The next game in the total war series?

[fomalhaut]

Those are solid ideas, but i don't think most TW players want to be so limited in their control. And the Greek Golden Age isn't that different to most consumers from the post-classical age of RTW, so it just wouldn't be viable.

[moonburn]

well there´s only so much one can do before it deplets the game ideas so i think this would make for a very sollid more realistic game altough not so appealling to everyone (particulary mindless kids wich seem the target segment of sega )

but somehow i see it working for a more specific crowd or then as a kind of expansion pack for a future rome total war 2: the greeks wars

[Populus Romanus]

I have gots the bestest idea'er evar. They should make EBTeam:TW!
Simply copy EB, claim it as their own, rename all the factions with the names of EB Team members, give them special units, and presto!, you have EBTeam:TW!

For instance, Foot would be Hayasdan, Atilius would be Romani, bobbin would be Eleutheroi, Moros would be Sab'yn, and Oudysseos the Casse!

It would be AWESOME!

[vartan]

As a fan of science-fiction, astrophysics, astronomy and many sciences, I would very much appreciate a revamped Total War game with a completely new interface and gameplay set in the future, even if in our own galaxy, with planetary systems at stake, and humanity on the possible verge of destruction if one does not act swiftly.

[Duguntz]

well there´s only so much one can do before it deplets the game ideas so i think this would make for a very sollid more realistic game altough not so appealling to everyone (particulary mindless kids wich seem the target segment of sega )

but somehow i see it working for a more specific crowd or then as a kind of expansion pack for a future rome total war 2: the greeks wars

Well moonburn, as well as we'd liketo see something like that happen, it won't. each and every game must be conceived to touch the most general public as possible, and not to target anyone in particular, in ordert o max out the number of sells, you pointlessly make a new game if only "connaisseur" buy it, as you'l loose much more money making it than you'd win selling it. hence, the game must be interesting, but enough hollywoodlike to attract as well the mindless kids that only wants to see thousand of soldiers hit each other with swords, no matter where, when or why!

[Cute Wolf]

well, Empire is CA at lowest ebb of TW gameplay, luckily for us, Napoleon are much more better.

[Fluvius Camillus]

Tomorrow S2TW collectors edition will arrive, I have to see if I'll play it, Empire drained my energy to play TW, might take up something different and play it later.

And I didnt even got to playing NTW, which everyone tells me is awesome... Busy life...

~Fluvius

[Duguntz]

Tomorow is the day I was breathing for since i know shogun II was being worked on...

[Cute Wolf]

shogun 2 is good, worth it, but damn, the AI is soooo easy to beeeaatttttttttt
(sun tzu? my ***, simple flanking tactics works and they are even dumber than RTW AI)

PS: Steam servers is noticeably works much more better

EDIT : maybe it was me that found the AI was too easy, years of blitzing RTW mods in VH battles difficulty allready took their toll

[Lazy O]

Ive watched videos, the AI stinks

[Duguntz]

What did you expected about AI? I don't wanna put any bad opinion on CA, but guys, you can't expect any REAL challenge while competing against a machine! it's still only codes! AI will NEVER be as challenging as an experianced human player. because even if well done, we can predict it. So stop blasting AI because of expectation impossible to acheive...

[Fluvius Camillus]

I played it a few hours with the Takeda, but didnt have time to start a serious campaign. I'll check it out later.

~Fluvius

[vollorix]

A true artificial intelligence has jet to be invented...
I wonder, though, if a battle "AI" would be programmed and run on kind of powerfull "supercomputer", how frustrating it might turn out to compete with it? If a chess computer can be created, why not a good AI for a game? :)

[Lazy O]

I don't wanna put any bad opinion on CA

That be a sin. I think its not that difficult to ATLEAST get the AI to do so much as flank, form a decent formation, pick sensible armies etc....No ones asking them to make an AI version of varus....

[moonburn]

That be a sin. I think its not that difficult to ATLEAST get the AI to do so much as flank, form a decent formation, pick sensible armies etc....No ones asking them to make an AI version of varus....

considering it takes me 5 minutes to rout an elite army with medium cavarly and militians while varus defeat took 3 entire days before he was defeated even varus is like 10 stars above the ai

and as someone stated how hard can it be for the general ai to build up "sensible stacks" with cavarly archers skirmishers and somewhat decent cavarly and then use them properly in the batle ... i mean the ai works greats for celts it´s all about clash and trying to break the enemy as fast as possible but for a roman army for instance they should try to tire up the enemy batle lines retreat the hastati use the princeps and try to win with them while moving cavarly and ascenssi around not to get flanked and not send their generals to fight against superior cavarly without at least a group of rorarii coming up for the suport in time (and thats when i decide to make it challanging and not let their generals die in a mass of infantry while my skirmishers shoot them in the back and my own cavarly is killing off their skirmishers and then turning around to clash against their bactleline backs)

[Lazy O]

Varus was/is one of the best RTW players of all time, I was not talking about the roman general :P

[vartan]

A true artificial intelligence has jet to be invented...
I wonder, though, if a battle "AI" would be programmed and run on kind of powerfull "supercomputer", how frustrating it might turn out to compete with it? If a chess computer can be created, why not a good AI for a game? :)

Classic myth: that competence in AI formulation for chess means competence in AI formulation, period. I've studied the history of decades of work in the development of chess algorithms and chess hardware, and I can tell you that the resulting product is very specialized, and rightfully so. It does a damn good job of what it was meant to do. Unfortunately, this doesn't translate into a "battle AI" for anything else, whatever that may mean anyway.

That be a sin. I think its not that difficult to ATLEAST get the AI to do so much as flank, form a decent formation, pick sensible armies etc....No ones asking them to make an AI version of varus....

Already done (see games in the Total War series by the Creative Assembly).

[Lazy O]

I must be blind then.

[vartan]

I must be blind then.

1) flank, 2) form a decent formation, 3) pick sensible armies

1) Observe the computer's movements in a custom battle. Computer chooses to ignore a good portion of your center and instead focuses on either flank. The computer surrenders the center to its own detriment. The moment you begin the battle the computer moves its entire army to one side, usually its left (your right), unnecessarily.

2/3) What is a 'decent formation'? There are no decent formations, only decent ladies and gentle men. Same applies to sensible armies. Moreover, the computer cannot be sensible in the sense humans can (no pun intended, really). Computer has some problems it can overcome, many it cannot.

[Lazy O]

1) No, most often it just targets one unit and the rest can be surrounded in seconds.

2) Thats a Darthmod invention, not included in Vanilla, very often they tend to form a blob

3) AI often send 2/3 unit stacks to besiege fortresses, picks levy armies, and almost never recruits cavalry.

[vollorix]

If you battle the AI for a prolonged period of time you could see it adapting to your army composition. But that requires some additional circumstances like enough money and availiability of apropriate units. Once i´ve started to harass Macs with some mostly cavalry based army in the Thracian mountains, they, after a couple of heavy defeats, have managed to field a counter force consisting heavy and medium cavalry ( Thessalians + Prodromoi ) as support for their heavy mobile infantry like Agrianians, Thureoporoi and Elite Peltastai. No phalanx units were send again in my direction untill the AI was planing to conquer Tylis. This time i was playing on VH campaign since on medium it was a catwalk to reach Athens by the year 250 bc, which seemed very unrealistic, and dissapointing, to me at the same time. It seems to me that AI needs some kind of "reaction time" to respond with alternative army compositions, but if one moves to fast, and to succsesfully, forward, destroying it´s infrastructure and limiting their recruitment bases, then one would mostly see that "levy spam", or simply a mix of units supposed to be superior than the players ones according to their stats, while 1 mediocore archer unit with low morale and melee skills, combined with a "hard hitting" cavalry, would easily beat multiple units of Pandas or Lugoai...

[Gaius Sempronius Gracchus]

If you battle the AI for a prolonged period of time you could see it adapting to your army composition. But that requires some additional circumstances like enough money and availiability of apropriate units. Once i´ve started to harass Macs with some mostly cavalry based army in the Thracian mountains, they, after a couple of heavy defeats, have managed to field a counter force consisting heavy and medium cavalry ( Thessalians + Prodromoi ) as support for their heavy mobile infantry like Agrianians, Thureoporoi and Elite Peltastai. No phalanx units were send again in my direction untill the AI was planing to conquer Tylis. This time i was playing on VH campaign since on medium it was a catwalk to reach Athens by the year 250 bc, which seemed very unrealistic, and dissapointing, to me at the same time. It seems to me that AI needs some kind of "reaction time" to respond with alternative army compositions, but if one moves to fast, and to succsesfully, forward, destroying it´s infrastructure and limiting their recruitment bases, then one would mostly see that "levy spam", or simply a mix of units supposed to be superior than the players ones according to their stats, while 1 mediocore archer unit with low morale and melee skills, combined with a "hard hitting" cavalry, would easily beat multiple units of Pandas or Lugoai...

I noticed the Ptolemaioi doing the same in my Sabyn campaign; because I used Sabyn noble cavalry (and Ethiopian medium cavalry) quite extensively, after a few battles they started sending stacks with more heavy cavalry. The battles became harder and harder.... of course, it didn't help that they had taken over half the world and the only other surviving powers either kept a peace with each other or became protectorates of one or other of the big guns. But, the composition of their armies did alter.

As to the question of what will be the next TW release. Shogun was the first, wasn't it? Rome followed and, let's face it, Rome sells. I wouldn't be surprised at all if the next release was R2:TW.

Wouldn't it be great if they contacted the EB team for some input? Now, if they (CA) make the mod tools available that they say they are going to then R2:TW could be a god-send, with the availability of more factions and naval battles....

[Rahl]

1) flank, 2) form a decent formation, 3) pick sensible armies

1) Observe the computer's movements in a custom battle. Computer chooses to ignore a good portion of your center and instead focuses on either flank. The computer surrenders the center to its own detriment. The moment you begin the battle the computer moves its entire army to one side, usually its left (your right), unnecessarily.

They don't do that every time and "bugs" like sending single units to attack an archer or your generals unit standing before the battleline can easily overshadow such good moves. But you're right, I recognised that too. Most times it makes me angry though, because it destroys my plans and forces me to use more micromanagement, but at the end I still beat them easily. Though, that's a problem of mine and not one of the AI, because it didn't made it wrong when the battle became harder for me and I think it's similar for other players.
When the battleline don't clash tidely it doesn't mean the AI made something wrong. Many players seem to be as annoyed by such things as I, but most times that increases my losses...

2/3) What is a 'decent formation'? There are no decent formations, only decent ladies and gentle men. Same applies to sensible armies. Moreover, the computer cannot be sensible in the sense humans can (no pun intended, really). Computer has some problems it can overcome, many it cannot.

I think the army composition could be improved without to much work. The AI already uses units according to its type (infantry, missile, cavalry...) what would be the problem to force it trying to mix such units on a "realistic" or historical way?

The biggest problem I see is the fact that EB players are usually passionate and often very experienced, but a battle AI for a TW game is very complex and thus hard to make it challenging, especially for EB players. I don't think that we will see much improvement there in the future. Though the Campaign-AI can propably become much better. AI for round-based games seems to be much easier to make, there're some with nearly unbeatable versions.

2) Thats a Darthmod invention, not included in Vanilla, very often they tend to form a blob

Something must be wrong with your install. I see a "blob" not very often.

[Lazy O]

The blob was all to common to me in all Mods, maybe it has something to do with me messing up their plans with light cav, but i dont care.

[Atraphoenix]

A true artificial intelligence has jet to be invented...

Never forget deep blue...
but CA is too poor for such a project...

[fomalhaut]

Shogun seems like it would be quite poor for modding as i'm realizing it is back to the SINGLE culture slot! and judging from the tech and building tree they didn't flesh out that one culture very much or am i just way too used to EB and RS? i haven't touched a vanilla TW in too long not when two of the best mods ever are around

[DeathFinger]

A true artificial intelligence has jet to be invented...

Semi-OT I know but try to defeat an AI in Demential mode in Starcraft II. Pretty challenging, and the AI doesn't cheat. They just were programmed like progamers are playing. I'm sure AI can be improved in this way, even if I'm all except someone able to understand how it could be made.

[WinsingtonIII]

Semi-OT I know but try to defeat an AI in Demential mode in Starcraft II. Pretty challenging, and the AI doesn't cheat. They just were programmed like progamers are playing. I'm sure AI can be improved in this way, even if I'm all except someone able to understand how it could be made.

I don't think it makes sense to compare an RTS AI to a tactical battlefield AI, though. In a pure RTS game, obtaining victory is all about micromanagement and outproducing your enemy. On a Total War battlefield, you can't just expect to produce a blob of units and throw them at the enemy en masse, because the morale system, the importance of proper formation, the effects of flanking, etc. mean that you can encircle and rout a tight blob of advanced units with a smaller number of less advanced units. You really can't do that in a pure RTS unless you are using units that specifically counter the enemy's in the game's rock-paper-scissors system.

[vartan]

In short, Rome: Total War's battle AI is complicated, but not complicated enough.

[jirisys]

In short, Rome: Total War's battle AI is simple, that means it's not complicated enough.

I corrected it for you.

~Jirisys ()