Stacks of Doom are great!

He said that he was yielding his first city, meaning the enemy will not be in a city with 100% defensive bonus. Second even a city such as you describe can be taken using the Verdun tactic, catapult down to 0% defensive bonus, move to the correct side of the river and bring plenty of troops particularly cheap anti-archer units.
HEY! Verdun's MY reference joke! :)
 
Which in my opinion is just the right amount of battle tactics to have in a strategy game like Civ. They tried to make battle tactics more important in Civ 5, but since the game is so heavily unbalanced (horsemen before patch, importance of iron after patch and super dumb AI), you hardly need to use any tactics, except when you lure the AI into a choke point, which in my opinion is more an exploit.

Well personally I'm a tactics guy so I would prefer some more tactics, but I don't want CIV to get as tactical as say Total War, if I want that level of tactics I play well... Total War.
 
You can't separate out the design from the performance of the AI. They chose a model, with whatever theoretical virtues, that couldn't produce a competent computer opponent. And even in this approach, too complex for the computer to do well at, the tactics are extremely artificial and limited because of the lack of basic ingredients from real tactical games (opportunity fire, combined arms on defense, etc.) It's design 101: a clever concept is only a good idea in a single player game if the computer can use it well.
 
Which in my opinion is just the right amount of battle tactics to have in a strategy game like Civ. They tried to make battle tactics more important in Civ 5, but since the game is so heavily unbalanced (horsemen before patch, importance of iron after patch and super dumb AI), you hardly need to use any tactics, except when you lure the AI into a choke point, which in my opinion is more an exploit.
You sound like you're switching gears from "Civ4 required good, thoughtful tactics on higher difficulties" to "Those dead easy tactics is all that should be required in a strategy game".
 
You sound like you're switching gears from "Civ4 required good, thoughtful tactics on higher difficulties" to "Those dead easy tactics is all that should be required in a strategy game".

Dead easy compared to what? The point of reference is the new version - you know, the one where the AI can't use navies and performs consistently worse at any given difficulty level than it did in earlier Civs. I'm having a hard time reconciling the overblown dismissals of the Civ 4 combat model with the wholly embarrassing replacement that we got in Civ 5; are you really saying that Civ 5 is, in actual practice, better? I just want to make sure that I understand your point.
 
Nono. I'm not talking about Civ5 at all, instead Bad Brett's view of Civ4 tactics. He said that that level of tactical thought is good for a Civ game. Sorry if I confused you.
 
They chose a model, with whatever theoretical virtues, that couldn't produce a competent computer opponent.

That is just an opinion not a fact.

It is my opinion that the 1upt model does not mean that we can't get a more competent AI opponent. I think Civ 5 not having a more competent AI opponent is more related to the release being rushed. Will the AI ever be competent enough so that without other bonuses it can beat the diety players? Obviously not but it is my opinion they can make it considerably better than it is now.

And even in this approach, too complex for the computer to do well at, the tactics are extremely artificial and limited because of the lack of basic ingredients from real tactical games (opportunity fire, combined arms on defense, etc.) It's design 101: a clever concept is only a good idea in a single player game if the computer can use it well.

I haven't figured out for myself if I think that bringing in more tactics as you mentioned would be a good thing or not... Though I agree with the fact that the whole concept is not very worthwhile if the developers don't make the AI be able to effectively use the tactics.
 
They chose a model, with whatever theoretical virtues, that couldn't produce a competent computer opponent.

That is just an opinion not a fact.

Hmm, it may not be a fact as concrete as, say, "water is wet" but it certainly rises above the level of mere "opinion". I would call it more like a "general consensus".

It is my opinion that the 1upt model does not mean that we can't get a more competent AI opponent

Right, but we haven't. The only hard fact about this is that the AI is currently pretty bad. There is as yet no concrete evidence to suggest it will be improved.
 
HEY! Verdun's MY reference joke! :)

Yes, but it was a really good one.:)

Really what we need is some concentration on AI. No combat model in the world (Other than a chess/checkers/risk model is going to help if the AI is a blithering idiot. We have already seen that the AI requires a numerical advantage to win a stack war, and a numerical and TECH advantage to win a 1upt war against a human opponent, other games posses similar errors on the AI side the TW series being an example; how do you kill 12,000 Mongols with 600 englishmen? Apparently by sitting in Acre and letting them come. It's ridiculous, the last time I lost an even fight with an AI in a strategy game was when I was like 8 or 9 (Warning: definition of even fight may very) A truly even fight between a human and an AI would be a slaughter if a human player were put in control of the AI forces. Few games posses AI that I truly feel is up to an fight against a human without advantages.

When the AI becomes capable of giving us a fair fight on noble then will the combat model become a serious issue.
 
Actually I am looking for something in between SOD and 1UPT, The two opposite approaches are both dull. My preference is small sized stacks to add tactical depth, for example a stack limit of 3 units per tile where you can make combos like:
  • 1 Spear, 1 Horse , 1 Archer
  • 1 Spear , 1 Sword , 1 Horse
  • 1 Infantry, 1 Artillery , 1 anti-tank

you see in these small sized stacks units support and complement each other, which is superior to 1UPT and more tactical than SOD.
 
Is tactics that much more complex in Civ5?
I like that someone asked that. I know it's *more* complex than Civ4, but I'm now questioning the degree.

Even though there's more factors than just this, generally the more complex the game the worse the AI is at it. I think that since the AI appears to be worse in Civ5 than Civ4 (at land at least), the tactics are more complex. It's also just the general amount of thought I need to put into a turn. In Civ4 I didn't need to put much thought in, and in Civ5 the longer I think the better war I wage.

I'm repeating myself. I think Civ5 is more complex, thus the AI can't cope as well. But it has more room for the AI to grow, so it's probably the better system in the long run as the designers improve it.

Me said:
A truly even fight between a human and an AI would be a slaughter if a human player were put in control of the AI forces. Few games posses AI that I truly feel is up to an fight against a human without advantages.
This is a good point, and the more I think about it, the less I actually care if the AI could ever keep up in a one on one fight against a human. I'd be happy if the AI could use every individual part (like naval, bombarding, sieges, fast and slow units) without many stupid mistakes. As long as it's halfway competent then we can always even things out by giving it more units (hopefully without carpeting the map). We'd have a really fun game then.

Ayt said:
Since the combat AI is so awful I'd say no. I think it is even easier than Civ 4.
I think bad AI is a result of a complex tactic system. The tactics system is a lot of visualization, which AIs suck at. Consider the game GO for example. It's a really visual game, and the best AI isn't better than an average human player.
 
The thing you have to realize about stacks in civ 4 is that they're NOT countered by collateral damage. Maybe that was the intention, but the way it's implemented, splitting up a stack won't reduce your collateral damage at all. Well, not unless you split it into less than 5 units, which is too small of an army to do anything.

The REAL counter to stacks in civ 4 is the fork- the ability to threaten multiple cities at the same time. If you've got an invasion force that can potentially hit 3 different cities or resource tiles, and they can't counterattack in time, they'll have no choice but to split up their army into three pieces. The more places you threaten, the more they'll have to split up their army. The "stack of death" where you put all your army in one tile just doesn't work.

Unfortunately, the AI in civ 4 still leaves a lot to be desired. It doesn't try to threaten more than 1 city at a time, and it always leaves behind a garrison in every single city. So, when fighting the AI, you don't need to worry about being forked, and there's little point in sneaking in to take out vulnerable cities. The giant stack of death works pretty well against the AI- but only because the AI is so stupid and has such monstrous production advantages.
 
He said that he was yielding his first city, meaning the enemy will not be in a city with 100% defensive bonus. Second even a city such as you describe can be taken using the Verdun tactic, catapult down to 0% defensive bonus, move to the correct side of the river and bring plenty of troops particularly cheap anti-archer units.

This isn't a limitation of SoD. This applies for every strategy game out there. Throw enough units at something and it will be defeated. Notice how vague the strategy is:

how to defeat a Civ4 ai stack of doom (any size, any units):
1. Get 6-10 catapults (with city attack)
2. Get a lot of swordsmen (with city attack)

So what exactly is "a lot"? Is 8 a lot? Is 20 a lot? And how would this strategy differ if used in Civ 5? If I build 30 Swordsmen in Civ 5, I, too, can defeat any AI army of the same era.

Sure, unit spam is a strategy, but it is a really inefficient one.
 
This isn't a limitation of SoD. This applies for every strategy game out there. Throw enough units at something and it will be defeated. Notice how vague the strategy is:



So what exactly is "a lot"? Is 8 a lot? Is 20 a lot? And how would this strategy differ if used in Civ 5? If I build 30 Swordsmen in Civ 5, I, too, can defeat any AI army of the same era.

Sure, unit spam is a strategy, but it is a really inefficient one.

But if a city is really valuable you can choose to do so. But the combat model should prevent such things from happening. It is really strange to loose 90% of your army but still atttacking at full strenght.

For example the new combat model should use a morale modifier. If you lose 10 units in your stack for 1 in an enemy stack the organized attack modifier will drop (or what ever the other combat model parameters are). You have to abort your attack and use more ranged firepower. (And hope in the mean time the enemy doesn't bombard your units strength and morale below freezing point).

A good combat model contains more that 4 lines (is complicated) and is well tested. And I don't see that happening. Especially in combination with an good AI.
 
Top Bottom