Early Armorer's Guild

[Seravy]

We already know this is a problem but here is another demonstration.

We've seen it with War Trolls so far but now Doom Drakes :
https://www.youtube.com/watch?v=8tLdmPJjfF0

It seems you don't even need to be able to buff them, just having Healing alone is enough. So the problem is bigger than I thought, it's not just something limited to life megabuffing rush.
As usual the gold is from draconian bowmen but that isn't really the problem, you can get early gold in a lot of different ways using lots of different treasure strategies, or even play Dwarves on rich and get lucky with your gold ores.
The problem is clearly the ability to outright get to maximal tier in no turns for a mere few thousand gold.
To be honest I think if the AI was intentinonally doing this, buying up to Armorer's Guild immediately, even low difficulty levels would be near unwinnable - it's so unbalanced. And unlike the human, the AI can simply gather the gold by not wasting it on other stuff, they have higher income.

Previous related threads : 
http://www.realmsbeyond.net/forums/showt...p?tid=8983
http://www.realmsbeyond.net/forums/showt...p?tid=8991

The original game contained this problem somewhat by having the gold cost depend on production already done, so going from 0 to 100% completion costed a lot more...but that system was flawed and easy to trick, as well as unfun - by the time those turns pass, you might forget you wanted to buy the building. Also buying the building is not a game balance issue for 99% of the buildings - getting any nonmilitary building 10-20 turns earlier "only" means that many more turns of resource production, and the cost of buying will always be way higher than the benefit, as well as gold being a more scarce and in demand resource than production. So I still think that was a bad system and removing it was the right thing - but we definitely need something else to fix it.

So, looking back on the old threads, we decided against :
Requiring fame for the buildings (very artificial, hard to explain?)
Raise production cost of the buildings (would be bad if you want to produce in multiple cities, bad for AI)
Make "Buy cost" higher for military buildings (change production trick works)

Then ignored the problem as we had no good ideas at that time (and still don't, unfortunately)

(also, unrelated but sad fact - raising the range penalties only really achieved Chaneller to be retort good for rush.)

Note : Obviously adamantium matters, but casting Holy Armor or Endurance can have the same effect or more so it isn't the actual problem. The problem is having a high tier unit that has good hp, defense and often additional abilities that help its survival (Regeneration and Fear we've seen so far, for Nightmares it would be flight and ranged, Nightblades are invisible, etc) way earlier than it's actually possible to kill one supported by 2-4 uses of healing, even for a full 9 stack of low tier units.

[zitro1987]

We could tie buildings to population size, might result in less overwhelming array of choices and also solve the mass-buying amplifying towers. As most things, into 3 or 4 tiers.

Tier 1 (pop 1) - Stables, fighter's guild, ship-yard, barracks, alchemist, economy buildings around the 250 and lower cost range, including builder hall ones. Buying costs 3X of prod
Tier 2 (pop 6ish) - War College, High-tier economy buildings (Cathedral, Sage's Guild, Merchant's Guild). Buying costs 2.5X of prod
Tier 3 (pop 12ish) - Wizard's Guild, Armorer's Guild, Maritime Guild, Fantastic Stables, Amplifying Tower. Buying costs 2X of prod

Rules and intuitiveness:
*Low to mid town size images to match the tiers
*Messages when you reach the next tier, buildings will be available, etc.
*Once a town reaches a tier, it cannot revert back due to pop losses, it gets locked in (invisible town enchantment might do this?)
*Buying formula tied to population may help balance - as a city grows, the city production increases, thus buying is a bit less important.

[Seravy]

Unfortunately, both this and fame would interfere with the queue system so we can't really do any such hard limits anymore.

[Nelphine]

What about tying the cost of buying production to population, based on building tier?

Tier 1 building, always costs x2
Tier 2 building, needs population 6 to be x2; lower pop it's x3
Tier 3 building, needs population 10-12 to be x2; pop 6+ for x3; lower than that is x4

Except you could buy the wrong thing, as you've explained ..

So, instead of changing cost of buying production, just add production required based on population.

So tier 1 always 100%.
Tier 2, pop under 6, 150%
Tier 3, pop under 6, 200%; if between 6 and 150%

[Seravy]

So, instead of changing cost of buying production, just add production required based on population.

So tier 1 always 100%.
Tier 2, pop under 6, 150%
Tier 3, pop under 6, 200%; if between 6 and 150%

Production required is...the amount missing until completion so I don't understand, can I have some examples?

Also note it probably takes like 8-10x the current cost at least to solve the problem this way. We need to delay it happening by about 50 turns, assuming 100 gold from taxes a turn and 2000 from treasure, that's 7k extra on top of what we already have.

My preferred solution would be "military buildings cannot be bought" or something like that, but that also isn't compatible with changing production and also quite artificial.

[Nelphine]

Right, this would just mean, a starting capital (pop 5) would require 825 production to build a fighters guild. As soon as it reached pop 6, that would drop to 550 (quite possibly meaning you suddenly have a ton of production leftover for whatever you do next - so it might be better to scale it 10% for each population you're missing).

If the capital managed to finish the fighters guild before pop 5, the armorers guild would cost.. uhh.. I have no idea what they cost these days. Let's pretend they're 600 base. So that pop 5 capital would need 1200 production to build it. When the city grew to pop 6, production required for armorers guild would drop to 900, and when it reached pop 12, it would drop again to 600.

The idea here is to 'change' purchasing power, but in a way that avoids people buying something else and then converting to the thing that should be more expensive. Since you're actually changing production required, then for the fighters guild at less than pop 6, someone could buy a farmer's market first (spending 360 gold), but they'd still need to spend 1290 more to buy the fighters guild because the actual production required is higher, not the cost of buying that production.

However, this would mean for your capital, even assuming you did it so fast that you were still pop 5, would 'only' require an extra 1750 gold (550 for the increased fighters guild and 1200 for the increased armorers guild) on top of the current gold requirement of 2300 for those two buildings. More realistically, it would only cost an extra 600, as most likely you wouldn't build either until pop 6.

If you're looking for thousands of gold more, you could specify military buildings get an arbitrary turn limit where the production required changes. But that would probably be very hard to make transparent. But something like a fighters guild costs an extra 25 production for each turn it is before turn 10, and an armorers guild costs an extra 50 production for each turn before turn 35.

(I personally think turn 50 is excessive - that's way after war declaration, and if you can't even build it before war declaration you can't get the unit out there fast enough to do a lot with it.)

[Seravy]

No, you're misunderstanding. I want an EXTRA 50 turns over the turn 20-30ish possible now.
The AI needs to have at least fighter's guild tier units in their more important cities and some uncommon spells to not get steamrolled (but still lose to the AG units, at a reasonable speed). I associate that with turn 80ish.
(and even that is likely far too early for Advanced and below difficulty but that's fine. However Expert and higher shouldn't be able to conquer all 4 wizards, or, any more than 2, using those units and common spells only. Wizard 3 should require at least uncommon, wizard 4 rare spells - whether to support the AG units or as a different tactic you upgrade into is irrelevant.)

Considering our research schedule was turn 72 for all commons and 1-2 uncommons done, turn 80 is realistic for armorer's guild units to still be powerful. It's also at least this much time the AI needs to actually be a threat through their unit production.

On turn 35 an AG unit is two tiers above what the AI uses, and is too durable for the AI to kill even if they used up all their casting skill on trying. Which would be ok if this was the last few such turns, but that stops being the case a lot later, around 80.

Or you can look at it differently : On turn 35 the AI's entire military and summoning production is like, 3 halberdiers and 2 hell hounds worth. You can buy one AG unit a turn and 1 AG unit is stronger by itself than those and beats them easily. 
Another viewpoint is that this early the unit effectively functions as a "semi-zero cost" battle, you need lose the unit as a single one already has more hit points than what the enemy can deal as damage. So your only cost is casting the healing spell(s). That would be relevant later, but this early, it isn't - the AI doesn't yet have all that many stacks. (you can also pick chaneller and laugh at this cost, as we've seen - this doesn't require anything except 2 Life for healing and something to get you treasure. So you have like 6-8 picks left to spend on whatever you like.)

The cost lowering linearly isn't a bad idea. We could define a "reasonable turn" for each building, and each turn before that increases the price in hammers by X%. It's a bit hard to explain but maybe we can get away with something like more advanced building and education technologies taking time to actually develop. Basically it's a hidden "research" kind of thing. All other 4x games you need to research to unlock more advanced buildings and unit types, here you don't that's the root of our problem.
We could in theory even use something like "total hammers ever produced (not bought) in the empire" as the base instead of turns, but I rather use the turns because the production capacity is fairly even between the races so using it pretty much just means whoever expands most gets the best discount which is the opposite of we want, and the same we already have (more expansion leading to more gold = faster buildings).

We'd need to teach the AI to not build them before it's time for them though, otherwise it's just wasting their hammers. It's pretty unlikely to happen anyway but let's not leave it to random chance.

...although I don't think any building except Armorer's Guild, Fantastic Stables and maybe Maritime Guild needs this feature in fact most buildings definitely don't.

[Nelphine]

Ok on lunatic, AI regularly have so many halberdiers floating around by war declaration turn that they can make stacks of 5+ for offense, in addition to the full stacks of Naga or water walking war bears they have.

On defense, having, say, a dragon turtle is a totally reasonable thing by the time war declaration rolls around.


Note: what about amp towers being bought in pop 1 hamlets en masse late game? Zitro mentioned it, and I know I actively play that, which this could stop.

[Seravy]

Why would we want to stop that?
I mean, we assumed the map holds as many of these buildings as it currently does, if we (almost) disable them for the low max pop cities, the whole math falls apart.
Also, "late game" so the turn modifier wouldn't be relevant for that anyway. Population would but I don't want population - It makes the gap between having adamantium early or later even more relevant, the opposite of what we want. Why should the capital get the AG dozens of turns earlier than any other city? Why should that pop 9 neutral get the AG when my capital still can't? Why should Klackons get the AG but Dwarves not?

For the turn count, Sapher's game. He bought the AG but haven't yet bought the first Doom Drake in 1402 November. That's turn 34. After that he steamrolled the entire plane easily and if part 3 was up we'd see he does the same to the other plane and probably wins the whole map before turn 80. (which is 1406 July btw, still not even midgame.)

The AI having 1 dragon turtle (from mercenaries usually) or making 1-2 stacks of 5 halberdiers is not enough. The AI needs to have filled at least their first 4-5 cities with halberdier tier units and not even Lunatic AI can get the settler production, movement, growing to hamlet, and fighter's guild, plus the units done in 35 turns.  That's just the bare minimum to not outright lose in 5 turns, but they'd need several dozens of more halberdiers if they want to actually do a relevant counterattack or keep enough doom drakes busy intercepting them to matter.

[Nelphine]

I'm fine with not worrying about amp towers, you just hadn't commented on it, and it seemed a valid point.

So doom Drakes or war trolls are powerful. If we go back to the original discussion, dwarven golems, we ended up deciding that swordsmen could do the same thing, which led to increased poison relevance.

I'd also say, particularly at expert, bezerkers do the same. So, assuming you do anything like a turn limit, what's going to stop sapher from using fighters guild units like troll halberdiers, bezerkers, golems, or galleys? 

Note, I'm not saying don't do something about armorers guilds, but the biggest take away from he last discussion was that it wouldn't stop anyone who really wants to play that way, but it would be a gameplay fun issue for anyone who generally only plays up to expert difficulty anyway. I never build armorers guilds, so it won't have any impact on my gameplay, but that implies that the core problem is something else.

For instance, could he do the same thing with minotaurs or stag beetles? If not doesn't that imply that it's actually certain units having too high of certain stats? I'm personally obsessed with endurance - could that imply that overland speed is too powerful?

I just feel we've had this discussion before, and while this is a YouTube video of it (and I'm a little surprised at doom drakes) I'm not sure what's new enough that you can actually say it's definitely the armorers guild, especially and then claim turn 80. That's just begging to simply use other units, like summons.