Test games played

[Nelphine]

Could we reduce powerbonus, and instead give AI a discount on combat spending? That way their power wold be more balanced when at peace, and the waste enemy mana in combat syratgey would be harder.

Something like 1/3 of the power bonus would go to combat cost so for expert power bonus would reduce to 155%, and combat spells would only cost 80% of normal.

[Seravy]

Not that high. They still have lairs to deal with, which is notociably relevant until late uncommon.

Still higher than arcanus wizards at war, but lower than arcanus wizard at peace.

In theory that's true for the spending but the treasure will pretty much always pay for the spending and more so in the end it doesn't mean a thing. Losing 200 or 300 mana to find 1000 is...a benefit, not a loss, unless units die in those battles and AI nowadays is probably good enough to not lose too many of them.

I have been thinking and rares in 1410 actually match our timeline goal. Well, technically it would be 1412 for the rare tier to start, but Expert AI is supposed to be slightly ahead, plus, that's the time when 2 rares are done, and they only had one. This has two major implications :

1. The AI managed to follow our intended timeline. Despite resource availability being only 50-65% of what is needed to achieve that. This shows the AI advantage is excessive, it is not only compensating the AI for the difficulty level, but also the missing amount from inadequate design while the human has to deal with that. So the AI power advantage is actually this high not only because they waste the power, but also because they, well, would be unable to afford casting properly the same way as the player always has been, we just didn't notice. So fixing power income on the whole might allow lower power multipliers for the AI.
2. Rares in 1410 feels excessively early when actually playing the game. So our timeline of ~6 years/tier is probably wrong. We might want to aim for 7 or 8 years instead. Unless having an excessively strong start, progress at this rate just isn't viable. Producing units to use up on conquest just isn't that fast, so unless there is a "no losses doomstack steamrolling" action going on with no proper garrisons in conquered cities (which the AI is no longer letting the player to get away with - in the example game I didn't lose because of the graphs, I lost because they found my main food producing city and it didn't have strong enough defenses (4 skeletons, a catapult, a pries,t a crusader and 2 swordsmen). If we increase the "Turns" parameter, that raises the number of creatures summoned in the tier but I can live with having to use uncommons or rares for an additional 6-12 summons produced. (46, 39, 39 respectively instead of 40,33,33 if we aim for 7 years, 52,45,45 if we aim for 8 years. Note that this is pure creatures - in reality you'd split the skill and use other spells as well, so you'd actually only get to have 50-70% of that.)
Furthermore, if "Turns" is higher, that will result in slightly lower necessary income values, which is good. It won't make a significant difference, but any difference is better than none.
Finally, you mentioned it might be better worth pumping power into mana, converting it to gold and buying research producing buildings if research is too slow. So any raise in "Turns" should consider that and incorporate it in the RP building costs. For example, A Sage's Guild produces 9 RP.  But it costs 4 G to maintain, which would cost 16 RP to create each turn, so assuming you are making money from power regularly to pay for your needs, it's a net loss to do it, even if the initial cost of the building is free. Even as an Alchemist it is only +1 RP a turn, which is not worth the 1200 RP initial investment. Universities will be similar. It might be worth investing into no maintenance, cheap buildings however, like Marketplaces, but normally you can build those from production which is more available and cheaper. So I don't think we need to worry about this, building from power is not viable for most buildings in the game. Military and units might be the exception if those forces allow you to do some good conquest but that involves risk.

Could we reduce powerbonus, and instead give AI a discount on combat spending?

So, like, lose only 1 mana crystal for every 2 spent in combat? That might actually be a good solution albeit it effectively means all AI has Channeller for free. Not sure I like that.

[Nelphine]

When i said i'd convert to gold, and buy buildings, i meant, i'd convert to gold, buy buildings, and literally never do any research. If I can just go life, buff city units like magicians, nightmares, or hammerhands, why should I spend the massive amount of required resources on researching rares? And if I'm not going to bother with rares because they're too weak.. why would I bother with uncommons? So then I just don't do any research at all, and avoid building any research buildings that don't do anything else (which allows me to put even more production towards strong city troops). (And with current research costs, this is already a problem; it's not quite something that is necessary to do, but if research costs go up by about 20%, and buildings did not, I would probably resort to simply never putting naything in research and relying entirely on starting spells for the whole game.)


For the ai discount, it would be more like only lose 4 mana for every 5 spent in combat. (Assuming you reduced power bonus from 180% to 155%). Which would mean it wouldn't quite that much due to rounding problems.

[Seravy]

Because we balanced out Life superbuffing a few updates ago. Meet a stack of ghouls or Nagas and the superbuffed unit is dead to poison.
Well, okay, not the Armorer's Guild unit. But you can already win by doing that, it's in the known problems for a reason. Regardless of the research cost. It's a fact that Armorer's Guilds come too early and probably are underpriced one way or another. We just haven't been able to fix it.

...maybe we have to just face reality and make it cost more anyway. (If the progress between spell tiers gets balanced (meaning slower than now), getting the AG units another 10-20 turns later is not that big deal. Whether it's worth upgrading to them and whether slow races have mid tier units to use meanwhile, is a question though still.)

[Nelphine]

IF you're going to increase research costs, then yes you'd need to increase city building costs. (However, on that note, if you do so, you need to increase fighters guild costs. But city troops weaker than halberdiers can't fight war bears and ghouls and naga, so you'd need to readjust those downward so that the human could even get units to fight those summons; so this entire process may just be a nice thought process, but the balance ramifications may go too far.)

[Seravy]

Playing the first game using the new research system.
Advanced difficulty, all Fair, 6 Death 5 Sorcery Astrologer, High Men.

No lucky nodes this time. I've set Magic Market to be a mandatory building before race specific military buildings but after smithy, 4 swordsmen and any settlers. The AI was set to have a power advantage equal to their gold advantage, meaning quite a bit lower than before (power was the highest category).

1404 now. I mostly focused on research but built some Magic Markets as well. My power production is roughly equal to the other two AI's I've met. One of them was High Men and I have quickly taken over their entire territory using Ghouls - the AI seems to have no answer to them at all. I wonder if a higher resistance race would have made a difference, but High Men sure are completely unable to do anything against Ghouls in the early game. I haven't attacked this wizard's fortress yet. They are Nature/Life so I'll likely find it full of sprites eventually, and a stack of Ghouls will be able to beat that. Considering the low difficulty and lucky enemy matchup I think this powerful start might be acceptable but it might also be a warning sight that Ghouls are too good. Not just for conquering  all of that but also for a good 10-15 undead units. Crappy ones but at this stage of the game, even 2 swordsmen matter a lot, as they can be left protecting a conquered small city while the ghouls take over another.

While I was busy doing this, the other wizard (Silver) attacked me from behind. I was able to block her settlers at a chokepoint earlier so they couldn't get into the area I wanted to settle, but now they sent nagas. The first stack of 4 attacked my capital thinking my skeletons are too weak to defend, but luring them there was actually the plan, as Skeletons counter nagas very well. I feel if these tried to attack my smaller cities, it would have been a smarter move on the side of the AI as I would have lost one, maybe two before killing that many Nagas. Of course, if I was a complete beginner who doesn't know how to counter Nagas, it would have been the good move from the AI, so it's probably better to leave it this way.

Another stack showed up with 2 Nagas and 2 magic spirits, and this time one of the Nagas had Focus Magic. Fortunately, they were within 1 time of 2 of my Ghouls so I attacked and beaten them easily with the ghouls and a zombie. So now I have a Focus Magic undead Naga.

By the way I'm researching AEther Binding as first pick, as it happened to show up by luck. This does prevent me from researching anything else for the first 50 or so turn even with going high priority on libraries so it might not be the best move, but I didn't need any of the commons I could have picked. Focus Magic Ghouls is not my cup of tea, they kill everything they need to even without the expensive and unnecessary Focus Magic. (and I have Darkness to support them if needed. Thanks to the new magic building, I could afford to push skill up to 25 without a node. So much more fun to play this way!)

   

I haven't lost a city yet, so I'm not at war with Silver, but unfortunately they refused my Wizard's Pact offer even though I gave them a Phantom Warriors spell. Sadly my military force is still slightly below hers, so there should be more Nagas in her cities and probably stuff other than those as well. They are Nomad ans stole the Halfling neutral city I was after, but I did get a High Elf one. As a somewhat longer term plan, I want to build Pegasai to take out the 9 Earth Elemental node nearby. I don't think there is any lair around that I can attack using ghouls and skeletons, at least nothing big. Maybe the one with 1 sprite and 2 bears when I've dealt with more important issues.

[Seravy]

Using the undead Magic Spirits I got, I started scouting the map. I soon found Flandre, the last Arcanus wizard. 6 Death 4 Chaos 1 Sorcery and Archmage. I was able to get a Wizard's Pact for free because we share 7 books and my army size is higher. She is Chaotic though so can't trust it too much - but it does prevent random corruptions unless a war starts. I managed to trade for Wraith Form in exchange for Blur. Silver and Flandre are at war which is good news and explains why I've only seen 6 Nagas so far.
I suspect we might need to limit the number of Magic Spirits an AI summons based on turn count. 3 of them this early is kinda excessive and wasteful, probably 1 for every 30 turns elapsed is good enough. I rather not tinker with it until we perfectly balance common creature balance though - unlikely but we might end up wanting the spirits to have a more relevant role.

[Seravy]

I finished researching and casting AEther Binding, and now researching Mana Leak. Already summoned 9 ghouls ready to strike the green capital. Meanwhile got into a war with red, but haven't lost a city yet to them. So far I have been focusing on research and magic, now I started building some crusaders as the enemy is playing ranged races. I also completed the first Pegasai. Meanwhile some undead Sprites I got have been wearing down the Earth Elemental node but the pegasai will be excellent against nomad pikemen, while the crusaders will handle horsebowmen and the skeletons will deal with nagas, at least this is the plan. Once I have enough forces around there, I'll start advancing into red territory.
   
   

Red has two nodes I see, yellow has one but based on the graphs they might have another as well. I have a wizard's pact with them so I'm not going to send scouts there.

Green and Yellow allied but I offered peace to green so I don't need to worry about losing the wizard's pact for that. (Yellow is still chaotic though)

The 9 Ghouls, despite the support from Mana Leak and an ongoing Bad Moon event, still can't beat the 8 sprites and 1 bard hero in the green capital. I was kinda expecting this and planned to reload but wanted to see the battle as Ghouls are on the "maybe overpowered" list. While they should in theory counter the sprites especially with mana leak and the event, the city walls and heavenly light are strong enough to let the sprites win anyway. I wonder how they would work against the Nomad wizard with their good resistance but I'm not going to find out - I'm not going to go for a poor tactic like that when I can use the much superior skeletons for that war. (Missile Immunity is huge against Nomads).
...now that I think about it, Magicians might work better than crusaders for this enemy. They are outright immune to missiles, and this wizard will not use strong direct damage spells to kill them unlike the last game. Although, against focus magic units, they might be less efficient than crusaders so idk.
A fighter's guild + alchemist guild combo is slightly cheaper than a wizard's guild though, but it does not produce anything so idk.

[Seravy]

Decided to use my 9 ghouls that fail to work on the wizard's capital to try a chaos node with some hell hounds and a chaos spawn. I failed to kill the spawn but damaged it enough that it run into a corner, allowing me to kill it next combat. So the funny thing is, I didn't need Focus Magic on the ghouls to actually get a Chaos Spawn (I did need Darkness though). This gives me even more doubts about FM+Ghouls being a problem, it's more likely that ghouls themselves are. So far they have given me...some 6-8 bears, 5 sprites, at least 6-8 magic spirits, a dozen swordsmen, cavalry or bowmen, a few horsebowmen and pikemen. That's amazing albeit if I consider the cost of the ghouls I could have had over 50 skeletons which are a similar military force and has no maintenance (the undead summons still cost a lot of mana.) so it might be fair due to the high price (and high risk, if you lose the quite fragile ghouls, you lose a very expensive unit! It might not be that easy to lose them against the right type of enemy but they die fairly easily to nomad armies or chaos wizards or even things like big groups of sprites. Even Hell Hounds eat them if you miscalculate and have fewer ghouls than needed.)

Conquered one city from red, found and conquered a neutral on an island (lizardmen), and got two nodes, soon getting a third.

On the turn 72 mark I was somewhat behind our goal on research, even though I built libraries, but I did spend a lot on wars, and had no nodes and barely any treasure during these turns, so I think that shows the system works fine. Now that I have nodes, I can research even uncommons in 6-8 turns if I push up the research bar (don't even need to max it) so it doesn't feel too slow. I could have uncommon creatures already if an uncommon summoning spell showed up on the list, sadly it has not. I'm hoping to reach Shadow Demons as early as I can but military spending and lack of nodes held me back so far. I expect that to change soon, as 3 of my cities are now producing meaningful units, and if I really want to, I can even build catapults almost everywhere which might actually be not bad against all those pikemen. So I'm probably fine even if I hold back on summoning and combat spending albeit with my new nodes I don't feel the need to do so.

At this rate I might get ahead of the planned "one and a half wizard defeated" goal for turn 144 unless something unexpected happens.
   

[Seravy]

1409 August, roughly halfway through what we designed as the "uncommon phase".
I was able to form an Alliance with Flandre which allowed me to scout the entire Arcanus plane finally. I made progress in conquering the other two wizards. I reached Shadow Demons and expect to take over the rest of the Arcanus plane using them. Thanks to the demons, I made contact with the Myrran wizard. They have Focus Magic Cockatrices which means I won't be able to use Shadow Demons against them.
Flandre declared war through Chaotic, so I'll have to start fighting her, but with Shadow Demons that is unlikely to be a problem. She doesn't seem to have any threatening forces, the difficulty setting simply isn't high enough for that. She does know Gargoyles but I've only seen like a single one of them summoned in her capital and no others.