I found only one reliable way of dealing with myrran sorcerer (besides killing him early) -- drain his mana reserves. Attack his units on other plane (for maximum range penalty) with invisible hero, let him summon air elementals, click done 25 times. Drag your feet in other battles. Transmute his resources, burn cities. Do it long enough and he no longer can afford these juicy control spells. Without mana reserves he is a pushover.
|
Test games played
|
|
Or one could reduce the growth of Myrran races accross the board? This would also make player Myrran games harder, which I think would make it closer to the 1 pick worth.
Also the AI bonuses could shift from growth to direct production (hammers+mana+gold)? This would make an AI "in the blue ocean" weaker, while making "confined" AIs stronger - if overall balance can be struck. Which I think could be worthwhile. It would also lessen the impact of player early conquest, as you cannot gain cities that are larger than your own due to living under AI growth bonuses (though they would still gain some indirect growth bonus due to more buildings). 
Unfortunately, there isn't much room for that - population growth AI bonus is already one of the smallest.
It's 75/100/100/120/130/140 only. It doesn't really matter - 3 times as many cities grow exactly 3 times (300%) as much overall, even without a bonus. ...an AI cheating mechanism that makes it so that the AI has the same total resources regardless of the number of cities could work but that completely defeats the purpose of the game - "more cities = stronger ruler" is a cornerstone we shouldn't change. Changing it makes expansion and conquest pointless.
Right but you may be able to set them to 100% across the board (aside from easy). As you say they already have more cities due to faster settler construction (and knowing where good city spots are and prioritizing intercontinental travel for settlers). So they already have a growth bonus due to that.
The only reason why I have any bonus at all, is because settler production, which is faster on higher production, does consume population, and the AI can find itself in a position where it can only build settlers in a city that does not have an infinite supply of that. In which case, low growth hinders the AI's expansion greatly, especially if some of those settlers are killed in battles.
...maybe it would be better if they had no extra growth and just cheated on population when making settlers? 
If that's the only reason you have the growth bonus, then yes you might try other things. Personally, I'd limit the cheating based on settlers being built to cities under population 4. Otherwise, early game, it will end up being a bigger population boost than already exists.
One more idea for future Myrran AI balancing :
What if landmass wasn't evenly distributed, but instead, based on the number of wizards? Example : Currently, a Huge game has 700 Arcanus and 700 Myrran tiles. In the new system, there would be 1400 total land tiles. Divide it evenly between all wizards, but counting each Myrran wizard as if it was 1.5 wizards instead (as we still want to retain the "Myrran wizard is the final boss" system, so they need some extra). So in a "normal" game, you'd have 4.5 as the divisor. That would result in 311 tiles per wizard. Arcanus has three, so it gets 933 tiles, Myrran has "one and a half" so it gets the remaining 466 tiles. The problem with this system would be the fact that if the player does manage to conquer all of Arcanus, the Myrran wizard won't be stronger than them anymore as Arcanus is twice the size. So we probably need a higher modifier than 1.5 - but then games with multiple Myrran wizards would have barely any Arcanus land. So, idk, but we can worry about it once we have enough test results of the current situation. Maybe this idea is good but needs different numbers/conditions to calculate the amount of tiles than that example. The OPness of the Myrran wizard clearly comes from the fact that they control as much territory as 4 wizards together AND/OR they lose no economy in wars for the first half of the game to unite the plane under their control. While this is intended, "4 times as much" is the source of problem (if we conclude it is one) and the only direct way to reduce the territory they control is reducing the Myrran land tiles. Maybe a much more straightforward "If the Myrran AI wizard is alone, Myrran size:=0.66*usual Myrran size" would work better. That would make it 2.5 times as strong as a lone Arcanus wizard, or 66% as strong as an entire united Arcanus plane (but without wars their economy should be far ahead despite that). Or something along the lines of this with different numbers (maybe +20% Arcanus and -20% Myrror at the same time instead?)
By the way, currently watching this : https://www.youtube.com/watch?v=E31-Bw2ytXw
Only in part 1 but the OPness of the Myrran wizard is already obvious on the graph. (They're Life Dark Elves, their power production was like 5 times anyone else's and it's still before 1410, they likely aren't even using the whole plane yet)
mm, I think your idea is almost right, but we have to think of the (choose less than 4 AI) option.
*We would also need to count the human wizard *Wizard starts in myrran sometimes and has 2 opponents in same plane. *So plane you start as human often or always has the most wizards (including you) *So starting plane as human has to have the most land in all circumstances. I think the best quick solution is to make an automatic 66% land in plane you start and 33% land in opposite plane (or 60%/40%) *Plane you start will have the highest number of wizards (including you) *Plane you don't start will have the lowest number of wizards, often just 1 You could devise a formula depending on how many wizards, but I would assume the plane player is on would most certainly require more than 50% of land. |