Test games played

[Seravy]

So current game, Chaos Pandora Elves.
It pretty much turned into a high level Thief hero abuse game. It was my first hero offer (came kinda late, but with a 833 cost that's not so surprising, if my other hero offers were also that expensive I could have missed them, only tried to save about 600) and she came at level 6 (and could summon rare stuff with the item).
..so one conclusion, Famous should be capped at a lower level. I'm considering to make it turn dependent, level cap = 4+ (turns/100). So early you get at most a free heroism, but later, it can be more. (Maybe make it turns/75?)
Using priests/shaman to heal heroes doesn't work. They are too slow to follow. Also, they would take away the slots from using Pandora. When you are using a Pandora's Box hero stack you generally want 3-4 heroes and no other units, as you need those slots for the summoning.
Some of the summons manage to draw enemy spells or creatures, others don't, so you are entirely playing for luck. If you have a high defense hero, the summons are pretty much useless, as the hero would survive and kill everything anyway. If not, then the summons might not help the hero (as they might be unable to block the flying enemy creatures from reaching it). So it's plenty of fun but I don't think it's a strong strategy, although it's playable (on lower difficulty). You most definitely need multiple heroes with the item, otherwise you might not get anything really relevant out of it - and for that you need lots of free slots too.
Unsure about the hero hiring chance being good as is, or needing a higher AI bonus than the original, as the jump from the 2nd to the 3rd hero slot is pretty steep, more so than before.
Having no answer to Crack's Call, I have to avoid attacking the Nature wizard using heroes (not that they have Crack's Call yet but better be safe), but I'm doing well against everyone else. On higher difficulty I don't think the heroes would be able to survive though (multiple reaper slashes and lighting bolts per battle).

[Nelphine]

I'd try to have it capped at the same level as heroism; then add 1 every 5 years or so, but starting in 1410. So, start-1410, cap as heroism; 1410-1415, heroism +1; 1415-1420, heroism +2; etc.

Yeah, my only major abuse of pandora's box really consists of attacking AI fortresses so the summons can soak up lightning. That's really my main concern, as that lightning was specifically designed to stop solo units from crushing fortresses because they had high stats, and heroes generally have the highest stats.

I was wondering about the steepness of later hero hiring chance. Honestly, I'd initially thought a lot more linear; 1/6, 1/12, 1/18, 1/24, 1/30, 1/36. But that ends up with too many heroes hired by the end.

So maybe what it really should be is pretty linear, but with the first 2 being extra fast. So, 1/6, 1/12, 1/24, 1/36, 1/48, 1/60? Even that's too fast.

I think you'd just need to teach the Ai to use summon hero more, which makes sense as that's part of what we wanted with these changes.

[Seravy]

They do use summon hero, but to have a decent hero in their fortress, that hero has to be hired early (otherwise no levels from turns and defending that fortress in battles).
Basically, we want the AI to have a good chance of hiring the Beastmaster early and Summon Hero is too slow for that. (It had a good chance of happening also because as the AI had more heroes, plus some died, so it had been using most of the fame 0 heroes.)
...oh old AI bonus was a flat +10%, not double the base chance. Yea, that would work, so basically we need the AI to ignore the calculated chance and always use the 1/6 to get the original behavior. (which was 14-15% during early turns with this bonus)

[Seravy]

Before starting a new game, I'll try to experiment with something. I tried increasing the number of entries in DIPLMSG.LBX. If it works, I'll be able to add more diplomacy texts because we'll have unlimited slots.
I'd really like it if the AI told the player why a treaty offer is rejected for example (too low relation to even consider, or just failed the roll), or used a different message if they had nothing to trade not you, etc.

[Bahgtru]

I find diplomacy to be too dependent on military strength. At the beginning it's quite strange to encounter a peaceful neighbour who has more than neutral standing with you but who is unwilling to make peace with you because... You're not building up your military. What gives?

[Seravy]

So in my current Gnoll test game... (5 Life 5 Nature Warlord).

I was doing really well until now. By 1410 I eliminated 2 arcanus wizards and conquered all their territory, and despite the overextension only lost two cities to razing. A pretty big contributing factor was the first enemy, Tenshi, Life/Chaos warlord, being Ruthless and triggering Jihad, so most her cities only had 6 weaker troops to defend them which is in no way enough against stacks of 9 gnoll units.

Last standing opponent is Byakruen, High Men, cities full of magicians and some Shadow Demons, but I outnumber her in cities like 26 to 7, so I should be able to slowly beat her. (Shadow Demons are a big pain, no webs. My only weapon against them is Star Fires and using my conquered Klackon cities to build Stag Beetles which will take a while to make happen)

However in 1410 I decided I'm doing really well so might as well take a little bit of risk and look at the other plane with Planar Travel. So I found Flandre on the other side, Death/Chaos, Ruthless. Well, turns out it wasn't a "little" risk but a huge one. Fire Storms started coming, plus Drain Power now twice as often as before (Byakuren already used some). So I have to keep MP maxed and research and skill on zero to even be able to cast spells.
But it's worse. Ruthless allows Jihad so Flandre also started one. Jihad allows attacking towers regardless of strength, so the turn limit gets ignored - she broke through towers 6 years ahead of schedule. Despite doing really well, I'm considering to drop this game because I definitely don't have the resources to fight on two fronts right now, and wasn't planning to fight Myrror any time soon. If I do and win, game is over with my victory and it's definitely too soon for that. If I don't win, then I lose so not much of a choice. Either way I consider this a broken game - towers should not be able to break like this anymore.

Two conclusions...
-Jihad on Ruthless needs to go. I already felt it was doing more harm to the AI than help, but with so many of its features specifically engineered for countering Spell of Mastery, it's just not suitable for generic use anymore.
-Planar Travel, we need to rethink.

So on Planar Travel, this is what I think right now :
-The AI cannot use it so why can the human? Spells that are human only shouldn't be a thing. (with the exception of spells that affect parts of the game where the AI has enough advantage to not need them, Earth Lore and Nature's Cures being those)
-It's Russian Roulette. Find a friendly Myrran wizard? Huge benefit, you can trade, make alliances, exploit lairs, all kinds of stuff at zero cost. Find a Chaos/Death/Sorcery wizard? You're up against painful curses you can't do anything about, as waging war when you need to buff units one at a time for 50 MP is not very realistic. Either you have an unstoppable doomstack to raze the entire plane or you can forget it, and will have to suffer the curses. If you do have the doomstack then it's just utterly unfair, as you get to kill the AI (and win the game as you monopolize the plane) while they still can't go to your plane.
-Rushing Planar Travel and getting an early Myrran city or even a dozen can heavily push the game towards your favor at almost no cost - the Myrran enemy will be much weaker while you'll be like twice as strong at the same time. And once more it's Russian roulette, because the resources spent towards it will be a waste and will cause you to lose if the wizard there is hostile.

So I question Planar Travel is healthy for the game. Shadow Demons, due to their high RP cost are slightly more acceptable I think.

(continued in the Life thread)

[Nelphine]

I find diplomacy to be too dependent on military strength. At the beginning it's quite strange to encounter a peaceful neighbour who has more than neutral standing with you but who is unwilling to make peace with you because... You're not building up your military. What gives?

what do you mean by 'make peace with' - if they're peaceful, they aren't going to attack you, so you're already unofficially at peace. Do you mean get a wizard pact?

A wizard pact requires investment - the other wizard needs to get something out of it.  If you aren't a military threat, and you haven't offered them anything, what are they getting out of it?

[Nelphine]

I'm continuing to try to get my 4 allied wizards to surrender to me (I have 9 cities and they have a combined 48 cities). I don't think I'll be able to do it before one of us casts SoM (at least 3 of us have enlightenment..)

[Nelphine]

aand i managed to get a freeze up with the screen showing all odd graphics again. I cast divine order. I haven't saved a few turns, so it will be a bit before I can try to reproduce it.

[Seravy]

Started a new game, same gnoll life/nature warlord.
Things went well at the start, but I might lose because I was stingy and didn't want to pay 550 gold for a wizard's pact. Seems like me and Vanilla H were expanding rapidly from the two sides while the two AI caught in between are pretty much down to their capital only. Problem is, since I didn't pay, my city got attacked and now we are at war. To make it worse she has early Crack's Call so lost my Assassin hero which I used to conquer like, the previous 5 cities. No items of high value were lost, I haven't found anything really good yet. Now I have no idea how to continue - buffing is an obvious no-go as Life buffs are prohibitively expensive and I can't afford losing a buffed unit to CC. Mass producing might work, but the enemy is lizardmen so questionable, gnolls are good but not that good. I'm still like 30 turns away from researching Transmute and another 30 for Prayer, albeit I could try pushing research a little, not by much, I'll need the mana. I could summon bears but I don't see that tactic working at this point against a pure Nature wizard. I had to abandon two conquered cities in addition to the one lost. So far no enemies are on my continent, the two AI who do have a city each are the harmless ones. I'd also need to research Stream of Life to make sure the High Men cities I got in my first war will have manageable unrest, but right now this is my lowest priority.