Heroes and abilities

[Nelphine]

Can you generate, oh, 1000 items easily and check that most of them come close to their predefined value? I know we had the recent discussion that some players are still getting too many items that simply aren't useful, and they're worth far less than their value in mana; so if we make that worse by having too many items not close to the budget assigned to the item, that could become a serious issue.

Actually, can we make the fourth slot item, arbitrarily roll differently?

List all the powers by value (instead of grouping them together based on type as your list seems to do); then for the 4th slot, calculate remaining value, find the cutoff for that value, and randomly roll on the next 10 highest value powers that are lower than that value?

(If it doesn't get to 4th slot, then its already used up its budget, so no need to worry about this for any other slot)

[Seravy]

Easily, no I can't, I'd need to duplicate the procedure in some real programming language.
More importantly, the whole point of the "more random items" option as is, is to reduce found item quality, so they shouldn't be as good as predefined. (in general, but on rare predefined abilities like Regeneration maybe yes, on stats and common abilities like Flight likely not.)

Note that the original procedure had no "safety" mechanism at all - it rolled 4 powers and the excess budget was wasted so a 3000 value item could end up as +1 defense +1 resistance +5 spell skill Bless.
Compared to that, ensuring 2 good powers on all items over 1500 value is a major improvement (even if predefined items tend to contain higher average quality).
Question is what qualifies as good or not, and what weights do we want, as this might be somewhat subjective. We can't have too much good for balance reasons but we also can't have too little otherwise all items will be junk.

For reference these are the current stats of the predefined set :

Flaming : 13
Lightning : 10
Doom : 12
Destruction : 9
Vampiric : 6
Death : 13
Divine Protection : 18
Exorcist : 13
True Sight : 18
Resist Elements : 11
Elemental Armor : 21
Inner Fire : 20
Stoning : 3
Phantasmal : 10
Guardian Wind : 10
Haste : 12
Resist Magic : 12
Pandora's Box : 8
Wraithform : 11
Cloak of Fear : 8
Bless : 11
Shadow : 8
Invulnerability : 18
Teleportation : 9
Necromancy : 5
Lion Heart : 14
Water Walking : 7
Regeneration : 11
Pathfinding : 9
Merging : 4
Flight : 12
Invisibility : 11
Total Attack items :117
Average Attack :3.90
Total Defense items :123
Average Defense :3.59
Total Hit items :41
Average Hit :2.10
Total Movement items :34
Average Movement :2.03
Total Resist items :60
Average Resist :2.90
Total Skill items :54
Average Skill :13.80
Total Save items :31
Average Save : -2.39

Do note these look better than they are because costs are not included - there are far more high cost predefined items than how often you find them.

[Bahgtru]

Ah that's very nice, every single sword I've found in the last 6 games was a +n sword of petrify and lighting.

[Dethedrus]

Quick question about Alorra (always my favorite bug-eyed archer). Is this a couple of visual bugs or is she actually using spells instead of arrows for her ranged attack?

And the larger potential issue... since she has a ranged slot, as per usual, will that screw with her attack?

[Seravy]

She does use a magical ranged attack, pretty much the same way as if you used Focus Magic on her. Basically she is like a wizard hero except you equip a bow not a staff.

[Dethedrus]

She does use a magical ranged attack, pretty much the same way as if you used Focus Magic on her. Basically she is like a wizard hero except you equip a bow not a staff.

I just got a different twist on Alorra in a new game... this one has constitution and might

It's too bad that she can easily end up with stats that aren't consistent with a ranged unit. Though in the 7 or 8 games I've played since starting back up, this is the first with such odd stats.

[Seravy]

Well, it's unfortunate but "Ranged" isn't a hero ability class. Might is probably the only Warrior category ability that's not very useful for her, but it's not worth disabling the entire category for that. (It's not entirely useless though - with Blademaster and Might both, she can be a pretty okay melee fighter after her ammo runs out.)

I guess we could special case the hero (or ranged heroes in general) to not get Might if there is high demand for that.

[Seravy]

I think Pandora's Box works as intended for the chance of summoning strong creatures, but in the more common scenario of summoning weak ones, we get far too many. As useless as they seem they do soak up a lot of ranged attacks and spells, so it's not as simple as 1 big creature = same total cost of small creatures. In many situations the many small creatures are a lot more effective.
So...the easiest way to solve this is to change the amount of budget actually subtracted after a summon. While we still require X points to be available to summon the creature, we can actually spend more than that, to reduce the quantity of summons.
For the time being I'll make it so that X+50 points will be spent, but I'm open to suggestions for something more elaborate while I work on other features.

[teelaurila]

Perhaps Pandora's Box is only the most out-sticking point in the root cause: The AI doesn't target combat summons with a properly reduced probability. Suppose ideally the AI would first determine whether it is winning the fight. If yes, shoot at the most opportune target regardless of permanence. But if no, do not shoot at combat summons.

And yes, Pandora's is probably too effective solo heroes enabler as-is.

On the plus side, pandora's tilts the balance towards developed early heroes instead of switching over to late-game champions.

[Pinkeh-_-]

Im currently doing a game where i got a Pandora wand early on a fairly weak druid(who is now a super hero) and while its not game breaking, i find the main issue is that it it summons twice per turn( not every turn tho so im not sure how it works exaclty) yet it allows me to turn a hero and 2 behemot into a full stack in 2-3 turns. Add a regen cloak on top and i cant lose most fightd unless im highly outmatched. Having it summon at the start of my turn only would make it more balanced imo