Heroes and abilities

[Nelphine]

I don't think heroism should make it 2-3 times as good as without. I'd really rather not base it on level. 

Or actually, what if you base it on the color of the heroes level? In other words, instead of a jump each level, you make it a jump each tier (level 1/no circles, level 2/white cicles, level 5/gold circles, level 8/red circles). That still leaves heroism + warlord as super strong, but at least that's almost as much committed to your build as it takes to build this power, and heroism alone is no better than gaining one level.

Values might be 100, 130, 280, 550.

Personally, I think your top end is too high - an item that can can summon an overland very rare creature? That's often better than an item that can cast annihilate 4 times. And not only does it not cost you a spell, or your hero a spell, but it doesn't cost an action, and can even happen before you get to play a single turn. Action efficiency alone makes this incredible.

I'm also not sure why you want it to be based on level at all. No other item power is, although I guess things like divine protection are, kind of. But defensive abilities are expensive. And things like spell charges don't scale, and those are easily the most similar abilities to this. Spell charges are already very strong - why do we need an ability that is much cheaper than most spell charges, AND scales as your hero gains levels, so that you don't need to make/find a better item?

[Seravy]

I could make it not be based on levels but then it would be more powerful, so it would need to cost a lot.
But it's a Chaos item power. The realm already has plenty of good late game options, so that would make it pointless. Especially as Chaos has no way to protect heroes in the late game, but can use them fairly well in the early/midgame in some cases.
So it's scaled by level because we want it cheap enough for Enchant Item, but good enough to be worth creating and equipping in all stages of the game, despite the "random" part of the deal.

Furthermore, other item powers make your hero better (exceptions apply) - so they scale up as your hero is stronger. That extra hit, attack, defense, invulnerability, immunity, haste, etc  all contributes more on a higher level hero than a lower level one. This however has an effect that isn't one of those "throw into the big hat labeled hero and let them amplify each other's effects" - you get additional units. A hell hound isn't +4 swords and hits on your hero, they are a separate unit. One that either manages to be relevant on their own or not, and more and more likely not as time elapses in the game.

Furthermore, where is the fun in "random creature" if you can be sure it will always be 3 bottom tier ones?

What makes this fun is you can't know whether a cockatrice shows up, or 4 hell hounds, or two gargoyles. Or if the hero is high level, maybe a sky drake. 

If heroism does prove problematic, I'll be willing to take the extra time to make the effect ignore it and calculate the unit's level directly from experience. (so no warlord or heroism or crusade, but you still get famous levels)
However this isn't trivial so I rather try it without at first. (It's not like you'll actually have this item before your hero gains a few levels anyway. Even then, we are the ones who decide which predefined item(s) come with it, if any.)
I'm fine with Chaos/Life/Artificer to be able to get better benefit from it through heroism. Not like that combination was ever a thing before, maybe it becomes playable now.

On the very rare creature thing, I estimate you have like 20% chance for one to come out on a level 9 hero on their first summon. And that's it, if they wasted the points, next time they don't have the budget for one. (maybe we could use 620 instead of 660 so even if they pick the cheapest common they don't get another shot at a very rare? Actually since when does a single very rare matter when you have level 9 heroes?)

[Nelphine]

So why not do the level tiers instead of every level?

[Seravy]

Because it's not fine grained enough?
Levels 5 to 8 is a far too large gap, that's like, 90% of the entire game's duration.
It also makes heroism warlord even stronger while normal heroism weaker (you'd get a value meant for levels 5-7 instead of just level 5. Which means it would be higher.) - I don't see that as a improvement, it just moves the problem elsewhere.

Also, there is no reason to. If heroism DOES become a problem, which I'm not expecting, I'm going to make the effect use base level instead, I already offered that twice. Alternately we can change the numbers on the first 4 levels specifically. I can't really judge this without having used the item in a few games.

Done adding the predefined items, cheapest with the ability costs 1400. Not exactly the price you'll find in the early game where you don't have a high level hero.

Ultimately, the item power costs 400. Even if you have a heroism hero, and get the best possible roll of 5 Hell Hounds, that's not so great for 400. You could summon 10 of them from that much. I don't expect the item to be particularly great until you have level 6+ heroes. (Albeit suiciding the creatures then retreating is a powerful tactic, so is equipping a stronger ranged weapons to do the same)

[Seravy]

So we might consider substantially increasing hero recruit chances with 0 (or even 1 and 2) heroes. So that one could even somewhat pick and choose the first hero(es). But more importantly stashing gold for a hero becomes more reliable.

I have been thinking the same, delaying heroes until turn 30 was the correct move, but only having a few % chance per turn even after that to get one, eh. And you even if you do refuse a hero you don't want, you r chances of getting another are low.

So again I think we need to set goals first, make a formula based on it next.

Let's see!

-I think it's a reasonable goal that we want 1 hero every year to offer for a non-famous wizard for their first slot.
-Famous shouldn't double the hiring chance - we don't want it to be mandatory for hero strategies, and if it does double then it's either mandatory, or the hiring chance is too high with it. Do note your heroes will appear at higher level - if you get too many offers, you get to hand pick the highest level heroes too easily, also each refused offer raises the level of the refused hero by one. So I think famous should be reducing the wait time between offers by 25-33%. (+33 or 50% bonus to chance.)
-Fame already increases hero quality offered, plus it's easy to have 20 fame at the start of the game, so it shouldn't have a very large influence, especially because it's easy to have hundreds of fame at the end of the game.
-Each occupied slot should probably increase the wait time for a hero exponentially, with less increase on the first, more on the later slots. Something like a 40% increase would mean 1, 1.4, 1.96, 2.744, 3.814, 5.37824 times longer waiting.

So let's merge these into one formula.

Chance to hire a hero, 1/X, where X=12*3*(1.4^(slots taken-1-0.5*trunc(fame/25))) /(2+1*Famous)

Probably needs to be capped at X=9 or higher. (So famous with all slots free is the best, and additional fame past that can only go towards removing the penalty for filled slots. Basically each 100 fame makes it as though you had one more slot free)

Defense in this case isn't just armor, it's all of the different things including agility, divine protection, and tactician.

And Counter Magic, Invisibility, Regeneration etc...

[Seravy]

Rethinking the above... so if you can have enough hero offers to even be picky on which hero to hire, what's the point of Summon Hero? Wasn't that supposed to be the solution of not finding heroes through luck?
I'm pretty sure that's why I refused to raise the chances before.

[Nelphine]

After seeing this in play for a while I do think the chances are too low. Multiple games with 1 offer in 15 years (with high fame) I think is too low. Admittedly, most games seems decent, but I don't remember the last game I hired 6 heroes before 1420. And my criteria is 'caster'.

In fact I don't remember the last time I had 6 heroes with caster (and before 1404 I'll take noble and certain other melee heroes too) from hiring or from treasure before 1420.

I think I'm probably averaging 3 casters.

Now, obviously since I'm rejecting certain melee types that's skewing the numbers a bit, but I don't reject them early, and there are more casters than non casters.

Note: the games I'm saying I've only had 1 offer in 15 years, those are actually one offer - I'm not discounting any heroes even if I would say no for that.

So I think the chances are too low, albeit not way too low.

They're definitely lower than prior to the famous change.


As for summon hero, I always skip to summon champion. However, as we know my hero style is unusual, I don't think that's super helpful.

[Seravy]

I also feel the current system isn't good but I find it hard to pinpoint what would be better.

So maybe we should try to define what we don't want to happen, instead, and try to get a solution from that.

1. We don't want games where a player is playing an early game hero strategy but gets no hero offers at all, or so few that they aren't able to even refuse the worst of the options in the early game.
Ideally, I think being able to choose 2 our of 4 should be good enough - with that much material, any hero strategy should be able to function otherwise it's a bad strategy.
...so we don't want fewer than 4 hero offers between turns 30 and 60.

2. We don't want the player to be able to "upgrade" to the highest tier hero available (or highest experience level for Famous) by repeatedly refusing to hire someone into the first slot. Assuming fame 20, that's about 1 in a 4 chance to get one of them?
..so we don't want 4 or more hero offers to happen in an affordable amount of time (say any period of 50 turns), even if all hero slots are empty.

ok so these two contradict already. Maybe we need to change hero offers to weight the chances by fame requirement, so higher fame shows up less often? Otherwise you can reliably hire a champion at a 1/4 chance (so 3 refusals) as soon as you have 40 fame.

3. We don't want so many hero offers that casting Summon Hero becomes unnecessary even when playing hero strategies - the spell should be desirable to cast.
So ideally we don't want the player to find or hire more heroes after the first 2 or 3. The existence of prisoners already contradict this goal...
Alternately we could rely on "but the player will lose heroes". Except, not really. You either have the knowledge and equipment to keep your heroes alive, or...you don't want heroes/aren't losing them because they are all in your fortress, giving you skill.


4. We probably want players who aren't playing a hero strategy to get less offers, but as those players won't be saving up gold, I think that already happens no matter what, so we are good with that.

5. We generally don't want more than a fraction of the existing heroes to show up each game, at least during the relevant time period before Summon Champion (so before turn 180?) otherwise the games get redundant. (If you always get to play one of your ~4 favorite heroes that's probably still fine, but not always the top 1 or 2.)

Anything I'm missing?

[Settemio]

you can reliably hire a champion at a 1/4 chance (so 3 refusals) as soon as you have 40 fame.

There are costs to this strategy: you need to hold gold and not spend it on improving your cities or military, and you need to have no current heroes at the time you have 40 fame, otherwise the chance of being offered a hero goes down so while you still may only need an average of 3 refusals before being offered a champion, those offers come spaced further apart. So, if you've cleared out several lairs and rescued one or two prisoners, and you want to force the game to offer a champion, you would need to dismiss any heroes and have been saving gold by the time you hit 40 fame. And it may not offer one of the champions you really want. I think that's a reasonable cost.

[teelaurila]

A couple ideas:

Hire weight based on player fame per hero fame, in addition to min fame. Eg. Very large player fame makes all heroes equally likely.

Summon hero gives you the chance to refuse once, and get someone else.