VH should be harder, or "For gods sake give the AI a full stack"

[chickenhawk]

The entire campaign game is easy busy work until the Mongols or Timurids show up with stacks capable of giving a real challenge on the battle map in single player. The AI needs work here more than any where else.

I am waiting for the patch to start any new campaigns but I thought i would ask for a general opinion on the subject.

Should there be an almost unbeatable VVH setting for campaigns?:inquisitive:

[Skott]

Make the game too difficult and they risk losing sales. My guess if you want a really tough campaign you'd need to go to the mods for that. They probably can do it better than CA anyway. :yes:

[Quickening]

To be honest I find VH/VH to be challenging. Im no "n00b" to strategy games either. But seriously, I sometimes wonder if my strategies in MTW2 are fundementally flawed somehow. What with folk saying how they've completed their billionth long campaign etc. I just in the last few minutes completed my first short campaign and I was only 5 turns away from the games end! :dizzy2: Im not shit, honestly.

On VH/VH the AI seems to always have another full stack army lurking around. The only way I can see it being any harder is having an "Insane" setting.

[TevashSzat]

The challenging part of vh/vh is not how eaily you can win a long campaign, but how fast. If you blitz on everyone around you, the game gets alot harder as you must fight every battle with as little losses as possible or your blitz will run out of steam. Just building up for the firsr 50 turns then attacking everyone isn't that dificult. My fastest campaign is my sicilian one which beat the long campaign at 41 turns.

[HoreTore]

Uhm, blitzing isn't very difficult... If you take it just a bit slower, the AI will start getting better and better armies. Just had a fight with the Spanish, they're first stack was 15 heavy infantry, about egual amounts of chivalric, feudal and swordsmen plus 5 units of knights of santiago. A GOOD ai army.

Blitzing isn't a challenge at all really. You'll get PLENTY of money from all your sacking and mercs can very easily replenish your losses. And the AI doesn't have any tactics at all to counter it. Plus, for the first 50 turns, all you're ever going to see, is AI militia armies. Not very hard to beat them with minimal losses...

But there's got to be a way to make this game harder. The problem isn't the campaign AI, it's the battle AI which keeps screwing up. Isn't it possible to give the AI a +5 or so bonus to everything combat stat or something? Like in MTW...

If the battles were harder, and you could actually LOSE an even odds battle, things would be much, much harder...

And about the "people won't play this game if it's too hard" thingy... Well, as long as you have easier levels, that's not true at all. Look at Civ4 for example. The Emperor, Deity, etc. levels are EXTREMELY hard, and civ4 sells more than enough. In fact, if you have higher difficulty levels who are very, very hard, they will ensure MORE sales, because people will play the game longer as it's more challenging.

[TevashSzat]

The problem is that the computer ai will never ever be as good as a human player because the comp doesn't have any flexibility to react to situations as they occur or to think of new unpreidctable strategies.

[HoreTore]

That's why you need AI bonuses. If their troops are significantly tougher than yours at all times, things will get much harder, and you're forced to play much better in order to win.

You may not be able to change the AI, but you can change it's soldiers...

[econ21]

But there's got to be a way to make this game harder. The problem isn't the campaign AI, it's the battle AI which keeps screwing up. Isn't it possible to give the AI a +5 or so bonus to everything combat stat or something? Like in MTW...

Personally, I think the problem is the campaign AI. It just does not have a "killer instinct". Try just pressing end turn in the middle of a war, turn after turn, autoresolving and building nothing. Often not much will happen for ages. I suspect you will get tired before the AI conquers you.

I think CA needs to work more on the AI - on VH, just get it to accumulate a huge, multi-stack army (cheating so afford it and that it can see yours and work out it needs 3x as many or something), then come barrelling at you, taking city after city until it hits your capital. The AI in strategy games like Homm3 or Civ can do this. The AI in STW and MTW did this. But with the RTW/M2TW open map, I think more work still needs to be done to get a strategic AI as dangerous as the one on the Risk style maps.

I haven't noticed the battle AI screwing up that much (unlike RTW). But anyway, a human is always going to beat the computer with even odds. The solution is to tweak the campaign AI so that the battles are not at even odds.

[HoreTore]

Yup, well, making the campaign AI better is the dream, of course. But that takes time and resources.... Simply giving them large bonuses is a much easier way of making the game harder.

But still, an enemy 3 times the size of your army is still something the player can beat with fairly little problems, without an improvement in the battles... Once defeated 4 enemy(milanese) full stacks with a single crusading stack as Denmark. And it wasn't particularly hard....

But I got to admit, the campaign AI in MTW was actually very good. I can still remember playing england and having the almohads kill my navy, then attack EVERY single coastal province I had with 1-2 stacks in the same turn. That's a well planned invasion!

[econ21]

Once defeated 4 enemy(milanese) full stacks with a single crusading stack as Denmark. And it wasn't particularly hard....

4 stacks simultaneously? I could not do that. I'd be struggling to beat 2. I tend to use fairly conventional tactics (line em up, go for good match ups, try to flank). Sheer weight of numbers will overwhelm a conventional battle line.

[gardibolt]

There are those of us who find M/M plenty challenging, so make it too much harder and we opt out of buying the game. Sorry, but it's true.:oops:

[Daykeras]

I think what the OP is asking is that when set to H and VH for campaign difficulty (or maybe tied to battle difficulty) the enemy will make large single stacks for attacking and defending, instead of a million tiny worthless stacks.

[pat the magnificent]

I find the campaign difficulty on VH to be trivial. The AI needs to be much more aggressive when it's at war with you.

[R'as al Ghul]

4 stacks simultaneously? I could not do that. I'd be struggling to beat 2. I tend to use fairly conventional tactics (line em up, go for good match ups, try to flank). Sheer weight of numbers will overwhelm a conventional battle line.

You can do that fairly easily if your general has 8 or more stars and you only face captains. Then you need Heavy Cavalry and Infantry. Surround the individual stacks and they rout easily. When routing you mow them down without taking any losses. You have to play that very offensively.

The campaign AI could do better. It should know exactly what to build and where to build. It should be able to plan further in advance than humans can. It should be better at calculating upkeep costs. It should frustrate the hell out of you in a good way because it's always better off than you and it has the better troops.
The battle AI should know exactly in which microsecond it needs to issue the retreat order for skirmish units. It should be impossible to catch an archer unit before it retreats behind a spear line. It should constantly adapt to your manouvering and outmanouver you. It should be able to read the battlefield better than humans because it knows its height and terrain values in exact numbers that translate in increase of kill rates for archers. It should be able to time the charge of the infantry line with the flanking cavalry exactly down to the millisecond. And so on, you get the picture.

[HoreTore]

4 stacks simultaneously? I could not do that. I'd be struggling to beat 2. I tend to use fairly conventional tactics (line em up, go for good match ups, try to flank). Sheer weight of numbers will overwhelm a conventional battle line.

Nope, you would have won that battle easily. My tactics aren't very special either....

To explain: it was on that hill section to the south of dijon. My army was pretty standard, my usual 5 missile(norse), 10 infrantry(crusaders and dismounted huscarls) and 5 heavy cavalry(crusaders and bodyguard). The milanese were all led by captains, and had their standard pavise/spear mix, with a few other units. The battlefield was with a big central hill. My army started on one slope, the enemy on the other side. I rushed to the top and quickly routed the first stack as they were almost on the top. I then formed a battleline and fired on the other stacks as they climbed to the top. When they had climbed about 3/4 of the hill, I charged everything at them, and they routed almost instantly. Result: about 3500 dead milanese, and about a third of my army dead. Easy as h...

Come to think of it, terrain advantages are extreme in this game....

However, fighting the hard stacks like the mongols are a lot harder... Manage to beat four full stacks of mongols in one battle, and you can call yourself the King of M2TW....


There are those of us who find M/M plenty challenging, so make it too much harder and we opt out of buying the game. Sorry, but it's true.

Well, we're not talking about changing M/M, or any of the existing difficulty levels. What we're talking about, is adding harder difficulties on top of the existing ones(at least I am...). So, you would have the ones today as they are, but add like level "superveryhard" and level "supersuperveryhard" on top of those. It wouldn't affect you the slightest until you start playing "level superveryhard"....

[dismal]

The entire campaign game is easy busy work until the Mongols or Timurids show up with stacks capable of giving a real challenge on the battle map in single player. The AI needs work here more than any where else.

I am waiting for the patch to start any new campaigns but I thought i would ask for a general opinion on the subject.

Should there be an almost unbeatable VVH setting for campaigns?:inquisitive:

I agree. VH/VH is not VH enough. I don't know that they need to make a level that is as insane as Civ4 on its highest levels, but more challenge would be nice.

At a minimum, I'd like to see a harder level where the AI focuses much more on building full stacks of top of the line units before attacking. And gets enough resource bonuses to pull it off.

I'm not sure if the problem can be attributed to the battle or the campaign AI. If the battle AI was better, much of what the campaign AI does might make mnore sense.

I doubt the campaign AI anticipates how badly it will get thumped in battle. It probably calculates it has a superior force when it sends it's stack of milanese militia out to get a 10-to-1 loss. It's impossible for the AI to make good tactical decision when it's off by several orders of magnitude on what's likely to happen when the battle is fought.

[chickenhawk]

Some further thoughts....:idea2:

Maybe they should seed each campaign so one particular AI faction is guaranteed to get big fast and present a real challenge. It could be as simple as giving one AI faction a LOT of starting money. So when the player ran into that faction it would be much stronger and holding a fair piece of the map.

Some basic coding to keep it troops together also seems in order. Rule one minimum garrisons, rule two no second stack until the first one is maxed, rule three keep best general with the best stack, and so on. That way you would at least get one real battle per AI faction.

And yes I know it is easy say and hard to program.:juggle2:

Or they could work on the multiplayer campaign some more...

[Foz]

Make the game too difficult and they risk losing sales. My guess if you want a really tough campaign you'd need to go to the mods for that. They probably can do it better than CA anyway. :yes:

I don't think the OP's point is that the game overall should necessarily be harder... but I think more along the lines that the maximum difficulty of the game should be higher. When I play a game, I expect the top AI level to beat me into the ground numerous times before I can learn enough about the game to contest it and maybe become good at beating it. That last bit isn't even a requirement: it should be very tough, and it's okay if the majority of players then cannot even beat the maximum difficulty setting. This game for most players seems not to deliver on giving a sufficiently difficult challenge at high difficulty settings... and so I think it's a reasonable suggestion to ask for the higher difficulties to be harder than they are, so the game can actually challenge those players who have become good at it. So not so much that everything should be harder, but that the AI should improve more with each successively higher difficulty level. This could take the form of better AI scripts, pumped unit stats, extra added AI cash, better AI cooperation against the player, and various other things we could call AI cheating, as long as it makes the AI sufficiently difficult to take down. I personally would prefer better AI play to make it more challenging, but I acknowledge the limitations of this at the current time, and so other options may be required.


But there's got to be a way to make this game harder. The problem isn't the campaign AI, it's the battle AI which keeps screwing up. Isn't it possible to give the AI a +5 or so bonus to everything combat stat or something? Like in MTW...

Personally, I think the problem is the campaign AI. It just does not have a "killer instinct". Try just pressing end turn in the middle of a war, turn after turn, autoresolving and building nothing. Often not much will happen for ages. I suspect you will get tired before the AI conquers you.

I agree partly with HoreTore and partly with what Econ said in response: I actually think neither facet of the AI is particularly good. What's at fault is just general overall lackluster AI gameplay, in all parts of the game. Regarding campaign, it seems to use strategy pieces decently (agents and such) but it is poor at fielding effective armies (it recruits far too much militia to be effective, and typically a bad balance even of that) and clearly employs what a human would likely call an inferior build order. One big thing I've noticed is that some buildings (especially the port) will not correctly predict their economic impact on the settlement details scroll, presumably because the game doesn't precalculate the new sea trade that will open. The AI building script probably relies on those improvement numbers to figure out which building is better, which may explain its often quizzical build orders. Other campaign considerations would be that real power blocks don't often form b/c of pointless AI bickering, and the generally nonsensical responses it gives in some diplomatic dealings. It apparently cannot fathom that it should beg for mercy when I am slaughtering it and have multiple stacks ready to finish the job.

As for the battle AI, its most common problem looks to be failing to account for a given threat. Even before the lines are engaged, you can most of the time ride some cavalry around behind the main enemy battle line and slaughter any generals/artillery/archers who happen to be there. Likewise you can often get a clean charge into the side of an enemy unit. In fact the AI mostly ignores cavalry attempting flanks, which also results in its units being engaged on multiple fronts very frequently. One big improvement might be if the end enemy units could better guard the end of the line from flanking. It might also be helpful if the AI could keep a few troops back as reinforcements just to plug up holes from fleeing troops and to specifically engage and deter flanking attempts. Honestly the AI's battlefield prowess would probably improve by lightyears if it could deter simple flanking maneuvers that will get it completely crushed if unchecked.

Here's hoping that some of these things are in the soon-to-be-released 1.2 patch.

[chickenhawk]

:balloon2:

I have read something similar to this idea in other games so it is not quite as crazy as it sounds.

step 1, the person playing the campaign checks a box that says they want to play a real opponent in their next major battle.

step 2, the basic odds present in the battle pop up on screen in the multiplayer lobby looking for a volunteer. The volunteer would know at least the general degree of challenge involved, basically get a look at the pre battle odds screen and get a chance to decide to take it or not.

step three, the more battles you take on in other peoples campaigns the higher priority your own get on the list.

step four, Help me out here gentlemen, I do NOT have all the answers.:help:

I do think not knowing if you were going to get a complete noob or a very experienced player in command of the other side in campaign that you were heavily invested in would be very "attention getting". it might make you think twice about four to one odds too.

[todorp]

Vanila MTW2 difficulty levels are aimed at the mass market players and are easy for players who become veterans :duel:

Download and play some of the mods,
http://www.twcenter.net/forums/forumdisplay.php?f=211
which significantly improve the campaign AI and the battle AI and will make your game much tougher. I play "Lands to Conquer" and "Darth mod", both are excellent and take the game to a higher level.

[stealingjoy]

What the game really needs is difficulty sliders for tons of aspects of gamplay.

Like, suppose I want really hard battles, really tight finances but somewhat easy diplomacy and assassination? Or medium battles, loose finances but difficult agent training? Stuff like that would really increase my enjoyment of the game a lot. Maybe even go so far as sliders for types of units (pikemen, cavalry, etc).