Early game strategies not balanced?

[Catwalk]

Of all things, neutrals?

We've already done that. Neutral cities are smaller, and for their size, come with stronger garrisons. Pretty much anything over 4 pop will have at least some ranged unit, where possible, magical. I don't think I can do any more to improve that, unless we decide even the tiniest hamlet deserves elite magicians in garrison, or neutrals do not come in small size at all, ever. I don't think they are a large threat to balance now, not any more than the remaining chance for easy nodes. (also, some races are just easy to conquer because they do not have good defending units, even if they have anything available. You can still take out 9 Jackal Riders with a few sprites, or a dozen phantom warriors.)

More sworsdsmen...well, that works. They won't get disbanded though, and I see two possible outcomes :
-The AI sees the city has enough garrison and is less likely to build troops. The swordsmen stay there instead of halberdiers replacing them, making the city easier to conquer for a while, even if the AI did rush the fighter's guild.
-The AI does produce the halberdiers, so the swordsmen are pushed outside. Ultimately they get used to attack targets - most likely the human player.

Certainly, stealing 1 pop cities is a problem of its own, but that's something you do without needing to rush. A single Gargoyle or smart use of your combat spells can do it. If we want to fix rush, we need to do something about conquering larger targets, where those rush strategies do matter.

Making neutral cities smaller is part of the problem. I'm happy to see 7 neutral cities at sizes 1-4 as that means I get them fast. Neutrals should bet between size 4 and size 12, and garrisons should be bigger with more military buildings present. Only two Arcane races have no ranged units (Gnoll and Klackons), so Sprites are not a big threat here.

Neutrals are how I start snowballing, and I can do it consistently. This is an easy and non-intrusive way to make early offense less potent, compared with the alternatives that you've listed.

[Nelphine]

I dunno. Wouldn't change my game at all. Doesn't really matter what the neutrals have, they don't have spells.

[Seravy]

It's not only two races.

Barbarians
All units have low resistance. If you have any resistance targeting spell, especially cloak of fear, you win.
They have no access to magic weapons. If you have weapon immunity, you win. Berserks might still do some damage but not that much.

Gnolls
No ranged. Medicore resistances.

Klackons
No ranged, medicore resistances. Only Stag Beetle is capable of being a real threat but it's still easy to kill (through resistance or illusion) and gives access to a free armorer's guild if defeated.

Lizardmen
These guys move slow. A single horsebowmen with guardian wind can take out the entire city.

High Men
These are good to go - priests are resistant, magical, ranged, and don't require a top tier building.

High Elf
These would need a top tier building (wizard's guild, fantastic stables or armorer's guild) to be a threat. I don't want those to be on the table frequently form the start, not only means if someone does conquer the city they win big, but also means raiders will be top tier units in the early game.

Orcs
No real outstanding unit here, wizard's guild or fantastic stables...

Halfling
These have no melee units. Protect yourself from ranged and you won.

Nomads
This one works, priests like for High Men.

[Catwalk]

From the two reports you posted, you lose out by not prioritizing neutral cities much. You underestimate how much they do for your economy. My mistake in game 1 (3) was not to have a credible semi-mid game threat and going on the offensive anyway. I had amazing economy, I just misread my build a little. I can do it even faster with Lizardmen, and Javelineers are a better semi-mid game threat.

[Catwalk]

Seravy, I invite you to try out an Impossible game where you go for neutral cities early on and report your results. I'm curious to see how fast you can rack up 3-4 cities consistently. You misinterpret what it takes to make them difficult to take early on cheaply.

In particular, Cloak of Fear and Wraithform are nowhere near as amazing as expected when most size 4+ cities have magic ranged units. They work sometimes and they're good, they don't work consistently.

[Seravy]

Certainly, the problem is you can pick out the ones you can conquer, and the spells that make it happen.
Magical ranged units? put Resist Elements and maybe Holy Armor on your unit and you've won.
Of course it depends on your books, if you don't have nature, magical ranged is a problem. If you have no sorcery, bow is a problem.

On the other hand... 3-4 cities consistently? Sometimes there aren't that many on the entire map. And definitely not in the area I can reasonably reach before other wizards. If it has "weak" garrisons, the AI will get them first unless it's very close. Even if they don't, I have no way to defend it on an enemy continent 20 tiles away from my starting location.

[MrBiscuits]

Certainly, the problem is you can pick out the ones you can conquer, and the spells that make it happen.
Magical ranged units? put Resist Elements and maybe Holy Armor on your unit and you've won.
Of course it depends on your books, if you don't have nature, magical ranged is a problem. If you have no sorcery, bow is a problem.

On the other hand... 3-4 cities consistently? Sometimes there aren't that many on the entire map. And definitely not in the area I can reasonably reach before other wizards. If it has "weak" garrisons, the AI will get them first unless it's very close. Even if they don't, I have no way to defend it on an enemy continent 20 tiles away from my starting location.

Yeah I played a game recently trying out this neutral strategy with normal land size and there were no neutrals on Arcanus whatsoever.

[Nelphine]

I don't lose due to not going after neutrals. I lose when there are no lairs to get me bezerkers early. Give me bezerkers (war college and fighters guild) in 2 barbarian cities, and I win impossible. I don't need economy, i need one boost of gold (a total of ~2000 gold).
I don't go after neutrals because I can't plan for them. So instead I try to use the same strategy each game to demonstrate it.

[Seravy]

...I went ahead and generated three maps, and looked at them, without playing.

Map 1 : 3 neutral cities total on the map. All 3 are roughly 12 tiles from my starting city, and two share a continent with another wizard (more like a large island), those are most likely going to benefit that AI, not me. All of them are small and easy to get. (tho the halfling one has shaman)

Overall, 1 neutral for me, 2 against me.

Map 2 : 7 neutrals here. Two on my continent, both easy. One at a nearby island, also easy.  One on a small island with another AI, this won't be neutral when I get there. Three more at "unreachable" distance, it's extremely unlikely to get those before someone else, and even then, not in the early game.

Overall, 3 for me, 4 for AI.

Map 3 - 3 neutrals total. One too far. One big lizardmen packed with javelineers - this is already the kind you want. One small, close enough but on an island. Probably would get it if I was playing.

Overall, 1 for me, 1 already hard to take, 1 for AI.

To be fair, these were all Fair landmass maps. Smaller would yield fewer neutrals, larger would yield more.

[zitro1987]

I am not too successful at rushing offense, with the exception of fairly close neutral towns (various tactics can work) and occasionally a few lairs (easier with nature). I think the game is set up pretty well in this case and can also, for balance, reward strategies of mass-producing your own cities with settlers thanks to sawmill-buying.

It seems the most successful players at this display excellent logic and comprehensive knowledge of 'Caster of Magic' mechanics. The average player will not be anywhere as successful. Making rushing difficult will sadly punish the average player.