Difficulty levels

[Seravy]

So. I need to play more extreme games. New resolution! (I usually play one or two extreme games after a bout of hard or impossible - after a hard, I always win, and think, OK impossible it is. After an impossible I seem to get wrecked by a 9 stack of sprites and bears and go ugh same early game issue as impossible, and drop to hard.)

The change in militarist should have helped that problem somewhat. Nature's strong early game can be harsh against Chaos or Death wizards who have low starting relation with nature but Death can get rid of some bears or sprites easily if needed (or summon their own forces of undead). Chaos...not so much.

Random events are disabled on Easy (even if enabled from the menu). Is this one necessary? It looks like it simply making Easy difficulty level unnecessarily boring even for the beginners.

Not at all, it's a leftover from the original game. Although...not sure how this works, I think it turns it off when loading/starting the game only, and it's possible to re-enable it. I don't play easy much, I'll need to look at this later.

[zitro1987]

I'd say channeler is not the most threatening retort in the hand of a very hard difficulty AI.

That is because of the maintenance reduction bonus not being as significant if AI already spend only about half as much on maintenance on the harder difficulties. It may only reduce upkeep of a stone giant by 1.5, a v rare summon or decent global enchantment by 3-5. I'd be more scared with retorts such as archmage, mana focusing (multiplies with AI difficult bonus) warlord, etc.

[Seravy]

I'd say channeler is not the most threatening retort in the hand of a very hard difficulty AI.

That is because of the maintenance reduction bonus not being as significant if AI already spend only about half as much on maintenance on the harder difficulties. It may only reduce upkeep of a stone giant by 1.5, a v rare summon or decent global enchantment by 3-5. I'd be more scared with retorts such as archmage, mana focusing (multiplies with AI difficult bonus) warlord, etc.

My thoughts on every retort :

Alchemy - Best - Allows the AI to fill mana reveres from gold alone, letting it spend the majority of income on research and casting skill. The gold not spent on buying troops makes minimal difference when the AI already has double production, but faster research and more casting power do. Add the magical weapons for early game (which the AI doesn't particularly go for on its own) and this is probably the strongest AI retort right now, even without factoring in longer Time Stops.
Warlord - Good - With the massive amount of units the AI produces, each unit being slightly stronger is a big deal...and it's sometimes more than slightly, an Elite Magician is hard to kill in one shot due to double health, and that matters.
Channeler - Good - Every mana not spent in battle is a point of power available for research and skill instead. Maintenance is irrelevant on high difficulty, the AI barely pays any, but it does spend thousands of mana crystals in combat. Also makes the AI resistant to mana draining tricks.
Archmage - Average - Late game battles are often decided by who runs out of casting power first, and more overland spells too. However fueling the casting power eats mana so this probably cuts into research a lot and makes the AI more vulnerable to long, unimportant battles draining mana.
Artificier - Worst - While the AI is much smarted at using artifacts, ultimately this still means there will be more stuff on their heroes to steal, and less other spells like Sky Drakes being cast.
Conjurer - Average - Like Archmage, allows summoning more stuff, but this one doesn't work on other spell types, and the effect is free, doesn't reduce research potential.
Sage Master - Good - Higher tier spells are no joke in this game, getting that Armageddon or Call the Void a lot faster can decide the game. Yes, I did say a lot, as the AI will spend slightly more on research so the overall speedup is more than 33%...the downside, ofc, less casting skill.
Guardian - Good - The player has to attack and take AI cities to win. Even if they go for SoM, they will need to eliminate some wizards to have the economy to get it done first. And the bonus from this is massive. Pretty much any player except Chaos will be severely slowed down.
Famous - Bad - The AI already has a high chance for heroes, loses its items to the player often, leaving only the more mercenaries and earlier high fame hero useful effects.
Runemaster - Good - If the player uses some enchantment based strategies, they are in big trouble. If the AI has strong enchantments, such as Magic Immunity, Invulnerability, globals, city curses, same thing. Neither of those happening is rare. This is unfun to play against so I rather not make it more frequent.
Charismatic - Very Good - The AI will be able to ally with other AI faster, make peace with them faster, and turn the game into a 4v1. What makes it extra dangerous is, unlike personality, this does not allow the human player to make an ally out of the AI. Might not be as bad in the new diplomacy system but probably still is. We don't want to make this more frequent, 4v1s aren't fun.
Specialist - Very Good - Faster research, more casting power, dispel resistance all in one. Chance of picking it is already very high if the wizard has only one realm and that's when it is best.
Tactician - Good - Mini Warlord.
Inquisitor - Probably bad - The AI is unable to expand its territory by conquest so if they had the potential to develop into the greatest threat, they might get held back. They are a lot more dangerous in the early game though, both because they'll have more units form the extra gold and because anything lost to them is gone for good.
Cult Leader - Average - Don't think there are any extra reasons that make this any better or worse for the AI than the human player. Can be scary on a Death wizard though, if they have Dark Ritual.
Mana Focusing - Average - The extra mana means the AI can spend more power on skill and research - but they most likely won't because the retort makes them want to spend more on mana. It does mean they'll be less likely to run low on it, and it gives a stronger early game but overall Sage Master is probably better.
Astrologer - Good - AI tends to get nodes early and double power is a lot. Especially considering the bonus to power income. Similarly the bonus casting skill is significant because they get a bonus on overland casting anyway. A lot of extra casting skill can be wasted if the wizard already has 500+ skill and the difficulty reduces spell costs, but at that point they've won anyway.

[Seravy]

Changes planned before starting more test games :

-Hard 12 picks, Extreme 13 picks, Impossible 14 picks
-City growth 100 for Normal, 120 for Hard
-No random events on Easy : Remove.
-Moving backwards before attacking enabled for Normal but Stalling is not.

On random events : Investigation shows loading any Easy game disables the random events immediately, but this remains invisible to the player. As soon as the Settings screen is entered, settings are reloaded from the configuration screen, which turns it back on (if it was originally enabled), making the impression that it was never turned off when in fact it was. Note that in the original game, this was the "Intro" difficulty level where it made sense to have no events.

On early game : If we do not want the early game to be so dominant in the game, reducing the AI's chance to pick Inquisitor (double gold for early buildings and troops) or removing the retort altogether (Even if the player picks it, it means they MUST play an early game strategy, as the retort pretty much disables late game.) can be a significant shift in the right direction. I'm not sure about it though, it's good for the game to have a retort that limits players to one race.

[Nelphine]

I actually take inquisitor specifically for late game strategies. Its how I won my impossible games originally - pick a race viable in late game, and burn everything else. I actually think its very suitable for impossible due to the disparity in early game strength making it unlikely to get enough neutrals to make a difference mid game.

However, I do not find inquisitor makes much difference to early game strategies for AI. They already get so many resources that I don't see a huge difference with it. 9 stack bears or sprites really don't need inquisitor.

[GermanJoey]

Same here, Inquisitor with a well-rounded race is one of the best ways to win late. High Men, Beastmen, Draconians, Trolls... Inquisitor with these races is so good. In fact, I'm surprised that you would even consider it an early-game retort, as I would default to thinking that Inquisitor with an early game race (barbarians, gnolls, etc) as pretty much some sort of challenge game.

That said, the AI kinda sucks with it unless they get really lucky with geography.

[Seravy]

Same here, Inquisitor with a well-rounded race is one of the best ways to win late. High Men, Beastmen, Draconians, Trolls... Inquisitor with these races is so good. In fact, I'm surprised that you would even consider it an early-game retort, as I would default to thinking that Inquisitor with an early game race (barbarians, gnolls, etc) as pretty much some sort of challenge game.

That said, the AI kinda sucks with it unless they get really lucky with geography.

Interesting. Not being able to conquer cities means where any other wizard can increase their territory, you cannot, which should have been bad for late game (you can by settling but it'll take another 100 turns before the city is of the size it would have been if you could just conquer it...and that's without considering buildings.)
I meant you play a late game race but with the extra gold you play it as though it was early : You can buy that armorer's guild or whatever you want and produce the late game units at the time barbarians would make berserkers. Hadriex's first impossible win was using this strategy, buying early doom drakes.
I'd say the larger the land size, the worse Inquisitor gets, because it means missing out on more enemy cities, and building your own will be behind even if you spend all your extra gold on it. But maybe I'm wrong? Playing Life for Stream of Life might fix this problem. Gaea's Blessing probably helps too. Nothing in the other 3 realms though.

About the AI, idk. I used to think they're bad with inquisitor but in my last game two of them had it and the game was really hard.

[Nelphine]

Late game, for me, hasn't come down to trying to match the AI income. If I try to do that, i have to expand to early, and I fail. Late game on impossible comes down (for me) to being able to have a mana income high enough to deal with 6 or so combats per turn, and having 3-5 stacks that can't lose to anything the AI has. Once you've reached those two goals, you don't need to expand. Inquisitor is ideal for that type of strategy.

[GermanJoey]

Same here, Inquisitor with a well-rounded race is one of the best ways to win late. High Men, Beastmen, Draconians, Trolls... Inquisitor with these races is so good. In fact, I'm surprised that you would even consider it an early-game retort, as I would default to thinking that Inquisitor with an early game race (barbarians, gnolls, etc) as pretty much some sort of challenge game.

That said, the AI kinda sucks with it unless they get really lucky with geography.

Interesting. Not being able to conquer cities means where any other wizard can increase their territory, you cannot, which should have been bad for late game (you can by settling but it'll take another 100 turns before the city is of the size it would have been if you could just conquer it...and that's without considering buildings.)
I meant you play a late game race but with the extra gold you play it as though it was early : You can buy that armorer's guild or whatever you want and produce the late game units at the time barbarians would make berserkers. Hadriex's first impossible win was using this strategy, buying early doom drakes.
I'd say the larger the land size, the worse Inquisitor gets, because it means missing out on more enemy cities, and building your own will be behind even if you spend all your extra gold on it. But maybe I'm wrong? Playing Life for Stream of Life might fix this problem. Gaea's Blessing probably helps too. Nothing in the other 3 realms though.

About the AI, idk. I used to think they're bad with inquisitor but in my last game two of them had it and the game was really hard.

Buying up to advanced units very fast is one option, yeah, while another is getting lots of settlers and infrastructure really fast. (as long as the geopolitical situation permits it) The more cash you have, the more you can cash rush, which means the more turns you can build housing to get really big cities (which are what bring in the really big bucks). Also, while an inquisitor civ won't have as many cities late game as a regular civ due to the fact they can't keep most conquered cities, they'll still end up settling a lot more themselves because they can rush settlers so much faster. Finally, they actually don't need as many cities to keep up late game because their economy will be really awesome already. (the one place they'll lack in less Amp Towers). But, of course, they'll still need some sort of good unit to carry them through the late game and can find themselves facing a crisis if their good unit doesn't work well against the enemy. (for example, High Men Inquisitor fighting Draconians)

[Domon]

Changes planned before starting more test games :

-Hard 12 picks, Extreme 13 picks, Impossible 14 picks
-City growth 100 for Normal, 120 for Hard
-No random events on Easy : Remove.
-Moving backwards before attacking enabled for Normal but Stalling is not.

On random events : Investigation shows loading any Easy game disables the random events immediately, but this remains invisible to the player. As soon as the Settings screen is entered, settings are reloaded from the configuration screen, which turns it back on (if it was originally enabled), making the impression that it was never turned off when in fact it was. Note that in the original game, this was the "Intro" difficulty level where it made sense to have no events.

On early game : If we do not want the early game to be so dominant in the game, reducing the AI's chance to pick Inquisitor (double gold for early buildings and troops) or removing the retort altogether (Even if the player picks it, it means they MUST play an early game strategy, as the retort pretty much disables late game.) can be a significant shift in the right direction. I'm not sure about it though, it's good for the game to have a retort that limits players to one race.

Is there any other way to make inquisitor work?

I dunno, conquering cities bring them down to 1 population with no buildings and converts theme to your race, instead of auto-razing?

Or, different races' cities capped at half population/half growth?

Or again, different races' cities get a negative population growth modifier, something like -100 or maybe -150, forever.

If this proves too powerful the gold bonus could go down to +50%

Of course, i'vo got no idea about how feasible are these mods!