Caster of Magic Release thread : latest version 6.06!

[m59gar]

Just downloaded 505. Where is this settlers menu?

[Kaiku]

Could the two tile warning during a Wizards Pact be looked at? It's nice enough in theory, but only a bit annoying and frustrating in practice.

Scouting with a Magic Spirit? I'm amassing troops at her border. (I can't know if it gets in range of a city as I can't see far enough ahead of me).
Conquering a forward city from a wizard she's at war with, then she plants a city next to it? I'm amassing troops at her border.
She plants a city next to a node I'm guarding? I'm amassing troops at her border.

Kali desperately wants a Wizard Pact with me (but under no circumstances a Peace Treaty), but this makes it impossible to maintain it. She breaks it, then comes begging me for a Pact again in a few turns... Before breaking it.

No human player with evil intentions would aim to taunt and piss her off like that before declaring war. We'd just waltz in and fight her. So while the warnings make sense from a realism standpoint, it doesn't work in the game due to reasons like the above, making the Pact often a hindrance to relations with the Wizards you want to be friendly with. I don't believe I've ever been 'legitimately' warned, when I'm actually planning to stab them in the back. It's always due to 'misunderstandings'. Is is possible to add criteria to the warning? Otherwise I don't think the AI is equipped enough to use this mechanic properly.

[Seravy]

Just downloaded 505. Where is this settlers menu?

I meant to write "settings".
Basically if this ever gets added, it will not be part of your save file and could be turned on/off in the middle of the game which is bad.

Could the two tile warning during a Wizards Pact be looked at? It's nice enough in theory, but only a bit annoying and frustrating in practice.

Then it works as intended. That's the price you pay for the benefit of not getting attacked by the wizard. Why would it be free? The AI isn't getting anything at all out of it - you are free to attack even the turn you break the pact so it doesn't give them any "safety".
There are exceptions btw : Settlers and Engineers don't trigger a warning. Everything else does.

[namad]

Could the two tile warning during a Wizards Pact be looked at? It's nice enough in theory, but only a bit annoying and frustrating in practice.

Then it works as intended. That's the price you pay for the benefit of not getting attacked by the wizard. Why would it be free? The AI isn't getting anything at all out of it - you are free to attack even the turn you break the pact so it doesn't give them any "safety".
There are exceptions btw : Settlers and Engineers don't trigger a warning. Everything else does.

I find this annoying as well. The problem is that human's don't necessarily want as many cities as possible, as some locations just kind of suck, but the ai will settle any tiny crack in your borders. The only way I can see to fix this would be to make it so that cities cannot be settled as close together as they can now. Make it the required distance between cities 1 higher than it currently is. It would be a large change to the game, but it would give bored humans fewer things to micromanage AND it would make wizard pact's make more sense (as they could then add 1 additional square).

However the pact can't add a square of range without the change I mention to cities, otherwise it would just give the human player free range to settle a vast number of uncontested border cities for free.

[Seravy]

I fail to see the logic there, we can add 1 additional range to the pact without requiring the same on city distance but extending the area would just make it even more annoying on the human. Wizard pact range has no effect on AI behavior so it cannot possibly give any advantage to the human.

[Seravy]

Regarding the "too many AI settlers" problem, I found the thread where the Fighter's Guild AI rule was added to counter it, here : http://www.realmsbeyond.net/forums/showt...1&page=182

Now, while I do think the amount of settlers was correct before and that still required adding the 2 free settlers, now that we have it and modified the "Fighter's Guild rule", we probably need to restart the discussion on how many AI settlers we want to see.

My current view is a bit different though - back when the "Fighter's Guild gap" was added, I assumed the AI settler production due to the resource advantage was excessive. However that never was the case, back then we simply had the "AI can only send settlers to tiles the human already scouted" bug. Obviously if all the world's settlers come flooding towards you, it'll seem like there are too many of them.

On the other hand, now everyone has 2 free settlers. Including the AI. Which means they'll have at least 2 more cities of their race than before (but this can be higher depending on how the race's FG rule changed) which essentially means AI with races they don't want to spread too much end up spreading their starting race more than before. In other words, the AI might need to learn that early military races aren't meant for settler spamming (but all others still are).

...except, there is this wall that the AI can't decide to attack others at will because there is a Diplomacy system. Meaning they can't conquer reliably and have to always rely on settler rush even for races where that's a bad plan.

Unless we are willing to say the AI should take the risk and skip 2-3 settlers worth of time for building up buildings/forces for those races, and hope they find a neutral or end up in a war and can then build cities of those conquered races instead. However in this case the "first turn for AI war" needs to be lowered - not building settlers for 40 straight turns which also not conquering is far too much.

Alternately we can let the AI spam Barbarian and Gnoll cities anyway, and hope the massive strategic military strength of the units produced from all those cities steamroll the map and win them the game through nodes, treasure and conquest. This might make the game more interesting (The AI will actually commit heavily to a strategy) but means the luck factor is higher (more snowballing) and might make it too easy to win for the human (gnoll/barbarian units have major weak points that are easy to exploit while the wizard will be left behind in magic power).

[Nelphine]

Considering my own optimal barbarian game is mono life (for inspirations) with 12+ barbarian cities, I don't think it will hurt if the AI has ~10 barbarian cities; gnolls can get amp guilds, so there's no reason not to have a ton of gnoll cities. (I'd rather have a ton of gnoll cities than a ton of dwarf cities.)

So you MIGHT put in a 10 city limit for barbarians, but I wouldn't worry about any other races. But you really need to reduce how many non bezerkers barbarian AI build. They should never build bowmen, and they shouldn't build more than 1 shaman per 8-12 bezerkers. (This also applies on defense. Bezerkers are flat better in every situation than either of the races ranged units.)

[m59gar]

I find the ongoing topic of free settlers and expansion very on point, since I was playing 202 until yesterday and my constant struggle in every single game was making enough Settlers as the AI expands like crazy all over the world. It got to the point where I was restarting games over and over just to get a capitol far enough away from AIs that I could make more than 2-3 cities before the 'tidal wave' of enemy outposts gobbled up all the land. Upgrading to 205 and having a free settler on Tiny and two on Huge really helped with that phase.

However, I find myself continually returning to certain strategies after finding out others don't work at all, because something seems to be off about unit balance. The vast majority of normal units are so bad that I've completely stopped making them. Nothing matters except stocking 8-9 ranged units together to annihilate all the useless melee units that never even get to attack anyone, so I always end up Halfling Slingers, High Elf Longbowmen, or Dwarven Steam Cannons. Hammerhands are laughably bad, because even though they are one of the only melee units in the game with enough defense and health to survive the charge across the battlefield, their movement is low and they can't hit air. I've actually found Hammerhands incapable of taking even low level dungeons because of the incredibly annoying new AI where the enemy just runs away forever. My poor crippled Hammerhands limp along after units that just run circles.

In the end, it seems that only ranged units or units with high flying speed matter. Pegasai are great. Summons are incredible. But normal units just kind of don't do anything.

Shrug.

[Suriname]

Note that threatening a wizard gives you a peace treaty if successful, which does count and prevents casting such spells until it expires. So if you are more powerful you can just tell them "hey stop doing that or you'll regret it" and they'll do for a while. Or course, a Wizard's Pact or Alliance is a more permanent solution if possible and desired.

While that's technically true, it's also pointless. AI seems to back off only with lower military strength. When this mechanic would be useful - at the beginning, just after the hostility breakpoint - they'll never back off and always declare war, as lower strength just doesn't happen thanks to cheating (barring berserker hacks, but then you're attacking anyhow).

In the end, it seems that only ranged units or units with high flying speed matter. Pegasai are great. Summons are incredible. But normal units just kind of don't do anything.

No worries, now high men now have a new speed 2 healer. Without ranged.
...Well, at least AI knights and paladins will be moving 2.

[zitro1987]

I do feel there's too many melee units too slow to catch ranged (especially with normal units, but even supposedly mobile units like hell hounds fare badly).

Speamen move 2 ... ok, swordsmen move 2 ... also ok, halberdiers also move 2 ... now that's a problem. Fighter's guild is supposed to make us progress, and halberdier is such a common unit.

Even glass cannons like chimeras move just 3. Melee units should have the advantage of higher movement than ranged units in average.