[Balance Suggestion] Prevent Steamroller in Mid Game

An easy overhaul of combat to make it more realistic?

Having now played quite a few games, and a few with the released version, I am still seeing game results that are a bit "uninteresting" in the mid-game and slog to the end.

The start can be tough. Sometimes I don't make it. But if I can get my 3-5 cities established and get to say, Armor, then I can build some quite good units. At this point, the game is effectively over, as the AI does not seem to be able to build a good stack, and I can defeat every stack and city with my best stack, without losses. The stack just gets better over time.

Obviously we can ask that the AI is better at its job, but that is hard work. What can be done with simple tweaks of numbers that might provide a better balance?

A better balance to me would be a situation where 'elite' troops were only just a little better than normal troops, defensive militia maintained a role throughout the game, and attacking somebody was likely to cause some losses. Some losses is important because it means you are getting weaker as you work through their defences (different cities), and might have to pause / stop for a while. 

Would it therefore be possible to change the balance for the better by reducing the 'power curve' of weapons, armor and veterancy, making upgrades slightly less good than they are now? Of course this still leaves the power curve for magic unchanged, but that can be tweaked IF we can get the mundane troop balance correct.

A change (down) in weapon and armor values is likely to make existing monsters quite tough. Such is life. 

The other change I propose is to reduce the effect of the various 'Barracks' type buildings so that at no stage are earlier troops worse than later troops. It is an annoying feature of many games that experienced veteran troops value is degraded by the impact of new 'training centres' which turn out better troops as a matter of course. Instead, veterans should be able to upgrade their armor and weapons to be the equal of any new troop, plus their bonuses for experience. The training buildings will need to provide other bonuses, such as reduced costs, quicker training, etc. 

Finally, it would be good to be able to include different types of weapons with different effects, and not simply each being a better version of the previous weapon. I have given the example before of short and long bows. Short bows shoot quicker, but have less penetration. Long bows shoot slower, but have more penetration. Xbows shoot slower still, but have excellent penetration and have a use against highly armored opponents. This is WAY more interesting than each bow just being better than the one before it.

So that sounds like a mod I could have a go at. Any thoughts?

 

22,646 views 15 replies
Reply #1 Top

You could try turning up the difficulty level a notch.

Reply #2 Top

LH has the ability to upgrade armor, weapons and the number of individual units per squad.  There are also some troops I have seen the AI make that scared the hell out of me as they eat through my well upgraded defenders if I let them get a couple attacks on them.

Reply #3 Top

weapon damage was already tuned several times in the beta. i think the balance is not too bad right now. you can't really stop steamrolling in a 4X game anyway. once you start to snowball, you'll wipe out the other factions - doesn't really matter if boar spears are 9 attack or 7 - if you get them and the enemy factions don't (or fail at using their stuff properly), you win. 

playing higher difficulties helps :) if you are too good at the game and run away in the mid game at your current difficulty, try the next higher level, until you find a setting where the AI has enough bonuses to present a challenge past the mid game. or try dense monsters and/or high world difficulty to keep the fights vs. the world more interesting for longer. or play epic research pace to keep the AI on even terms for longer.

Reply #4 Top

The other change I propose is to reduce the effect of the various 'Barracks' type buildings so that at no stage are earlier troops worse than later troops. It is an annoying feature of many games that experienced veteran troops value is degraded by the impact of new 'training centres' which turn out better troops as a matter of course. Instead, veterans should be able to upgrade their armor and weapons to be the equal of any new troop, plus their bonuses for experience. The training buildings will need to provide other bonuses, such as reduced costs, quicker training, etc.

Finally, it would be good to be able to include different types of weapons with different effects, and not simply each being a better version of the previous weapon. I have given the example before of short and long bows. Short bows shoot quicker, but have less penetration. Long bows shoot slower, but have more penetration. Xbows shoot slower still, but have excellent penetration and have a use against highly armored opponents. This is WAY more interesting than each bow just being better than the one before it.
End of quote

 

Good ideas :yes:

and definitely doable via mod.

Reply #5 Top

I do not think weapons and armor at higher tech levels should be made weaker. I think the balance in that regard is really well done. But I do agree that barrack, forge, and other unit enhancing buildings should effect older units. Would it be possible to have fortress-specific buildings upgrade a units equipment by adding a subtype trait to the unit? For example:

  • Sturdy Armor (+2 def) to Masterwork Armor (+4 def) to Legendary Armor (+6 def). Weaponsmith could add Masterwork Weapon (+2 attack).
  • For the accuracy bonus from the Training Yard line, make it Practiced (+10 accuracy) to Disciplined (+15 accuracy) to Well Trained (+20 accuracy) to Elite (+25 accuracy).
  • The "+25% attack against the opposite race" already has a trait. 

So if a unit is created in the fortress with this building, it would be applied automatically on creation. If a unit does not have the trait, it would be given.

Of course, the big problem with this is that that means that any unit created in another city could reap the benefits of the fortress. X| Perhaps the bonuses to unit creation, levels given, and cost inherent in the fortress would discourage this. Or a trait could applied to units created in a fortress (any fortress) that would only let their equipment be upgradeable, to stop conclaves or villages from spewing out troops.

Reply #6 Top


It should be way harder to attack and approach cities- they do this well in civ 5 giving cities a ranged attack each turn before the siege. During the "battle on the streets" even people throwing chairs / rocks / tables from buildings at invaders would help (the city attacks once per round). then each level higher the city is it should get an upgraded attack. different types of cities also should have differing attacks, the fortress being the best or hardest to conquer. bonuses to these attacks could be enhanced by what buildings you have created as well, like an alchemist adding +1 acid burn for 20 turns on successful attack of enemy unit.

Reply #7 Top

there are already several strategic spells that stop, slow down, divide, or damage enemy stacks that are in your territory heading to your cities.

Reply #8 Top

Higher difficulties don't fix the problem, it just makes it take longer before you get to the point where you are more powerful than every opponent put together, and makes it so that all opponents hate you because they think you are weak even though you are like 100 times stronger than them.

Most of the options here are great, but they will simply give the player more options which will have the opposite of the intended effect. 

The following things really need to happen imo before the AI even has a chance of providing a fair fight, and only one of them is moddable:

  1. Every pre-designed unit needs to be ripped out of the XML with great fury. Replace it with stuff that isn't optimally bad.
  2. The evaluation function that is used to determine the strength of an individual unit needs to be redone entirely. The game thought my Level 46 lightbringer champ with scythe of the void was only "strong."
  3. The evaluation function that is used to determine the strength of a faction needs to be redone entirely. I suggest this be done by analyzing   the most powerful hypothetical groups that a faction can create.
  4. There needs to be some sort of a "underestimation coefficient" that raises every time the AI loses a battle against a player that they expected to win. This coefficient is applied to their evaluation of a player's stack.
Reply #9 Top

...there are already several strategic spells that stop, slow down, divide, or damage enemy stacks that are in your territory heading to your cities.

 

yes there are plenty of magic options but you cant tell me that cities of 100/1000/10000/100000 populations cant assist in turning away an invasion or a random monster that's walking around my territory, this should be accounted for in city/territory defense. also if i have specialized buildings in my city i have access to better defensive options depending on what is present.

just think of the greatest fortresses in history, they are very hard to take over during a siege.

Reply #10 Top


I propose a fix to snowballing and boring end game:

If Player is above AI is score, and game has past X amount of turns, and Player is in a war that the player is winning, the following must happen:

[Note the following should be placed as a toggled settings in match setup]

1. All non-friendly/close/allied AI must unite together against the player, and all close/friendly AI are given a choice to join the player, or if the player rejects, become hostile.

2. If by turn 50, AI is still losing hard, A Event must trigger that converts ALL Enemy AI units into Demons(reskinned normal units) [stats = X% greater], say its a last ditch ritual by the enemy sovereign, lore wise.

3. A Turn Counter must initiate for say 100 turns, by which point, if the AI has lost more territory than gained, all enemy AI must immediately forfeit, thus granting the player immediate victory (to save the boring grind that is the cleanup of the map of enemies)

Think of this as a Final show down stage. A lot about the events can be made to have a different variety, but point remains though, the cleanup phase should be given an option to be avoided

Reply #11 Top

AI will never be as good at protecting its top units, or creating a doom stack.  Its advantage, at higher difficulties comes simply from out producing the player.  

Some ideas:

- Make it easier for AI to wear down an elite stack with its waves of junk armies....

- Limit healing per season (This is what often lets a player survive limitless enemy waves)

- AI needs to be better at using spells.  Never had "kill," mana burn, or strategic spells like tremor cast at me.   (Granted I play mages mostly, and blindly counterspell).

Use of tremor/freeze alone would make it much harder for the player to rely on just one elite army.

 

Reply #12 Top

Quoting sxyz123, reply 11
AI will never be as good at protecting its top units, or creating a doom stack.  Its advantage, at higher difficulties comes simply from out producing the player.  

Some ideas:

- Make it easier for AI to wear down an elite stack with its waves of junk armies....

- Limit healing per season (This is what often lets a player survive limitless enemy waves)

- AI needs to be better at using spells.  Never had "kill," mana burn, or strategic spells like tremor cast at me.   (Granted I play mages mostly, and blindly counterspell).

Use of tremor/freeze alone would make it much harder for the player to rely on just one elite army.

 
End of sxyz123's quote

I agree with all 3 (In my opinion, units shouldn't regenerate unless in town). As a side note, I've noticed Ai of custom factions cast pillar of flame and freeze on me.  A lot of times, it freezes my army next to something nasty, like a dragon.

Reply #13 Top

they actually started using strategic spells around patch .90

i don't think the perceived problem is a real problem. i mean, it's quite obvious that you are supposed to win the game before you reach the end of the tech tree - otherwise the tech tree would be longer. what some people consider "mid game" is actually the phase where the game is already won. whether you play it out or not is up to the player, but once the game is won, there's no real point throwing in artificial roadblocks to slow the player down. i tend to just stop playing when i have clearly won. doesn't matter at all if i'm at the end of the tech tree or somewhere in the middle.

winning a 4X game is about eXpand/eXplore/eXploit/eXterminate -those 4X can be summarized as one "X" eXponential growth. once your snowball outpaces the others, it's effectively over.

play higher difficulty if you get to that point to soon/too easily.

Reply #14 Top

Azunai_, in my opinion the problem is the game is won before 1/3 of the turns are over, and when every AI has a significantly higher "score" than me, due to the reasons listed above.

Reply #15 Top

Quoting Azunai_, reply 13
they actually started using strategic spells around patch .90

i don't think the perceived problem is a real problem. i mean, it's quite obvious that you are supposed to win the game before you reach the end of the tech tree - otherwise the tech tree would be longer. what some people consider "mid game" is actually the phase where the game is already won. whether you play it out or not is up to the player, but once the game is won, there's no real point throwing in artificial roadblocks to slow the player down. i tend to just stop playing when i have clearly won. doesn't matter at all if i'm at the end of the tech tree or somewhere in the middle.

winning a 4X game is about eXpand/eXplore/eXploit/eXterminate -those 4X can be summarized as one "X" eXponential growth. once your snowball outpaces the others, it's effectively over.

play higher difficulty if you get to that point to soon/too easily.
End of Azunai_'s quote

 

That is one way to put it I suppose, but sometimes, I just want my 4+ hours of playing to worth something more than just "Oh, I am now unstoppable, so ... new game?"

Sometimes I want a new challenge to arise so my elite units that I worked so hard to get, can be put to good use, more than just a slow game of wack the mole.