Archers need a nerf.

Arches OP?

  • Yes

    Votes: 82 27.9%
  • No

    Votes: 102 34.7%
  • Buff Armor instead

    Votes: 139 47.3%

  • Total voters
    294

Users who are viewing this thread

Proportionally raising damage resistance of armor will disproportionately favor player armies, because they have disproportionately larger number of units with high armor ratings. The effect it will have and you ignore is that it will drastically decrease damage that AI armies can inflict on player units, reducing their casualties. But will have little effect on damage player armies do to AI units with next o no armor anyway. More alive/un-wounded player units will just slaughter AI units faster. Battles will be even more one sided then they already are.
The solution to that is to make AI armies level better so that they aren't bringing stacks of low tier troops to battle (totally doable through tweaks to the AI experience formula). A high tier army SHOULD wipe the floor with a low tier army. Combat should not be balanced around the need for a low tier army to compete with a high tier one in order to create artificial challenge for the player.

I dont think it is a great or reasonable assumption that a game in Early Access/testing/beta phase is not going to change. I think at this stage we should ask for mechanics to change or features to be introduced for better not just quick fix to fit our desired outcome (I think)
I have already explained that saying "well the AI should be better" is a cop-out that makes the discussion moot because such a thing is totally out of our hands. We are discussing concrete ways in order to fix the imbalances in the game the way it works currently, not in a theoretical future iteration of the game. These fixes will involve much less work then rewriting the combat AI. If you have anything constructive to add, please do. If you do not, please stop pushing a solution that will possibly never happen, and renders the discussion of combat balance pointless until it does.
 
Last edited:
Well with all due respect i do not agree with the idea of AI cheating by not having recruits after you maul them.

That would destroy the whole balance of the game because if my foot general loses one battle, i have to start all over again with recruits so I dont see why the AI should not especially since i play on max difficulty and they can recruit tons of troops from villages, many of which are not recruits whereas the player is locked to basic recruits for the most part unless you do villages missions for ages ( aside from prisoners/tavern mercs).

Also, if you do not execute lords in max difficulty, for a good part of the campaign ( before you are very strong ) the main way to weaken a faction is to prevent it to have high tier troops by destroying their individual smaller armies one by one and hoping as a small vassal that your faction will capitalize on it and manage to survive when they siege enemy cities from factions that outnumber them.

If the AI would come back all the time after a few days full of mid tier, it would almost break the game because it would make the stronger more numerous faction almost impossible to defeat for the smaller AI faction ( they already get chain war declarations on them when they re weak currently ) or when the player wants to invert the "power bar" by siding with the smaller faction against the much stronger one.
 
Last edited:
Well with all due respect i do not agree with the idea of AI cheating by not having recruits after you maul them.

I agree and disagree at the same time, having them spawn instantly with mid tier troops is not a good idea and agree it would very tricky to balance, however having them running around with too many lower tier troops is also bad and makes fighting them rather boring.

Personally I'm in favour of having them spawn with mid tier troops but with a 30 day delay while they are off training those troops, refills from garrisons is also fine but there should be a limit to that so weak lords don't totally destroy empire defenses.
 
The solution to that is to make AI armies level better so that they aren't bringing stacks of low tier troops to battle (totally doable through tweaks to the AI experience formula). A high tier army SHOULD wipe the floor with a low tier army. Combat should not be balanced around the need for a low tier army to compete with a high tier one in order to create artificial challenge for the player.


I have already explained that saying "well the AI should be better" is a cop-out that makes the discussion moot because such a thing is totally out of our hands. We are discussing concrete ways in order to fix the imbalances in the game the way it works currently, not in a theoretical future iteration of the game. These fixes will involve much less work then rewriting the combat AI. If you have anything constructive to add, please do. If you do not, please stop pushing a solution that will possibly never happen, and renders the discussion of combat balance pointless until it does.

It is not constructive because you are not hearing what you want to hear.

AI usage of formations is in the devs priority list, so I think it is reasonable and constructive to ask to fix a problem properly instead of using an artificial boost or nerf.
 
Well with all due respect i do not agree with the idea of AI cheating by not having recruits after you maul them.

That would destroy the whole balance of the game because if my foot general loses one battle, i have to start all over again with recruits so I dont see why the AI should not especially since i play on max difficulty and they can recruit tons of troops from villages, many of which are not recruits whereas the player is locked to basic recruits for the most part unless you do villages missions for ages ( aside from prisoners/tavern mercs).

You remember we had that on release, right? And how nobles after losing a major battle (once two or three in-game years had passed and bandit parties had spawned) would repeatedly get destroyed by bandits or looters? And how that meant if a faction lost two major battles, it would be crippled afterwards?

Also, if you do not execute lords in max difficulty, for a good part of the campaign ( before you are very strong ) the main way to weaken a faction is to prevent it to have high tier troops by destroying their individual smaller armies one by one and hoping as a small vassal that your faction will capitalize on it and manage to survive when they siege enemy cities from factions that outnumber them.

The better way to weaken a hostile faction is to form a clan-only army that you use to follow around other armies in your kingdom, then jump into their fights so they always win. The AI isn't smart enough to factor in the player's party/army being nearby and jumping into the fight mid-way.
 
The better way to weaken a hostile faction is to form a clan-only army that you use to follow around other armies in your kingdom, then jump into their fights so they always win. The AI isn't smart enough to factor in the player's party/army being nearby and jumping into the fight mid-way.

Of course you must do that, it is part of the general strategy of depleting high tier units from the enemy so that when he inevitably comes back a few days later, a faction with 4-5k men can stand a chace against one that has 15-18k troops.

Once your smaller faction sieges a city, sometimes it has to fend off 2 defending armies coming at us before it can actually assault (if enough are alive to assault). That is unless it is the AI commanding and assaulting with just a battering ram.

You can defend easy once you have more than 500 men because you can retreat and beat much larger armies but before you reach that point, if the AI comes back a few days later with more men you are toast.

There has to be an actual point of winning field battles. If everytime you destroy an army , the AI comes back with lots of mid tier units a few days later, the game is broken for me, might as well not fight any battle and just wait behind your walls :smile:

My point is whatever they come up with with patches, decisive battles should still carry serious consequences for the player and AI.

A smaller faction with a more elite army who just beat their enemy main army should retain this advantage of having clearly better troops and even increase it.

Because this is the result of previous battles that allowed them to slowly build up this life saving advantage for them. It shouldnt be ruined by artificially allowing defeated factions to invent veterans everytime their army gets wiped out IMO.

I agree it can be boring if you purge them too hard because their armies would have lots of recruits but that is the way it should be when a faction has lost too many battles and is losing the war.... however in my opinion what could be done which would not break immersion ( maybe it is already the case i dunno ) is to allow the AI to resupply from their castle and city garrisons if they have better units there.
 
My point is whatever they come up with with patches, decisive battles should still carry serious consequences for the player and AI.

A smaller faction with a more elite army who just beat their enemy main army should retain this advantage of having clearly better troops and even increase it.

You do realize they are getting most of those mid-tier troops out of their garrisons, right? They only spawn with either 10% or 25% of their total party size and about a third of those are low-tier.

I agree it can be boring if you purge them too hard because their armies would have lots of recruits but that is the way it should be when a faction has lost too many battles and is losing the war.... however in my opinion what could be done which would not break immersion ( maybe it is already the case i dunno ) is to allow the AI to resupply from their castle and city garrisons if they have better units there.

They already do that. That's where they get most of their non-recruits on respawning. It takes them a few days of running around, of course, and usually players aren't being hit by the same armies, but by the kingdom's second or third or fourth army.
 
It is not constructive because you are not hearing what you want to hear.
I'm hearing "The solution to archers overperforming is the AI being better in general, and being able to use shieldwall specifically." That statement isn't constructive to this discussion because it renders this entire discussion moot. It is pointless to talk about combat balance at this juncture if such a major change will be implemented, and it is not clear if it ever will be.

An analogy: If a man asks how to become wealthy, "Win the lottery" is not a constructive response...

Is it possible that the AI might be improved? Sure, but certainly the opposite may be true and the AI might never be improved. Either way it isn't germane to the discussion about DEFINITIVELY ACHIEVABLE ways to improve the combat balance. Once again, we are discussing obviously achievable ways to balance the game we have RIGHT NOW, not the game we want to have, or MIGHT have at some point.
 
Last edited:
I'm hearing "The solution to archers overperforming is the AI being better in general, and being able to use shieldwall specifically." That statement isn't constructive to this discussion because it renders this entire discussion moot. It is pointless to talk about combat balance at this juncture if such a major change will be implemented, and it is not clear if it ever will be.

An analogy: If a man asks how to become wealthy, "Win the lottery" is not a constructive response...

Is it possible that the AI might be improved? Sure, but certainly the opposite may be true and the AI might never be improved. Either way it isn't germane to the discussion about DEFINITIVELY ACHIEVABLE ways to improve the combat balance. Once again, we are discussing obviously achievable ways to balance the game we have RIGHT NOW, not the game we want to have, or MIGHT have at some point.

It is as possible to improve AI as it is to degrade archers or upgrade armor. Besides, I have a feeling that devs didn't made AI use shield wall -or otherwise hold their shield up like they did in Warband to make player archers viable to begin with. It's painfully obvious how harmless archers are against the player. Degrading them further would only make things worst.
 
Last edited:
It is as possible to improve AI as it is to degrade archers or upgrade armor.
No it isn't. Changing damage resistance granted by armor is as easy as changing the formula within the code. Changing archer shoot speed may perhaps be a little more complicated depending on how the animation system works, but is likewise relatively straightforward. Both changes may not be "one line code changes" but they are certainly less work than rewriting a portion of the combat AI.

If I know anything about software development (and I assume I do because I went to university for it) I know that as a general rule developing AI systems is the most time consuming and complicated part of ANY project, sometimes even by ORDERS OF MAGNITUDE. It is not as simple as changing a 0 to a 1 within the code, and all of a sudden the AI use shield wall intelligently.

Besides, I have a feeling that devs didn't made AI use shield wall -or otherwise hold their shield up like they did in Warband to make player archers viable to begin with.
You may very well be correct, in which case "just let the AI use shieldwall" is not the solution to the balance issue because it is intended NOT to be that way by the devs.

It's painfully obvious how harmless archers are against the player. Degrading them further would only make things worst.
You are wrong. Archers (especially archers that can hold their own in melee) are the most dangerous units against the player outside of getting swarmed by infantry in melee. Furthermore, we aren't talking about how well archers do against the player. We are talking about how well archers perform as a part of player army composition against the AI armies, where they outperform every other type of army composition regardless of situation, opponent, or sometimes even tactics.
 
No it isn't. Changing damage resistance granted by armor is as easy as changing the formula within the code. Changing archer shoot speed may perhaps be a little more complicated depending on how the animation system works, but is likewise relatively straightforward. Both changes may not be "one line code changes" but they are certainly less work than rewriting a portion of the combat AI.

Do you realize that all code is just a set of formulas and that shieldwall AI is already codded in the game? You don't even need to change it.

If I know anything about software development (and I assume I do because I went to university for it) I know that as a general rule developing AI systems is the most time consuming and complicated part of ANY project, sometimes even by ORDERS OF MAGNITUDE. It is not as simple as changing a 0 to a 1 within the code, and all of a sudden the AI use shield wall intelligently.

Actually yes, it may as well be as simple as that or close enough, because AI already knows to shieldwall, therefore all that is needed is to trigger it. Which may as well be as simple as changing 0 to a 1. Or couple of them.

You may very well be correct, in which case "just let the AI use shieldwall" is not the solution to the balance issue because it is intended NOT to be that way by the devs.

Well, assuming you would manage to convince devs that there is a balance issue to begin with and archers are overpowered, then it would be solution. And much better then reducing damage or raising armor.

You are wrong. Archers (especially archers that can hold their own in melee) are the most dangerous units against the player outside of getting swarmed by infantry in melee. Furthermore, we aren't talking about how well archers do against the player. We are talking about how well archers perform as a part of player army composition against the AI armies, where they outperform every other type of army composition regardless of situation, opponent, or sometimes even tactics.

When I was talking about "player" I was talking about both, player and his armies. As he obviously commands them. And as I said, archers are almost harmless against player -and his armies.
 
If I know anything about software development (and I assume I do because I went to university for it) I know that as a general rule developing AI systems is the most time consuming and complicated part of ANY project, sometimes even by ORDERS OF MAGNITUDE. It is not as simple as changing a 0 to a 1 within the code, and all of a sudden the AI use shield wall intelligently.
Actually yes, it may as well be as simple as that or close enough, because AI already knows to shieldwall, therefore all that is needed is to trigger it. Which may as well be as simple as changing 0 to a 1. Or couple of them.
The AI does not know how to shieldwall. The only time a shieldwall is ever used by the AI in the game is behind the gate in the siege, and that is only because the troops start there when the map is generated and never move. The AI may not even have a system in place to recognize when a good time to use shieldwall would be, or how to react to archers on one side and flanking troops on another, or any other myriad of scenarios. All of that would have to be coded from scratch if they do not have the necessary systems in place. Frankly, your assumption that the developers could just change one line of code and the AI will suddenly start using shield wall intelligently in combat shows how little you know about software development and coding in general.

Well, assuming you would manage to convince devs that there is a balance issue to begin with and archers are overpowered, then it would be solution. And much better then reducing damage or raising armor.
Which would be ignoring why the devs took it out in the first place, if they indeed did. Overperforming archers is better than useless archers.

When I was talking about "player" I was talking about both, player and his armies. As he obviously commands them. And as I said, archers are almost harmless against player -and his armies.
Archers are most certainly NOT "harmless" against player armies. And... at the risk of sounding like a damn parrot: THIS DISCUSSION ISN'T ABOUT HOW WELL ARCHERS PERFORM AGAINST THE PLAYER, IT IS ABOUT HOW WELL ARCHERS PERFORM AGAINST THE AI, EVEN IN SCENARIOS WHERE THEY SHOULDN'T BE PERFORMING WELL.

I put the last part in bold because you seem to keep ignoring it.

Do you realize that all code is just a set of formulas ...
lol
 
Which would be ignoring why the devs took it out in the first place, if they indeed did. Overperforming archers is better than useless archers.

Eh... as a general game design rule, it is better to have an underperforming element than an overperforming one. A single underperforming (or nonperforming) element means every other element remains viable and suitable as a point of balance otherwise. But if you have one overperforming element, then the others are non-viable because your balance considerations have to account for the overperforming element first and foremost -- or risk even more imbalances in other places.

That's my issue here with archery: the developers absolutely watch YouTube videos and read comments (there are reddit memes in the game files, for proof of this) made about MnB, so they know just how far a player can push things with -- for example -- noting but Fian Champions. "Well, OK, if they can do that much, then I'll adjust the economy to match!" But what about players with actual balanced party compositions? What about the poor suckers who read into the lore and decide their Vlandians should have lots of melee cav and sergeants and, oh, just a few crossbowmen? For them, they spill gallons of blood and get relative scraps of influence and loot for the trouble.
 
The AI does not know how to shieldwall.

F3-F2 ...vuala, shieldwall. Or do you think that AI on your soldiers is different then that of the enemies?

The only time a shieldwall is ever used by the AI in the game is behind the gate in the siege, and that is only because the troops start there when the map is generated and never move.

So AI does know to shieldwall after all.

The AI may not even have a system in place to recognize when a good time to use shieldwall would be, or how to react to archers on one side and flanking troops on another, or any other myriad of scenarios.

If getPlayerArmy()->haveRanged() == true then
self->getInfantry()->setShieldwall() = true
else
self->getInfantry()->setShieldwall() = false
endif;

Which would be ignoring why the devs took it out in the first place, if they indeed did. Overperforming archers is better than useless archers.

If my assumption is true, then devs turned shieldwall off for AI because they thought archers are undepowered. Which is exact opposite of what you think.

Archers are most certainly NOT "harmless" against player armies.

I said almost harmless.

And... at the risk of sounding like a damn parrot: THIS DISCUSSION ISN'T ABOUT HOW WELL ARCHERS PERFORM AGAINST THE PLAYER, IT IS ABOUT HOW WELL ARCHERS PERFORM AGAINST THE AI, EVEN IN SCENARIOS WHERE THEY SHOULDN'T BE PERFORMING WELL.

I put the last part in bold because you seem to keep ignoring it.

Which ...at the risk of sounding like a damn parrot: SHOWS THAT PROBLEM YOU THINK YOU HAVE IS IN AI AND NOT THE ARCHERS.

lol
 
I get its blunt but how does STONE being thrown by the worst unit is able to deal a good chunk of damage against armored units let alone heavily armoured ones? Its silly.

As for me charging straight at them, am not they just turn and get really lucky shots in.In the Original warband they would deal little damage or no damage (in regards to peasents throwing rocks) which is more then fine they shouldt be able to. Lowly troops being able to deal good damage against top tier troops is absurd.

A large rock hitting your head with force and speed while you ride at them is very dangerous.

I didnt see looters kill elite units with rocks to be honest, just rarely they can down a light cavalryman running at them or an archer if multiple looters throw many rocks at his face.

What i saw is they can sometimes kill a higher tier unit wih their melee weapon. But sometimes i just stand in front of 10 looters throwing rocks at me and it does not do much damage ( mid/late game with good gear) , id say i could have more damage in real life perhaps.

I respect your opinion but i think tier 1 units like looters are very weak as it is, normally some of them should have slingers and throw knives/cheap pointy things which would make them far more of a threat ( and some of them should wear better things than garments) because as they are now, they are more like starving cavemen :smile:
 
So AI does know to shieldwall after all.
No, the game knows how to deploy the unit in shield wall at the start of the battle and leave them there... or are you too stupid to see the difference?

F3-F2 ...vuala, shieldwall. Or do you think that AI on your soldiers is different then that of the enemies?

If getPlayerArmy()->haveRanged() == true then
self->getInfantry()->setShieldwall() = true
else
self->getInfantry()->setShieldwall() = false
endif;
Apart from your mangling of anything remotely resembling proper syntax... (lol using assignment operator for boolean operation, no semicolons at the end of statements, no scope control, and calling an object like a function and then using a pointer access operator).

Simply turning on shield wall if archers are present in the enemy army is not "using shield wall properly" it's mindless, and allows players to potentially exploit the mechanic... It also reduces the effectiveness of infantry in melee, and their speed. There are going to be times when the AI will need to not be in shield wall even if the enemy force has archers. A player is capable of toggling when needed, the AI can't.

If my assumption is true, then devs turned shieldwall off for AI because they thought archers are undepowered. Which is exact opposite of what you think.
And now they are overpowered.

Archers are most certainly NOT "harmless" against player armies.
I said almost harmless.
Pointless distinction. You are using semantics and strawmanning to troll, I don't know why you haven't gotten a ban yet.

Which ...at the risk of sounding like a damn parrot: SHOWS THAT PROBLEM YOU THINK YOU HAVE IS IN AI AND NOT THE ARCHERS.
lol
The AI will likely never be able to use shield wall up to the standards of a human. More then that, it may even be an intentional design choice that the AI CAN'T. To balance archers around assuming it will happen at some point is idiotic.
 
Last edited:
Simply turning on shield wall if archers are present in the enemy army is not "using shield wall properly" it's mindless, and allows players to potentially exploit the mechanic... It also reduces the effectiveness of infantry in melee, and their speed. There are going to be times when the AI will need to not be in shield wall even if the enemy force has archers. A player is capable of toggling when needed, the AI can't.

Pretty much sounds like you simply want an arrow-immune infantry that doesn't suffer any negative impact from the use of formation.

You simply want the infantry to flick shield up every time an arrow comes in, and then set it down when there are now arrows in the air, and thus, suffer no penalty from a defensive formation, move at regular speeds, fight at regular capacity, and yet still be impervious to arrow fire.

And don't you dare say this is strawmanning, because from what ridiculous mess you've been arguing on and on so far, there's no other implication to be had.


You simply want the archers to be useless, down to a non-factor. For some inexplicable reason.
 
I just fought a battle against Vlandia with my foot general no cavalry campaign. Vlandia had 40 crossbowmen against my 35 archers left at the end.

Once we beat their infantry , i ordered my 20+ guys left standing to advance at the enemy with shield wall while my archers were sent to flank them from the right.

Problem is they totally murdered my sturgian vets in shield wall and my archers barely won it in the end... the battle ended with like less than 10 guys left standing but it helped me realize that if something needs a nerf, it is the ai crossbowmen rate of fire... insane rate of fire, my sturgian archers could barely compete and shield wall didnt save my infantry :smile:
 
Pretty much sounds like you simply want an arrow-immune infantry that doesn't suffer any negative impact from the use of formation.

You simply want the infantry to flick shield up every time an arrow comes in, and then set it down when there are now arrows in the air, and thus, suffer no penalty from a defensive formation, move at regular speeds, fight at regular capacity, and yet still be impervious to arrow fire.
No, and that isn't what I'm advocating. I am just saying we shouldn't balance archers around the AI using shieldwall because it can't and probably never will, and if you have read my other posts in this thread you can see I have been disagreeing with people who think low tier archers should be useless against heavy infantry.

Archers are overperforming right now, but I'm not saying they should be nerfed into uselessness.

Way to jump into this thread and totally misrepresent the argument that I am trying to make...

Which is: That archers are so dominant right now because cavalry (the counter to archers) sucks at the moment (at least in relation to what they should be), and armor across the board doesn't do enough to mitigate damage (including in melee).
 
Last edited:
I would not say cavalry is useless to be honest, the high tier one is pretty dangerous, it is just that cavalry charge effect on morale and the damage of horse collision with humans is quite underpowered. If they get bogged down in melee it is normal that they lose quickly, it is more of an AI issue than cavalry IMO.

What is true though is that spear thrusts damage is underpowered so it affects cavalry somewhat, although high tier cav can still one shot when hitting at full charge.

If you play a campaign on foot without cavalry in max difficulty you will see what i mean, you will notice that good cavalry ingame is actually quite powerful because they can run down your archers unless they charge head on against superior numbers. It is the lower tier cav that is quite weak.
 
Back
Top Bottom