Current Impossible Strategy

[Nelphine]

Interesting about your preference for expansion. I play on poor minerals so I almost never worry about building settlers (usually only after I've already got at least one amplifying tower). The exception is if my capital is not on a shore, I'll build one settler for warships. (Although my current game I picked up archangel before I got a ship yard, so I'm using them instead of warships. Archangels are so beautiful as escorts.)

Basically, humans can't produce fast enough. You need to spend all your production on troops and amplifying towers, so you make the AI do all the settler building, and then take them away once they've grown into hamlets with at least sawmills. The majority of the time I completely ignore expansions as threats. AI settlers and the cities they turn into are just my cities, but growing with huge production and growth bonuses. I don't conquer them until I don't need to particularly worry about garrisoning them; generally, the fortress falls first, then destroy as many major unit stacks as required so that significant garrisons won't be needed.

The main reason to conquer cities earlier, and therefore waste time with garrisons, is if a different AI is doing particularly well, and you need more power production to keep up with them. (Usually this happens if the AI manage to get nodes - an impossible AI with a single node is about as good as the human with 6 nodes, chaos omniscient and astrologer. If they get 3 nodes things can go very poorly indeed. Note that I always play on max power in order to have the strongest node defenders possible.)

As a side note, my current game, 1407, I have 9 nodes, my second main offensive stack is being led by an archangel, I've got consecration cast on 4 of my cities, and I have enlightenment cast. Despite having 3 very rare spells in play, the game is not won yet. I'm actually weaker than the only remaining arcanus AI, and the myrran AI has been free to expand without any neighbors. (Myrran AI are really hit or miss though for being dangerous. Myrran races just grow so damn slowly.)

[Nelphine]

Catwalk, is water walking plus increased max pop better than dropping the nature books, picking up a second death book for wraithform (for flying ships - costs more per individual than water walking but much cheaper per stack), and a third life book for endurance?

I tried 2 chaos for a while, but found I could rarely afford to cast flame blade (too busy with spirits and city buffs, but I play with more life than you; you wouldn't need to spend time on city buffs, so flame blade should be viable). Unless you're doing flame blade during combat?

[Nelphine]

I've decided to try to conquer my last opponent without him taking a single city. Specifically, whenever I conquer a city, I use the offensive stack as a garrison. I'm also trying to hold every single myrran node.

The first few years he stole a few nodes from me, and managed to sneak out a tower and destroy a city on arcanus. But i've now got full garrisons on each tower, and I'm in the process of getting a full garrison on each node in Myrror. I only banish him whenever I get too frustrated with spell blast (only 1 sorcery book and he has spell blast).

But, I am very pleased to report that he actually hit the same sorcery node garrison with 3 fire storms and then 4 stacks of 9 units (all in the same turn) and managed to take it away from me. So despite my power garrisons, the AI can focus enough to do this, at least with chaos and trolls.

I'm concerned though that he won't end up with any strong summons. 5 death 5 chaos 1 sorcery, myrran, archmage; dark elf, draconian, beastman and troll cities. (Dark elf was capital so the vast majority is dark elf.)

I've seen pestilence (I have consecration) and a few very rare combat spells; he's got zombie mastery and doom mastery, which is very good for the AI.

But without warlord or life, I don't think he can threaten any of my main stacks (7 bezerkers, 1 archangel and 1 unicorn) seriously. Any of the ones I want to actually keep alive I pour lots of bless and lionheart and endurance on; and it definitely becomes 'either I can dispel the buffs and barely hurt the units, or I can be strong enough to hurt the unhurt units'; he can't do both at the same time.

[Nelphine]

Also, Impossible AI continues to try as hard as it can.

April 1414:
My army strength is highest, Sharee is about 75% of mine; but it was around 60% when we first met, and I don't lose many battles against her. Doom mastery plus war mammoths/nightblades is just awesome.

Power production, Sharee is about 70% of me.. but was around 50% when we met.

Spell power, Sharee is around 60% of me (she's archmage), and has pretty much been holding steady there since we met.

Now, consider: This is fair land size. I have 36 cities. (1 dark elf, 1 troll, 2 barbarian, 5 orcs, 3 klackons, 5 lizardmen, 19 halfling)
This is max power. I have 18 arcanus nodes and 9 myrror nodes. Remember i have Astrologer too (that's 756 power production right there, without the actual node power, equivalent to over 30 fully developed cities), and 1 chaos book Omniscient.

I know of at least 2 other Myrror nodes that haven't been conquered yet; not sure about the other 1, but they aren't on Sharee's main continents so my GUESS is she doesn't control them.

That I know of she has 23 cities. It's probably closer to 30 but I haven't fully explored yet.

So even if she has the unaccounted for node, she still only has 1 node and ~30 cities to match against my 36 cities (ok, only 34 of these are helpful, since 2 are barbarian) and 27 nodes (plus astrologer and omniscient). Now, true, most of her cities are dark elf; with ~25 dark elf cities averaging population 12, that's an extra 300 power that I don't have. And she has dark rituals and is finally being liberal about casting it.

But the fact is she's gaining on me; impossible AI just won't quit!

Edit: OMG HE GOT GREAT DRAKES! THINGS JUST GOT REAL! (That and Spell Binding. How the heck does he have spell binding??)

[Catwalk]

I've seen your setup and explored it a little. It takes a little longer to get going IMO but certainly offers powerful units in a reasonable time frame. I tend to focus on obtaining critical strength as soon as possible. If I ever find the time we should try a series of games on the same maps with our respective setups and strategies. Alas, my time will only be decreasing for quite a while.

I do indeed place some importance on expansion, at least for the sweet spots. With Omniscient, I think those are very good investments, often better than marketplaces.

I'm also not sure if having both Holy Weapon and Flame Blade as early buffs is doable. The damage increase is tremendous, though.

As for Water Walking, I really like it for early speed. You can spot neutrals in no time and gather your troops effectively. Wraithform is prohibitively expensive early on, and it doesn't help horsebowmen as much as it does wolf riders.

[Nelphine]

I agree about wraithform being prohibitive on units - that's why I use it almost solely on ships, that fly the stack around. 75 mana to give an entire stack pathfinding and water walking is huge. It does require getting that ship first which is either 110 production (if you don't get a big chunk of gold right away) or ... 690 production? Something like that to get your first warship.

However, since I use alchemy, and I specifically don't use heroes (except to sit in my capital and grant overland casting skill and power production and nobility), that means I can destroy the first item or two I get - and getting 1500 gold/mana/items from a few lairs is extremely easy.

I suppose it does rather depend on getting a half decent sized continent for your first continent.

(By the way, have you ever defeated an impossible AI, after.. 1407 or so... without razing most of their home continent? I just find they have too many piles of things lying around, so it's easier to just run around their home continent burning everything. In my current game I could easily defeat them by doing that, I'm just actively trying not to do that.)

[Catwalk]

How fast would you say you take neutral cities, and how many can you get before the AI snatches them?

[Nelphine]

As per my comments earlier, I don't care terribly about expansions. Going with this setup its all about getting nodes, which means bezerkers and flying warships. I prefer bezerkers over other units because they hit air, move 4 with endurance, and they're cheap. Horde would be better for late game with buffs (and they're the only one cheap enough), jackal riders are better if you can deal with the air thing, paladins air, trolls air, dark elf and draconians are actually both good. But ALL of those are too expensive. I can build a bezerker every one or two turns, starting about 3 turns after I take my first or second lair.

And bezerkers have the advantage that the ai thinks they're super strong and won't attack them. So one bezerker will hold a node until an intercontinental death stack finds it.

Bezerkers also usually guarantee my army strength is higher than any AI, even on impossible, so I can avoid militarist and 'equal military' wars.

Anyway, point of all that is: I don't need neutral cities - any city that's not barbarian is really a node that has low power output. Still very important (especially amplifying towers), but not my focus.

That being said, I've had games where I've taken 3 neutrals before 1402 (which is before I even get bezerkers). Once I get bezerkers, 'if I see it, its mine' becomes the catchphrase. But really, the goal is: take a few lairs, get bezerkers and a warship. Find enemy fortress and kill it. Replace bezerkers. Find next enemy fortress and kill it. Repeat. Only take neutrals and AI expansions if it won't slow down taking enemy fortresses. Without buffing the bezerkers, I can reliably take two arcanus wizards. Since I usually have to buff the bezerkers for the third one, its ~1406-1408 before the third one dies.

With the first two, I collect all the expansions after the fortress dies - the first is almost invariably defeated from first fortress strike (not banished), the second is a toss up, which means they're just a bunch more neutrals. The third one I usually have to raise the home continent.

About the time the third one dies is when I start colonizing arcanus - usually ~2/3 of it is already settled by AI cities I stole, so I pick the best race (halflings if I can) to cover the rest.

Lairs are wonderful. Some games I already have 3 very rare spells before the third wizard dies, and I've already spent tens of thousands of gold and mana.

Note, if a life wizard 'survives' on myrror, I still consider that a kill. They're weakened far too much to be a threat, and they actually weaken the myrror wizard.

[Catwalk]

What is your evaluation of horsebowmen vs berserkers? Half the price and no fighter's guild needed.

[GermanJoey]

Do Horsebows really have the punch to deal with lategame threats? A lightly buffed up berserker will murderlate anything that gets into melee range with it and doesn't need to have nearby adamant. I've done lategame with massed centaurs though... it works pretty well but still needs that Fighter's Guild.