07 Aug, 2012, Runter wrote in the 1st comment:
Votes: 0
I wanted to talk about some of the game I'm working ons features. Feedback and suggests on the idea is much appreciated. I'm especially looking for other places it can be used that makes sense that I've overlooked aside from areas, npcs, and items.

My generation system is used across multiple concepts so I'll talk about broad concepts it uses.
Subject = Thing being generated
Affix = A property the thing can have

Selection of Affixes
Generation requires 3 groups of affixes.

Group A: Primary
Group B: Performance
Group C: Mechanic

There's 3 rolls to determine which affixes may be rolled.

Subject must roll one and always one affix from group A.
The remaining affix must be rolled from group B and C.
Group C must only be sampled from one or zero times.

Number of Affixes
Subject quality determines affix count in all cases.
I'm using trash, common, uncommon, rare, epic, legendary.
0, 1, 2, 3, 4, and 5 affix respectively.

Magnitude of Affixes
The rank of the subject determines the maximum value an affix may roll. The minimum value is always 1.
Example: level of subject

Banding of Subjects
When multiple subjects are generated at once they must band with pairs of the exact same quality as long as there's more remaining subjects of a lower quality than what would be generated. For example, cannot turn 10 commons into 1 uncommon and 1 epic.

Given 9 commons the process is like this:

9 commons first become 3 commons and 3 uncommons. Only 3 pair up and band because 3 is the highest possible that remains supported by the sum of lower ranking subjects. Next, 2 uncommons band to rare giving 3 commons, 1 uncommon, and 1 rare

Conversion of Subjects
Any time the game converts for various reasons subjects from one to another, the conversion rate quality is the subjects quality minus 1. Quantity of conversion is equal to quality of subject. So 1-5. Consumption of the subject isn't demanded, depends on the specific API. Killing NPCs to convert to an items would consume the subject. Other cases may not.

Explanation
So this is applying to my items, my areas, my npcs, and some other things.
Conversion of subjects differs, but an example is killing an NPC. This counts as converting an NPC to item. So legendary npc guarantees at least 1 rare item, and potentially (See banding) 5 items in total.

Conversions are fleeting, so I can think of a few interesting possibilities. Converting NPCs to areas for the purpose of quests involving sleeping NPCs, their mind, demonic possession, etc. Living sacrifices to open gates to areas. Trapping living things and converting them into an item. Dungeons being converted to momentos that can be used later as keys back to a similar dungeon. Converted an area to an NPC could also be interesting. Could be used for boss generations.

Oh, I should mention that areas are "floors" and collections, following the banding rule, are floors. The quality of a floor determines, no surprise, how many affix it has, Boss quality, count, etc.

I'll post more information if I left anything out or is confusing. My main goal is to maintain a standard that players and devs can count on when it comes to the generation of stuff in the game.
07 Aug, 2012, Idealiad wrote in the 2nd comment:
Votes: 0
Runter said:
Given 9 uncommons the process is like this:

9 uncommons first become 3 commons and 3 uncommons. Only 3 pair up and band because 3 is the highest possible that remains supported by the sum of lower ranking subjects. Next, 2 uncommons band to rare giving 3 commons, 1 uncommon, and 1 rare


I'm a little confused by this. Should it read 'given 9 commons'?
07 Aug, 2012, Runter wrote in the 3rd comment:
Votes: 0
Idealiad said:
Runter said:
Given 9 uncommons the process is like this:

9 uncommons first become 3 commons and 3 uncommons. Only 3 pair up and band because 3 is the highest possible that remains supported by the sum of lower ranking subjects. Next, 2 uncommons band to rare giving 3 commons, 1 uncommon, and 1 rare


I'm a little confused by this. Should it read 'given 9 commons'?


Sorry, yes. Updated it.
07 Aug, 2012, Rarva.Riendf wrote in the 4th comment:
Votes: 0
When I started to think about random item generation, the main thing was to not make them more powerful than what was already in the game; but as powerful in a different manner, genrating items that you could not get from mobs. I think the thing to start with is to value all of your existing item.
A resist worth xx numer of point, +1 to stat y, +1 to mana y etc
Once you manage to find point that reflect the real use value of an item , and it is by far the most difficult thing, then you can easily generate random items.
All you have to do is generate a 'value' roll, go into your table and get each stat roll% till your 'value' is empty.

(oh and as a side effect it has the nice bonus that you can keep in check yoru builders very easily, as if they create an item that is off the mark, it is automatically spotted)
07 Aug, 2012, Runter wrote in the 5th comment:
Votes: 0
Rarva.Riendf said:
When I started to think about random item generation, the main thing was to not make them more powerful than what was already in the game; but as powerful in a different manner, genrating items that you could not get from mobs. I think the thing to start with is to value all of your existing item.
A resist worth xx numer of point, +1 to stat y, +1 to mana y etc
Once you manage to find point that reflect the real use value of an item , and it is by far the most difficult thing, then you can easily generate random items.
All you have to do is generate a 'value' roll, go into your table and get each stat roll% till your 'value' is empty.

(oh and as a side effect it has the nice bonus that you can keep in check yoru builders very easily, as if they create an item that is off the mark, it is automatically spotted)


I don't plan on having any non-randomly generated items.
07 Aug, 2012, Rarva.Riendf wrote in the 6th comment:
Votes: 0
Huhu that helps. I would go Diablo2-3 kind of randomization it works pretty weel, and it easily tweakable, and quite beta tested already :)
10 Aug, 2012, KaVir wrote in the 7th comment:
Votes: 0
How do you plan to determine what affixes something can have? For example, perhaps a breastplate can have "improved armour", while a sword can have "cleaving", and a mace can have "smashing". Presumably you don't want swords of improved armour, or breastplates of cleaving.

Will each object store a list of available affixes? Or will affix "sets" be defined separately, so they can be reused for similar items (eg a longsword and shortsword might have the same set)? Or will you use some other solution entirely? Will material type also impact the available affixes (eg red dragonscale giving fire-oriented bonuses)?

In addition to Groups A, B and C, it might also be worth having a Group D: Enchantments. This would allow you to define what sort of enchantment buffs can be cast on each item, so someone can cast a 'sharpness' spell on a sword or axe, but not a mace or helmet, etc. Although this isn't part of the random item system, it's a useful option that can piggyback on the same implementation.
10 Aug, 2012, Runter wrote in the 8th comment:
Votes: 0
Right now I'm putting those things into group C. I consider cleaving or smashing to be mechanic affix. So in other words, group C (or any other groups) is a collection of objects that implement a common interface that builds the available affix. So the system relies heavily on principles of polymorphism and higher order operations like collect/select. The item type object contains some of the information for this, but the implementation details are rather boring. I think what's important is the concept that group C can be built both from item types group C set, a global group C set, and any other number of sets. Maybe even possibly with granularity down to a specific item.

I've decided to actually make legendaries share the same number of affix as epic with the difference of legendary loading with access to group C affix from other types. Like meta gems on items other than helmets (if we were using diablo 3 as an example).

The group D is an interesting idea. I haven't thought too much so far about how modifying items would fit into the system through spells, abilities, or other mechanics.
0.0/8