Defending on clear tiles is a DEATH sentence

[Luthervian]

No he hasn't sufficiently covered the cons.

What he is saying that there is a grand strategy of unit placement in the current system.

This hex you live and that hex you die does not make for a grand strategy game.


Or am I missing something here?

I'm lazy so I'm not going to scroll back to his first post (before everyone's tone started going really downhill) so I'm going to paraphrase, and likely paraphrase wrongly.

I believe some of the cons (which means downside to your plan - not "your plan is wrong" -- every plan has its pros and cons, its how you weigh decisions in life) he may have observed are (in my words):

1) Removing the open terrain penalties could run the risk of homoginizating the different tiles and remove percieved benefits of highly defensible terrains. I.E., is it worth looking for hills and forests if the penalties for fighting in clear terrain aren't sufficiently bad enough to go look for them?

2) Removing the open terrain penalties runs the risk of unbalancing other parts of the game in some manner of another (again, I don't know or care, what the exact words are).

3) etc. I am sure there are more, and I am sure Ahriman will return and point them out and correct my own wording. It's fine.

Please note that I pointed out that he observed the "cons". That does not translate into meaning that I agree (or disagree) with him. I percieve you as now being very defensive of your views and also borderline hostile, and in jeopardy of misinterpreting anyone's criticism as an attack. This of course, is understandable, given the tone of some of the criticisms to your ideas. They could have been phrased more constructively - doing so certainly would make the topic feel less hostile than it needs be.

Thanks to Krikkitone, I've gotten some useful ideas from this topic, and will toy with the 0% open terrain defense modifiers (I'm intrigued enough to try it), but I'll withdraw from posting further here -- there's too much bitterness and hostility around these halls nowadays for this lurker.

Carry on.

 

[Mongolia Jones]

1) Removing the open terrain penalties could run the risk of homoginizating the different tiles and remove percieved benefits of highly defensible terrains. I.E., is it worth looking for hills and forests if the penalties for fighting in clear terrain aren't sufficiently bad enough to go look for them?

Well right now the difference is between rough and clear terrain is 87.5%. Thats more than 3x promotions. Is that acceptable to you?

How much do you think the difference should be?

2) Removing the open terrain penalties runs the risk of unbalancing other parts of the game in some manner of another (again, I don't know or care, what the exact words are).

Right now the game is soo imbalanced that we can runaway AI steamrolling continents. I'm trying to bring some balance back to the game.

3) etc. I am sure there are more, and I am sure Ahriman will return and point them out and correct my own wording. It's fine.

I'm sure he will too.

Thanks to Krikkitone, I've gotten some useful ideas from this topic, and will toy with the 0% open terrain defense modifiers (I'm intrigued enough to try it), but I'll withdraw from posting further here -- there's too much bitterness and hostility around these halls nowadays for this lurker.

Carry on.

That's all I want. For people to try it without the penalty

BTW, I have tried it with the 0% and the game works pretty well. There is still some steamrolling, albeit at a reduced rate.

I now want to go one step further and give all defenders at least +20% bonus. And that's where the firestorm started.

It never hurts to try.

 

[Sonereal]

Fixing AI problems by changing defensive modifiers=/=a sound idea.

The defender has the advantage in this game most times anyway. The defender can pull back and then counterattack the invaders on open terrain. The defender can set up in rough terrain and force the attacker to enter open terrain and get slaughtered by ranged attacks. The defender has use of its road network. The defender has city bombard. The defender can get one promo that negates the -33% penalty right off the bat.

By the way, the EUIII's attrition bonus for defenders made perfect sense in the context of the game. It doesn't here since the fights take place at a tactical, not strategic level. You can wage a war of attrition pretty easily in Civ5.

 

[jwallstone]

To add my two cents, copied from my mod where I tweaked the value to -15%:
http://forums.civfanatics.com/showthread.php?t=389887

The high disadvantage for defenders on open terrain is too strong, making ranged sniping too easy and is also possibly one of the major reasons for the high power of mounted units, since on open terrain they could hit and run heavily penalized defenders with impunity. The base penalty also overshadowed fortification, the Shock promotion, and even attacking across rivers. However, removing the penalty is possibly too big of a change. Furthermore, defenders can often Fortify on open terrain immediately upon moving there, something that can't be done on rough terrain. I believe this provides a careful medium, where fast advance on open terrain is penalized, but a slow advance with fortification keeps just a slight advantage on defense.

 

[Sonereal]

To add my two cents, copied from my mod where I tweaked the value to -15%:
http://forums.civfanatics.com/showthread.php?t=389887

The high disadvantage for defenders on open terrain is too strong, making ranged sniping too easy and is also possibly one of the major reasons for the high power of mounted units, since on open terrain they could hit and run heavily penalized defenders with impunity. The base penalty also overshadowed fortification, the Shock promotion, and even attacking across rivers. However, removing the penalty is possibly too big of a change. Furthermore, defenders can often Fortify on open terrain immediately upon moving there, something that can't be done on rough terrain. I believe this provides a careful medium, where fast advance on open terrain is penalized, but a slow advance with fortification keeps just a slight advantage on defense.

Mounted units are supposed to be powerful on open terrain. A swordsman unit would get slaughtered in open terrain. The real issue is that horsemen have 12 strength and can take down spearmen or at least stalemate them on even terms.

 

[player1 fanatic]

To add my two cents, copied from my mod where I tweaked the value to -15%:
http://forums.civfanatics.com/showthread.php?t=389887

The high disadvantage for defenders on open terrain is too strong, making ranged sniping too easy and is also possibly one of the major reasons for the high power of mounted units, since on open terrain they could hit and run heavily penalized defenders with impunity. The base penalty also overshadowed fortification, the Shock promotion, and even attacking across rivers. However, removing the penalty is possibly too big of a change. Furthermore, defenders can often Fortify on open terrain immediately upon moving there, something that can't be done on rough terrain. I believe this provides a careful medium, where fast advance on open terrain is penalized, but a slow advance with fortification keeps just a slight advantage on defense.

I agree on this. Swordsmen that moves a bit and fortifies should have slight edge against attacking swordsmen. As long as that edge is smaller then hill/forest bonus.

-15% does that nicely (25-15=10%).

 

[Folket]

Mongolia Jones, I agree with your changes. Where could I get hold of your mod?

 

[grzybek]

In my opinion we discuss 2 matters here, and I'm going to address both

1. Whether the current system is correctly balanced or not.
I think it's more like a "personal point of view" issue and I think everyone should accept that there are people who think differently.
The only conclusion, here, is that people that don't want their battle system to be changed should not take part in this discussion and not stop people with different mind to reach their happiness Someone doesn't like the idea - don't download the mod and your good

2. Problem with deciding game balance.
I agree that the game is imbalanced currently and it needs to be fixed.

While reading through all comments in that thread I have identified following parameters that have influence on the battle/war results, which were mentioned in the discussion, and these are:

- Attacker offensive bonuses
- Defender defensive bonuses
- City strength
- Terrain bonuses

People mostly addressed only one of those parameters while I think that we have to think about this problem globally, because if we focus on a single thing we will just make the game imbalanced, only in a different way.

My proposal is:
- remove defensive penalty for units in clear terrain (or even make it a bonus like +20)
+ it will make some unit promotions unusable (so what to do with them)
+ makeing +20 bouns in clear will also cause us to make the rough terrain bonus bigger (otherwise there will be no difference between terrain types)
- give an automatic promotion for Mounted and Archery units vs. flat terrain (or address the issue of Mounted/Archery units advantage in some other way)
- increase city base strength together with the power of city defensive building.
My units taking -1 demage done by the city is just a laugh. The cities are completely defensless, only taking the city takes 4-5 turns instead of 1, but there is no question whether will I take it or not

 

[dmieluk]

Wow, what a mess of a thread... And, here is the problem right here.

Why would you want to favor defense like this?

To stop 1cpc (1 ai CIV per CONTINENT) spam.

I agree that 1cpc (or the steamroll) is too prevalent. IMO, it is silly to try to address it by flattening the tactical terrain. This game has suffered enough from flattening, please no more.

There appears to have been a mis-diagnosis as to the cause of the steamroll. It is caused by bad AI. I would argue it is not the combat AI (as bad as it is) which is the problem. It is the strategic/diplomatic AI which is completely broken, and causes frequent steamrolling.

When the AI is at war, it knows only to attack, and that means that it should always send the troops it has towards the enemy. It does not keep a reserve.

The AI will rarely consider peace if it is winning a war.

Therefore, if two AI players are at war, it is likely that they are sending all their troops at the other. At the end of the first conflagration, if either AI has has decent amount of troops remaining, they will neither accept peace nor be stoppable by force. It will be a steamroll.

But, what should stop this steamroll? Shouldn't the dominant military player take all those cities? This is where the unhappiness effects of taking cities should come in. For the human player, there is a cost to continually acquiring conquered cities (albeit too small). For the AI, they have so much surplus happiness that taking cities would have no penalty for them. Therefore, the negative feedback mechanics designed to prevent a human player from stomping the board (and don't even do a good job at this) have absolutely no negative feedback effect on AI steamroll.

Why is this? Well, the AI is so unable to compete with the human player, that they had to give it massive arbitrary bonuses. But these very bonuses are the mechanism for negative feedback. By giving the AI these particular bonuses, the AI has become immune to the built in negative feedback. It is terrible terrible design. But at least this makes it clear why 1cpc happens so often.

How should it be addressed?

- the best way would be to improve the overall AI, so that it doesn't need such enormous bonuses to compete with human players.

- failing that, improved strategic AI would let the AI realize when it was losing a war, and have it turtle (with reserve troops) while suing for peace. The attacking AI should be able to recoginze turtling AI, and decide whether to try to go for an overrun, or agree to peace with benefits.

- lastly, if the AI will stay bad and you must give it such huge benefits to compete with the human, then the penalties which relate to negative feedback should be increased proportionally. In the very least, this will preserve the negative feedback for the AI, which in theory at least, should provide the internal logic for a peace settlement before 1cpc occurs.

If you doubt what I have said, think of it this way.

1cpc is a consequence of the failure of negative feedback. That is a truism.

What is the cause of the failure of negative feedback? Do you really think it is caused by flatland penalty? Or is there a better hypothesis?

 

[eric_]

Right now it's checkers. It is always to my benefit to attack rather than reserve my attack because i will always cause MORE DAMAGE.

Which then leaves you stuck on that tile awaiting your own certain death (assuming the AI even has any units to back up its front line, which it never does, which is the real problem, imo). That is the tactical (not strategic...unit placement = tactics, not strategy) trade off that is lost if you remove the -33% defense modifier.

 

[eric_]

It does not keep a reserve.
...
Therefore, if two AI players are at war, it is likely that they are sending all their troops at the other.

This is even true when the AI attacks a single City State. Monty DoW'd a militaristic CS in my current game and, judging by the "state" of his military (he had none) when I DoW'd on him a few turns later, he threw everything he had at them and no one came home. (Well, there was one lousy pikeman left, who just flitted about the fringes of the battles, passing up two opportunities to attack my exposed horses before I realized he was there). After conquering him, I approached the militaristic CS, as well as a nearby maritime CS, and both had shockingly large armies, far larger than I've seen fielded by any rival civ in any game so far.

It definitely seems broken when it'd be more challenging to defeat either of two single-city CSs than it was to conquer all three 3-city civs on my contitent on King.

 

[grzybek]

Response to Dmieluk

To be honest - I think you are partially right in here I would even say that you are more right than wrong but still not completely right.

It is also a problem that:
- attacker is soooooo much better then defender on the flatland. As explained - it is true for archery and mounted units, but not for other. And this problem also needs to be addressed
- city defenses are not really a defenses but more like a mosquito

In the end - it would be great if we were able to fix the AI so at least on problem (1cpc) would be gone, but we can't. This is why we discuss "possible" solutions to the problem

So the only place were you are wrong is - "We cannot fix AI"

 

[Dizzy75]

Which then leaves you stuck on that tile awaiting your own certain death (assuming the AI even has any units to back up its front line, which it never does, which is the real problem, imo). That is the tactical (not strategic...unit placement = tactics, not strategy) trade off that is lost if you remove the -33% defense modifier.

Exactly! As I found out (painfully) in my last game, when Hiawatha outnumbered me 4-to-1 in units.

It's definitely NOT always to your advantage to attack, except with ranged. You can slaughter an enemy on open terrain, but then your (damaged) unit is open to being slaughtered. Do you have another unit nearby that can kill the unit that counterattacks you? It reminds me much more of chess than any previous version of Civ did.

Horses have a decent advantage because of this, but that's a different story, because they're a bit overpowered in general.

Maybe the difference is TOO extreme, but I wouldn't want it reduced by too much.

 

[dmieluk]

In the end - it would be great if we were able to fix the AI so at least on problem (1cpc) would be gone, but we can't. This is why we discuss "possible" solutions to the problem

Possible solution which we can probably do...

Increase the cost of unhappiness for AI players for number of cities (eg. by +4 per city)
Make colosseums +8 instead of +4 for AI players

Obviously you'd have to tinker with it a bit to find the amount which would reestablish the negative feedback loop. But the point is, there is no net change here to AI happiness, but more cost when a city is taken, and it would take a while for the AI to overcome that extra by building the extra collosseum...

The point is to try to solve it by reestablishing the negative feedback loop which the developers clearly intended, but failed to provide, rather than by simply dipping the game into a barrel of tar (make defense much much stronger, so everyone is kind of stuck).

 

[Sonereal]

A good solution would be to change how luxury resources work so some of that give a lot of happiness but little gold while others give a lot of gold but little happiness. Then make those effects stack so that the player with gold (low happiness, high gold) has a reason to trade for wine (high happiness, low gold).

 

[Abegweit]

@Krikkitone

In defense of Abegweit where the defender in Civ4 had a bigger advantage.

1) Vastly superior movement within your own borders
Because of road/rail spam, infinite rail move, and a default unit move of 1 (tops 2) square(s), the defender had a huge movement advantage.


2) Cultural protection for cities
It is true that cities have an inherent protection now, but in Civ4 you could put huge stacks in cities where they became an absolute pain in the ass to conquer. Not so in Civ5. Yea you get a bonus, but I haven't had any problem taking any city up to now. Not like in Civ4 where sometimes I had to back off completely until i amass enough forces to take it.


3) War weariness
In Civ5 there are ways of getting around unhappiness. You can puppet or you can raze. In civ4 you don't even have to conquer any cities and you still get hit with unhappiness. Either end the war or trash your gold/science to keep your pop happy.


4) The ability to make the first strike, if chosen
In Civ4, if I am not mistaken the Defender ALWAYS gets first strike as a default. The First Strike was a unit special and part of the Drill IV advance (I believe) to counter the inherent defender bonus.


5) The ability to choose the battle location
I guess in connection to point 1. With road/infinite move rail spam the attacker can choose the location of the battle much more effectively.


6) Far stronger city defence (SoDs)
In civ5 cities surrounded by clear tiles are sitting ducks. There is no effective defense of a city in such a case. The only defense is a good offense (i.e. quickly wipe them out before they wipe you out.) Again, cities in Civ5 aren't really that much of an obstacle, hence the mod to make cities in Civ5 stronger. From what I can tell it is quite popular a mod.

Nice post. I have little to add to this but I do want to make a clarification. When I said that the defence has the ability to strike first if it choose, I meant exactly that; the defence gets the choose to strike out at the enemy if and when it wants. The possibility of confusion with the Civ4 concept called "first strike" never even occurred to me.

In Civ4, the AI trundles through your lands one square at time. Forget the two-movers; the AI doesn't take advantage of them except to pillage the tiles in its path. When you have maneuvered it into a place of your choice, you can strike it. First! Move up the artillery and deal collateral damage. Follow through with melee and other units to destroy it.

Alternately, if that seems right, you can hide behind your walls and watch the AI impale itself on them.

The defence always has the initiative in Civ4. The opposite is true in Civ5. The initiative lies with the attacker.


@Krikkitone

The city defence in Civ5 is laughable compared to that of Civ4. Cultural defences plus garrisons were far more powerful than the pathetic auto-defence and bombard capabilities in Civ5.

WRT to roads, I'm not exactly sure what your point was. If you were saying that the attacker can't use roads in the defender's culture, well I didn't know that. If true, it's completely irrelevant because roads don't exist anymore. Instead we have these funny trading routes that appear on the map. Units fall over each other if they try to follow them. In practice, wars take place on open fields and that's true both for the attacker and the defender.

 

[Sonereal]

1) Vastly superior movement within your own borders
Because of road/rail spam, infinite rail move, and a default unit move of 1 (tops 2) square(s), the defender had a huge movement advantage.

There was no infinite rail move in Civ4. As a matter of fact, it made defending almost too easy since you can move an entire stack anywhere in a normal-sized empire.

In Civ5. No one's forcing you not to build roads to the border.

2) Cultural protection for cities
It is true that cities have an inherent protection now, but in Civ4 you could put huge stacks in cities where they became an absolute pain in the ass to conquer. Not so in Civ5. Yea you get a bonus, but I haven't had any problem taking any city up to now. Not like in Civ4 where sometimes I had to back off completely until i amass enough forces to take it.[/QUOTE]

That's mostly an AI problem, not game problem.

3) War weariness
In Civ5 there are ways of getting around unhappiness. You can puppet or you can raze. In civ4 you don't even have to conquer any cities and you still get hit with unhappiness. Either end the war or trash your gold/science to keep your pop happy.

War weariness was flawed. Did you really have problems with it? I didn't. I ignored its existence in most of my games and I was fine having wars that last thousands of years.

4) The ability to make the first strike, if chosen
In Civ4, if I am not mistaken the Defender ALWAYS gets first strike as a default. The First Strike was a unit special and part of the Drill IV advance (I believe) to counter the inherent defender bonus.

No. The Defender didn't always get first strike. Units with the Drill promotion or were special units got first strike right away.

5) The ability to choose the battle location
I guess in connection to point 1. With road/infinite move rail spam the attacker can choose the location of the battle much more effectively.

What are you talking about? Attackers can't use your roads or your railways. You get to choose where the battles are fought and you can strike when you want. With ZoC, this makes perfect sense. It was horrible in Civ4. You only got to choose where to attack, not defend but there was no ZoC and a 50-unit stack could actually ignore another stack and pass right by it.

6) Far stronger city defence (SoDs)
In civ5 cities surrounded by clear tiles are sitting ducks. There is no effective defense of a city in such a case. The only defense is a good offense (i.e. quickly wipe them out before they wipe you out.) Again, cities in Civ5 aren't really that much of an obstacle, hence the mod to make cities in Civ5 stronger. From what I can tell it is quite popular a mod.

Cities sitting out in the open are pretty much sitting ducks in the real world as well. Which is why you're supposed to make sure the enemy doesn't get that far. Again, its the AIs fault for not relying hard enough on ranged units in cities and units in rough terrain that can strike at the right time.

By the modern age, taking cities really isn't that hard. Cities more often than not have little defense against rocket artillery spam these days. Which is why having units is important.

 

[Mongolia Jones]

There appears to have been a mis-diagnosis as to the cause of the steamroll. It is caused by bad AI.

Well I don't think I have made a mis-diagnosis, but who knows really.

Somewhere in this thread (page 1 or 2 maybe) I detailed my observances which I will repeat in a nutshell:
1) Continent has a primarily flat region on one side and a primarily rough region on the other.
2) The flat region quickly gets steamrolled by a single victorious AI.
3) Eventually the civs in the rough region get conquered (albeit more slowly) due to the fact that the now big superpower civ (which has swallowed up most of the landmass) has 4x the economy, more units, better tech.


In conclusion: Clear terrain get rolled. Rough does not. Hence my focus on clear terrain.

 

[Sonereal]

Well I don't think I have made a mis-diagnosis, but who knows really.

Somewhere in this thread (page 1 or 2 maybe) I detailed my observances which I will repeat in a nutshell:
1) Continent has a primarily flat region on one side and a primarily rough region on the other.
2) The flat region quickly gets steamrolled by a single victorious AI.
3) Eventually the civs in the rough region get conquered (albeit more slowly) due to the fact that the now big superpower civ (which has swallowed up most of the landmass) has 4x the economy, more units, better tech.


In conclusion: Clear terrain get rolled. Rough does not. Hence my focus on clear terrain.

That's a broken chain of logic. Its an AI problem. Not terrain problem.

 

[Keitan]

I also dont like the -33% on open Terrain. I know, i dont have to
defend there, but the Ai doesnt know that, and thats one problem why
the Ai is so weak.

So thats why i made a Mod for me last weak with following values.

v.2
Open Terrain -10% defense from -33%
Marsh -15% defense from -33%
Fallout -33% defense
Ranged Units like Archer Siege Air Units receive Promotion: +20% Ranged vs open terrain, or -15% Ranged vs rough terrain
Horseunits receive Promotion: -25% City and -25% rough terrain
Ironclad/Panzerschiff RangedCombat 18->20, Cost 220->200, Moves 4->5
Tanks/Armor receive Promotion: +30% Defense vs Ranged
Great General Moves 2->3

Download by Modbrowser->Gamerules or here