face to enemy army, go for it, release regiments and every one is to attack its objective. There was a number of stupid behavior like generals charging from the very start and regiments running up and down for no reason, but it was a good game and its sequels were a great opportunity to improve it. The strategic AI needed much more work but everyone understood then that the strong point of TW games was the tactical battle. Since then, every TW game has been more of the same. Superb graphics, mediocre tactical AI, dumb strategic AI. I’m sure it’s not that hard to code a pseudo intelligent behaviour for a few abstract entities like regiments or groups but i guess that doesn’t sell games. Graphics sell games so graphics is what you do. The rest is a bluff. But hey, i’m not complaining, for a few moments one is watching the screen and is amazing to feel the appearance and sounds of an ancient battle.
Posted by Aitorinsky on November 5th, 2011 at 11:39 am