Test games played

[Nelphine]

See, I played ghouls on large, without going after lairs (pre nagas buff, so might have changed), master, and ghouls just wrecked AI early. Take all the hordes of AI city troops, and watch your own military grow as fast as you can find fresh troops to attack.

[Nelphine]

Also, life barbarians, as per my original version of them in the impossible strategy thread, was to have 3 AI dead by 1408. I play it slower now, because I can and I like the lawful (I've realized I'm not really peaceful) perfectionist playstyle, not because the strategy is inherently slow - you can easily be a ruthless militarist with it. As a note, naga buff has definitely made the fast version much harder.

[Seravy]

I wonder whether it wouldn't be possible to normalise a little the starting conditions, by tweaking the landmass generator?I'd try to make the continents more similar in size, at any given number, if at all possible.

Why? I intentionally tweaked the numbers to make sure it's as random as possible from fewer big to many small continents - Exploration is more interesting if you don't know anything in advance.
Obviously this follows the landmass settings and is mostly true for small/medium size- don't expect 50 continents of 15 tiles each for Huge.

How are two ruthless AIs that have been at war allied is beyond me:

Yellow is Charismatic, that's a major positive modifier to making treaties. That, plus the roughly neutral starting relation from books was enough to have a fairly good chance for a treaty.
Basically, if you see Charismatic on an AI wizard, assume they'll get as many alliances as a Peaceful or at least Lawful wizard.

My guess is those two things are going to mean any game on tiny is really not a great one for strong testing.

This is mostly true because Tiny has a very low amount of resources and unlike the human, the AI can't play well from small amounts of resources, advantage or not. Naval travel is also a problem but nowhere near as bad as lack of resources. It's subjective but I'd say Tiny is about 1 difficulty level easier than Fair and Huge is 1 level harder.

Also you answered your own question - how do ruthless become friends? Trading spells. One gets weak, they beg for peace, other sees massive strength difference, agrees, then they trade a ton of spells, and now they love each other.)

Nope, the AI won't get a bonus for trading spells, and in fact they are unlikely to trade if their relation is bad. They do get a bonus for free at random but it's small and is the replacement of any other possible positive diplomacy interaction, because the AI gets none of those. They can't pay tribute to each other or anything like that.

[Nelphine]

Oh, interesting. Welp, I continue to know little about diplomacy aside from 'carry a big shovel and fortress strike them all'.

[Nelphine]

So, my death game.. I have 55 overland casting skill. It's november 1410. I'm researching very rares. 10 books + specialist + sagemaster is a HUGE amount of research. It'll probably take another 3 years just to cast any spell I research but THEY WILL BE VERY RARE.

[Tlavoc]

lmao! Been there myself. I have this amazing spell, now hold still for 20 turns while I cast it!

[Arnuz]

I wonder whether it wouldn't be possible to normalise a little the starting conditions, by tweaking the landmass generator?I'd try to make the continents more similar in size, at any given number, if at all possible.

Why? I intentionally tweaked the numbers to make sure it's as random as possible from fewer big to many small continents - Exploration is more interesting if you don't know anything in advance.
Obviously this follows the landmass settings and is mostly true for small/medium size- don't expect 50 continents of 15 tiles each for Huge.

Sounds good, I hadn't realised that was intended. Guess I'll limit testing to huge then. That's where the roaming hordes of summons should be most effective, in fact.

What do you think of my conclusions above on the subject of difficulty and cheating? The more I play the more I'm convinced that it's too much and random at turn 15-40 and irrelevant - or at least much less so - once you've gone past the turning point and you can defend yourself. Some more testing at large-huge wouldn't hurt of course, but so far testing seems to be confirming this.

Are your objectives on win chances changing? Are you shifting to desiring a more random/difficult game?

[Seravy]

You fortress spiked an AI early using a slow strategy (ghouls are expensive!) and you say it was too much advantage at early? That does not compute, it shows the exact opposite.

[Nelphine]

Arnuz, I think you're possibly just not used to the late game. The whole point is for the game to continue being difficult as wizards are removed. Which means the next AI should generally be stronger than you - once you're ahead of a given AI, they generally become much less of a threat - even with all the changes, AI still needs to be in a position of superiority to be a threat in most cases.

Although, I think I may be confused as Seravy - you've said you had no problem with the early AI, that the late AI were the strong ones, yet you're saying the power in the early game is what you're worried about, and the late game was won regardless of overall strength.

[Arnuz]

Haha bad at comms :*)

I'm saying that the early advantage was due to the tiny landmass, I actually did not fortress spike the second IA - it turtled up with megabuffed units and I could only end it when I got the undead behemoths or hydras (I don't remember which).

OTOH, I fortress spiked the first one with - lol - beastmen archers. No need to wait till I had a stack of ghouls, which would've taken forever in this setup with no resources. It had I think sprites, I had resist elements that I casted while boating there.

Basically I could not follow my usual ghoul strategy due to the lack of FM, and survived despite the AI cheating advantage only due to better island hopping strategies: first I sent 2 triremes full of archers to the first one, then I conquered each city (1/island, see the map) of the second with the ghouls, slowly, as soon as they became hamlets. That was the only way to get power. I could start grabbing some of the very strong myrran nodes only around 1404, and soon after I finished off the second wizard.

Conclusions: without tiny landmass I wouldn't have stood a chance. Therefore, cheating is excessive, outside maybe of the case of tiny landmass and enemies that are boat-bound (the second AI was life/death and the first one couldn't live to become a threat).

Arnuz, I think you're possibly just not used to the late game. The whole point is for the game to continue being difficult as wizards are removed.  Which means the next AI should generally be stronger than you - once you're ahead of a given AI, they generally become much less of a threat - even with all the changes, AI still needs to be in a position of superiority to be a threat in most cases.

Although, I think I may be confused as Seravy - you've said you had no problem with the early AI, that the late AI were the strong ones, yet you're saying the power in the early game is what you're worried about, and the late game was won regardless of overall strength.

The early ones weren't a problem only because of the tiny landmass , in this case. So yes, despite having no problems, I'm concerned with the initial (15-40) conditions. Anyway, I'll do some tests with larger landmasses to be 100% sure of this.

The late ones had the sea of arcanus literally covered in stacks that were completely irrelevant at that point, therefore no amount of cheating would be enough on its own, something more would be needed to make those stacks a threat. Or maybe, a higher cheating later in the game, so that AIs get higher rarity spells easier and remain a threat.

For example, I just had an idea. Look at the military strength bar of the last one: it's double mine, yet it doesn't scare me one bit because it won't attack my hero stacks for the stack military strength. What if they attacked despite stack military strength? When their overall military strength is so higher than the player's. This, with the idea of whittling the players' stacks down gradually.

Well, maybe this is a bad idea, it's kind of only part of the old multi-stack strategy idea I had some time ago that Seravy said is impossible.