Fewer wizards on Myrror = easier game?

[Seravy]

First test game, Lunatic difficulty. Myrran AIs in 1507 have roughly the same number of cities as Arcanus wizards and are overall weaker on the charts.

I'm restarting and watch turn-by-turn what is happening for the one that had the fewest cities.
Race is dwarves which should in theory, build fastest...

1501 March.
Initial 2 cities settled, one more settler produced and stating to move. University built in fortress. City produces 33 hammers, settler took 6 turns to build (cost 200).
Second settler is being produced, first is enchanted by Wraith Form. Two initial cities are building a ship yard/trireme. 8 Ghouls are summoned.

So far everything looks ok.

1501 Sept.
Third settler builds a city. Fourth settler on the way. Fifth is being produced.
Took out an easy lair, raised 2 undead hell hounds and a fire elemental.
Initial two settlements aer building marketplace/swordsmen.

1502 Feb.
Fifth city built, sixth settler produced and moving. Fortress started building an Engineer then a Library. Why? Should still be building settlers? We don't have at least 2-4 already in production/in play?
...oh, that's why. Before turn 60, the fortress city stops building settlers if the AI has at least 5 cities or fewer than 3 population left.
Makes perfect sense, they need at least one city producing stuff other than settlers and this early on the fortress is the only place that can do that. They need economy and military, especially the latter to be able to fight a war. So beyond this point, for a while only other cities can make a settler.
They are only population 2 though so they can't, not yet, they are building Library and University. (one of them has adamantium, the other just wants to unlock Steam Cannons.)

1502 August
Adamant city starts to build Alchemist Guild after 2 more boats. (AI is expansionist so they are a bit trigger happy on triremes)
AI still has 6 cities and no settler being produced, but I don't think it's possible to have any more at this point no matter what. The AI is playing as intended.

1502 Nov.
Fortress starts mechanician's guild. AI still has no cities large enough to build a settler. AI hasn't built a steam cannon yet but is making his fifth Trireme. Ok, this is definitely not optimal. Time to adjust the "ships wanted" formula a little.
Replaying the previous few turns.

1503 Apr.
First Steam Cannon is being produced in one of the small cities. Fortress completed the Mechanician's Guild, now building Barracks
Adamant city has Alchemist Guild, now building Miner's Guild. Every city except fortress is still size 2.

Fortress has +0 city growth at pop 4? Looks like there is a growth script bug...time to find that and start over the whole test I guess.

Edit : Ok, so the problem was (food>pop*2) in the script instead of >= so any city with the exact correct amount of food didn't grow. Ouch. (This is why I didn't upload immediately, when a lot of new code is added, these things happen...)

[Seravy]

They still only have 9 cities in 1507 after fixing the growth bug. o_O
The other Myrran player has 12 though and the last one has 17. Rampaging monster issues? Time to start one turn at a time again.

[jhsidi]

Oh, speaking of rampaging monsters -- another trend I've seen on Myrror is neutral cities actually managing to build neutral empires. Typically what happens is:

- Neutral city spawns on a continent alone and has time to build up (on Arcanus, there'd always be an AI neighbor to quickly conquer it)
- Settlers arrive; neutral city begins spawning out raiders
- If the landmass is large, the neutral city will often manage to spawn multiple stacks of raiders
- The new wizard settlement is defeated, built up and then also begins spawning stacks
- The really bad part is, those stacks hang out on the landmass permanently, just waiting for new settlements; I've played games where I took one neutral settlement on the landmass then had to kill 4-5 armies in a row that all ran to attack
- Finally, the neutral stacks are probably also much stronger than they would be on Arcanus; 9 troll or dark elf swordsmen are a hell of a lot harder to handle than 9 high men

I saw Master level AI having problems with those neutral raiders, so I'm sure that would affect lower difficulty AI too. On Lunatic, though, the wizard AI seemed to have no issue with them (not surprising as Lunatic AIs are a massive leap in strength over Master and often run around with full stacks of gargoyles / spiders / etc very early on).

[Seravy]

There is nothing wrong with the AI as far as I can tell. (other than the continent being kinda long so there is a delay for cities built that are being sent to the end of the continent.)

1506 Feb.
This AI already has Golems in the fortress and steam cannons everywhere. Armorer' Guild is being built in fortress. pretty much every city except the last one has cannons.
Can't really cann this AI weak or left behind or anything. This is as good as it can reasonably get. Summoned creatures are a lot of gargoyles, as well as the plenty of Ghouls. 2 nodes cleared already.
Steam Cannons seem to be slowing down military advancement towards more nodes/lairs for now but soon enough there should be some stacks with only gargoyles in them.
9 cities total.

Other dwarf AI has 7 cities only, as there is no more room on their starting continent which is somewhat far from other landmasses. 2 settlers are getting near the shores of the nearest land. Gargoyles, golems, ghouls, steam cannons here as well.

Draconian AI has 14 cities surprisingly. Armies are mostly bowmen and magicians. They play Life/Sorcery so no noteworthy summons, just Nagas. This is probably easy to conquer but this race/realm combo doesn't really have anything better to use in garrisons...not the AI's fault.

I think the AI is working reasonably well. Maybe raising the amount of parallel settlers in use could be bumped up on Fair and higher landmass instead of Large and higher for high difficulty. The draconians did better most likely because their flying settlers left the initial continent faster and thus the "settler per continent" quota wasn't reached. of course having no population growth issues and less settler costs also help them and hider dwarves.

The problem is more like, what are we expecting the AI to have in 1506-1507?
Armorer's Guild units only unlock in 1505+, and they need to research then cast the spell, roll Armorer's Guild for production - they don't prioritize this without the Tree Mastery setting" -, build it, then produce the related unit, and in more than one city too, to have a relevant quantity. All of that can't happen in just 1-1.5 years, it's literally impossible. So the AI only has halbderier tier and below units - Dwarves will have cannons and golems but other races won't have anything a smart human player cannot beat easily. As far as summoned creatures go, Myrran isn't making a difference, the advantage in city count won't be active until after this date so the AI will only start to get ahead in research and skill around 1510.

Overall, what the AI has is more than adequate for the 1506-1507 but I understand why a Master+ level player would think they are weak. Halberdiers and bowmen, even magicians are trivial if you are a veteran player and while the AI has a high quantity of these units guarding their cities, they don't yet have a large excess roaming the lands and counterattacking invaders. Filling up each of 10+ cities with 9 defenders with basically no city production from half the cities is already a miracle (or more like the result of their production bonus).

This game doesn't seem to have relevant neutral cities. I'll pay attention to those in the next test(s).
But first, I'll run this same map with increased settler quota to see if that makes a difference.

[Seravy]

With the increased settler quota, 1507 January (1 year later than the last test so a bit higher numbers expected).

13 cities for first Dwarf wizard. (Massive improvement)
11 cities for second (minor improvement)
20 cities for the Draconian (minor improvement)

Overall the plane is about 60-70% colonized vs the previous ~40%.

No large difference in the tier of units, but there are more of them, although this is likely due to the extra 11 turns that passed.

It's clear the extra settler quota did improve things.
Question is where should we draw the line to apply it?

Currently it is Master and above, and Fair land size and above. It's definitely useless on Tiny and Minimal, likely not important on Small either so the question is the difficulty level.
I don't think the AI does need to be stronger on Myrror for Expert and below?