Fewer wizards on Myrror = easier game?

[jhsidi]

The idea that No Overlap makes the game vastly easier sounds speculative to me. This seems similar to how Plane of Earth is somehow rated as 0.5x score despite there being no substantive evidence at all that it's a huge benefit to the player. I believe you said you played it yourself and didn't notice that the gameplay was particularly easier or harder, yet the score penalty suggests that less ocean = same thing as the AI literally being hard coded not to attack you (Sandbox Mode, at 0.4x).

So specifically with regard to Myrror and No Overlap your assertion seems to be that 1 more space results in:
- AI not being able to build large armies or have enough resources
- AI not being able to navigate the vast distance (wait, it's just 1 more tile)

... but is that based on intuition, or evidence?

As far as I can tell, the main issue here is the travel time to destination and the sheer amount of unsettled space available for 3 AIs (max). When it can take from 20-40 turns just to reach a distant location (particularly when massive oceans are in the way), then another 10-20 to found the settlement, grow it and set up a defense, why is it surprising that the AI has poorly defended settlements on the distant landmasses during the early-mid game? They're too far away to gain any benefit from the stacks on the home landmass, and the only thing they can build early on are the early game units like swordsmen.

I could also imagine that No Overlap might make the issue better -- with the locations filled on their home continent faster, the AI can move to the more distant locations earlier in the game and start building there. It would take a side by side comparison to tell.

[Seravy]

Plane of Earth is actually going up to 80% score in the coming update and it definitely doesn't help the AI anywhere as much as I expected.

Smaller land size is already worth less score. Whether it means easier game or not isn't really speculative, it's the metric we choose for calculating the base score. Conquering more territory is more work, even if it's not necessarily harder work although in the past 5 years it was painfully obvious how much easier the game is when there are fewer land tiles and enemy cities.

No Overlap does the same thing, except instead of directly reducing the amount of tiles avialable, it increases the amount of tiles required for each city. The net result is the same. You have to conquer fewer enemy cities, so assuming that is the metric for base score, it should be worth less, regardless of whether it is actually easier to do or not, but it very likely is.

In particular :

- AI not being able to build large armies or have enough resources

This is a proven fact - the AI consistently performs weaker on smaller landmass settings in the DOS version. It's not really the "large armies" part though. It's that they can be taken out or crippled in fewer moves due to fewer cities, so a counterattack is less of a threat. Fewer cities also mean less research speed, so the AI is stuck playing low tier spells much longer, removing the pressure from the player to advance their spell research, making any early game focused strategy, which people already almost always play for higher difficulty, stronger.
And while the effect of course also affects you, it still means you are playing a shorter game which automatically makes it easier in my opinion because there is less time/chance for human error - and exploiting human error is the only way the AI can win as it doesn't have an overall strategy.

- AI not being able to navigate the vast distance (wait, it's just 1 more tile)

No, the AI is perfectly fine navigating the extra 1 tile distance.
But that means their first city is 1 turn later. The cities in the next "circle" are 2 turns later. The next one, 3 turns later.
When your own strategy doesn't rely on using settlers to expand or generally, win the game, this is a direct advantage to you/disadvantage to the AI, as their cities will be a few turns behind in progress while your own strategy that relies on conquering nodes, neutrals, or other players, is mostly unaffected. It's a small effect, but it's there.

A larger effect would be the drastically increased probability of the AI losing territory to not optimizing settlers for maximal number of cities. While on a 4 city distance map, that's usually fine and the overlap mechanics guarantee almost no tile gets unused, with a 5 distance the amount of wasted tiles is much higher and while the human player will optimize and always build 2-4 cities where they can instead of one in the middle blocking the entire area, the AI does not, leading to much greater losses proportionally.

As far as I can tell, the main issue here is the travel time to destination and the sheer amount of unsettled space available for 3 AIs (max). When it can take from 20-40 turns just to reach a distant location (particularly when massive oceans are in the way), then another 10-20 to found the settlement, grow it and set up a defense, why is it surprising that the AI has poorly defended settlements on the distant landmasses during the early-mid game? They're too far away to gain any benefit from the stacks on the home landmass, and the only thing they can build early on are the early game units like swordsmen.

That's exacty how it is and is why it's expected that the AI is left alone until after they filled up the plane. It's almost like in an RPG where you attack the to-be final boss while they are still level 1 and busy getting to 99. The difference is in an RPG you literally can't do that because the final boss is not there to fight until the game near the end.
Reminds me of Warcraft where you could ambush the other player while they were building their initial cities with like, 3 swordsmen and win the game. Don't think there was a good solution to that either other than agreeing not to do that with the other player because we actually want to play for more than 2 minutes.
Master of Orion solved this by not spawning the final faction on the map (Antarans, was it?) until the endgame I believe.

[Anskiy]

I dunno I think trying to claim that breaking open a tower in 1506-1508 on higher difficulties is a "rush strategy" is a bit rich. After all most towers are actually not very tough for a competent army to take out, and opening the other plane is a legitimate gamble since it does give the player more access to land, but also contact with other wizards who may or may not like them. Besides, I've found that on higher difficulties you *need* to expand through conquest at around that time, else the AI just leaves you hopelessly in the dust thanks to their absurd resource bonuses.

[Seravy]

What am I supposed to do about it then?
Make towers all have 9 Sky Drakes?
There is no way a Myrran player with 3-4 times as much territory to settle than Arcanus players can possibly be ready to fight in 1507 at least not in the sense of having no vulnerably cities or continents.

[Anskiy]

Well maybe alter the Myrran AI specifically for when the equal wizards on planes thing is not selected? Or atleast remove the scoring penalty for the equal wizards thing, since so far that I'm seeing, that option is actually more challenging than not having it on.

[Seravy]

Ok, let's try to summarize.

On lower difficulty everything works fine. Players don't break towers before 1510+ and by then attacking the Myrran player instead of other Arcanus players is more difficult (as intended) and isn't a good strategy.
On high difficulty, players have to play better to win so they are able to break towers much earlier. The Myrran AI isn't ready until at least 1-3 years later. This makes attacking Myrror first the winning strategy which is not intended.
When the player doesn't break into Myrror first, the difficulty works as expected, even on high settings as it gives the AI the time needed to develop.

If all of that is true, we can't make the Myrran AI stronger. Any advantage we give to them for the early game carries over and makes them even harder in the late game.
So then we get a case where attacking the Myrran wizard early is difficult, but attacking them later is outright unwinnable. So it just makes attacking Myrror first even more of a winning strategy and we want the opposite.

...we need a solution that makes breaking a tower early undesirable unless playing on World Equalized/Race to Unknown/I am the Boss options, we don't need to buff the AI, UNLESS they also underperform on the other plane past midgame. (1510+)

Is that correct?

I don't see how breaking the tower can be made undesirable though as it rewards you with treasure. Maybe a diplomacy penalty that specifically ONLY applies on high difficulty that causes instant war with your plane if you open a tower? Harsh but other than being forced to use your troops in a war to defend your home plane, I see nothing else that could lead to a solution.
Maybe a new war declaration type with a new roll, for opening towers early, could have something like this :
"I cannot watch idle as you open a portal to a different world and unleash untold horrors on our world. I must take control of this situation by adding the area around that cursed tower to my empire. Prepare for war!"
Or the same thing as a warning only but with a large penalty that easily escalates into war.

Ultimately, the AI is given extra territory on Myrror which takes extra time to expand into and exploit. We need to find a way to provide them with that time. The older versions of the mod had stronger towers but that wasn't very popular so now towers are much more reasonable, but leads to them breaking in 1505-1507 if a player really wants to.

[Anskiy]

Well I think a possible solution would be to make the Myrran wizards have a low initial opinion of any Arcanus wizards breaking into their side, and vice versa. That way you would be almost guaranteed to have 3 enemies on max AI settings when breaking a tower to Myrror early, and it would also make the Arcanus wizards legitimately recognize Myrran wizards breaking into their world as a threat too.

Though I must add that forcing the player to defend with most of their army is hard to do, in practice the best way to defend is to simply gobble up more territory. After all the AI is eventually going to make stacks that you can't reasonably defend with generic city garrisons, like 9 water elementals, shadow demons etc, so a few cities being lost is to be expected and is a reasonable price to pay for heavily damaging the AI in turn.

[massone]

Ok, let's try to summarize.

On lower difficulty everything works fine. Players don't break towers before 1510+ and by then attacking the Myrran player instead of other Arcanus players is more difficult (as intended) and isn't a good strategy.
On high difficulty, players have to play better to win so they are able to break towers much earlier. The Myrran AI isn't ready until at least 1-3 years later. This makes attacking Myrror first the winning strategy which is not intended.
When the player doesn't break into Myrror first, the difficulty works as expected, even on high settings as it gives the AI the time needed to develop.

If all of that is true, we can't make the Myrran AI stronger. Any advantage we give to them for the early game carries over and makes them even harder in the late game.
So then we get a case where attacking the Myrran wizard early is difficult, but attacking them later is outright unwinnable. So it just makes attacking Myrror first even more of a winning strategy and we want the opposite.

...we need a solution that makes breaking a tower early undesirable unless playing on World Equalized/Race to Unknown/I am the Boss options, we don't need to buff the AI, UNLESS they also underperform on the other plane past midgame. (1510+)

Is that correct?

I don't see how breaking the tower can be made undesirable though as it rewards you with treasure. Maybe a diplomacy penalty that specifically ONLY applies on high difficulty that causes instant war with your plane if you open a tower? Harsh but other than being forced to use your troops in a war to defend your home plane, I see nothing else that could lead to a solution.
Maybe a new war declaration type with a new roll, for opening towers early, could have something like this :
"I cannot watch idle as you open a portal to a different world and unleash untold horrors on our world. I must take control of this situation by adding the area around that cursed tower to my empire. Prepare for war!"
Or the same thing as a warning only but with a large penalty that easily escalates into war.

Ultimately, the AI is given extra territory on Myrror which takes extra time to expand into and exploit. We need to find a way to provide them with that time. The older versions of the mod had stronger towers but that wasn't very popular so now towers are much more reasonable, but leads to them breaking in 1505-1507 if a player really wants to.

Although stronger towers may be unpopular, I am personally in favor of this option, or at least having it as an adjustable setting. In particular, I think that if Towers are meant to be opened late, then they should be harder than almost all nodes. This means:
1) They should have Very Rare creatures, and multiples of them.
2) There should be a combat effect that makes them harder to fight in, similar to node aura, but stronger. For example, a strength 70 counter that applies to all Realms, and unaffected by Astrologer or any retort, just like the actual Counter Magic/Philosopher's Stone. And it goes away after being opened, so wizards don't have any defensive advantage when keeping units on the Tower. Or it could something like a special "uncontrolled planar distortions" effect, causing wizard's creatures to take irrecoverable damage every turn.

[Seravy]

...actually, if stronger towers are enough to be a solution, isn't it the easiest to simply recommend changing the tower budget in modding.ini?
I mean, I can of course hardcode a bonus to the budget that scales on difficulty, or add it to modding.ini but ultimately that's just a convenience feature at that point, those who want stronger towers might as well just raise the number for themselves even as is.

Would it solve the problem though? It for sure makes the tower unbreakable for the AI even way beyond when they would be allowed to do it as it forces the AI to have a 9 stack of very rares to do it but how much does 9 great drakes stop the human player?
Sapher has repeatedly demonstrated that taking out 9 drakes only cost like 2 nomad horsebowmen and some fire bolts or whatever other common spells he used to do that. There is no such node or tower garrison that can't be cheesed.
Which brings us to the counter effect which is pretty much mandatory and has no alternative solutions. But having a counter effect on a tile that's not a node and one that doesn't get affected by the spell and retort specifically meant to do that, idk, not a fan of that idea.

...a more drasctic but actually working solution without side effects would be to not have towers at the start of the game and have them spawn over time. But that also comes with an issue in the sense when the game tells you "cannot build cities on towers" and the tower is hidden until 3more years that's super confusing. And also, where do the towers come from? Ok, maybe they were invisible and the magic is wearing off or something... still feels a bit unnatural. I think the diplomacy consequences are more intuitive and might work better.

[Cruel100]

Well I think a possible solution would be to make the Myrran wizards have a low initial opinion of any Arcanus wizards breaking into their side, and vice versa. That way you would be almost guaranteed to have 3 enemies on max AI settings when breaking a tower to Myrror early, and it would also make the Arcanus wizards legitimately recognize Myrran wizards breaking into their world as a threat too.

I like this idea. And I also think stronger towers might be at higher levels, above master. I also prefer less sea or ships with more movement points.