To start crafting in our ECS, you would choose a pattern. Each pattern would have different material slots that would accept a certain type of crafting reagent. So, a dagger might have 2 blade slots and a grip slot. Different metals could go into the blade slot and a leather or cloth could go into the grip. Based on the type of reagent you placed in each slot, and the quality, you would affect the base stats of the crafted item.
So, let's say we want to craft daggers. We would gather or buy various materials, head to the crafting station and choose the dagger pattern. A window opens where I can see the material slots (2 blade and 1 grip slot) and can choose materials to place in there.
I decide to start cheap and choose 2 stones for the blade. Now stone is a terrible crafting reagent for most purposes and as I put the stones in the slots, I can see that the blade becomes slower, does less damage, has poor armor penetration, has an accuracy penalty and low durability.
I am not happy with the stats on that blade and so decide to try a better grade of material, switching iron for stone. The iron blade is still pretty slow (not as bad as the stone), but has decent armor penetration, damage and durability. The stats I get out of the iron blade are good enough for now and so I move onto the grip where I have similar choices. I can create a wool grip which does very little for my weapon's accuracy and speed, or choose a hide grip, which boosts those two stats a little.
I could see crafter skill coming into this system several ways. Each crafter would have an overall pattern skill level. So, I might have a Craft: Dagger, or Craft: Sword skill. I think crafters would also have skill levels in the various materials they work with. So, as I worked in stone or iron, I would gain skill in those specific materials.
When you pick a pattern and place the materials into their slots, the skills would affect the final result in various ways. I think both pattern and material skill would play into the overall success chance of the crafting. A master swordsmith working with a known material would almost never fail whereas a novice working in a difficult material for the very first time would possibly waste quite a bit. Next, I think a high pattern skill would give a small overall (randomized?) boost to the stats of the final product; a low pattern skill would lower the final stats. Finally, the material skill would change the rate at which the material added/reduced the item's attributes. So a master of Quickiron would get more quickness and less reduction of damage than a novice.
The basic system is very similar to what is in Ryzom and sounds a lot like what the Mortal Online team is creating. I think this system has a lot of advantages. The first is that you get thousands of combinations of materials that lead to tons of slightly different items. More items with more variation is a good thing. Crafters would tailor their choice of materials to their customers' needs, creating lighter armor for rogues, or more damaging swords for warriors.
Another advantage is that crafters with higher skill and access to better materials will tend to create better items. I could see this system leading to players who flock to Crafter Bob for fast swords, or Crafter Sally for the most durable armor. Even a mundane sword produced by a master swordsmith who had mastered the proper materials would be sought after for its exceptional base stats.
If we take the raw materials and add refining skills, then we can see another possible advantage to this system which is an effective use of combines and interdependent trade skills. Now normally, I hate excessive combines in crafting recipes, but in this case, you would not be required to use them. You could use the base materials and still get an item -- the combines would just lead to better items with higher stats.
So, let's say I am a swordsmith and I have a lot of raw iron in my inventory. I could use the iron to make swords, and get a few nice ones, but I really need to up the stats on my product to compete with what is already on the market. I could contact a metallurgist, hand him the iron and some coal and he could use his skills to make steel. I would then use the steel to make swords with higher attributes, getting an edge on the market. However, I pay for this quality by having to provide coal and a fee for the metallurgist.
Now, all of this is for creating "mundane" items. Items with character stats, procs, special damage, or other special powers are created via crafting "aspects" into our items. Every crafting material item would have an aspect score which would generally be hidden, but might be visible to characters with high crafting skills or special abilities. The aspect score is tied to an individual item, not a class of items, so if you loot 3 different rabbit's foot items off of corpses, they all have different aspect scores. Typically, low-level, common items have very low aspect scores, while rare, high-level items have high aspects.
The aspect score of an item is roughly the chance it has of imparting a specific quality onto an item. An quality would be something like "Quick: +2 speed" or "Caustic: 10% chance to proc 10-20 acid damage." Every item has a list of these qualities that it can give to each type of pattern.
In addition to the normal material slots you see when creating an item, you would also have 2-5 aspect slots based on your pattern skill level. Pretty much ANY crafting item can be placed into the aspect slots. When an item is placed into an aspect slot, the computer chooses one of the item's qualities like a reel in a slot machine (though it does not show the player). When the crafter hits the button to finalize the crafting process, the randomly chosen qualities each have a chance (based on the item's aspect score) to be applied to the final crafted item.
So... that needs an example badly... Back to crafting iron daggers.
I am crafting daggers using 2 iron for the blade and leather for the handle. I am happy with the stats I am getting, but I would like to try my hand at creating a magical weapon. Unfortunately, all I have is some granite I gathered from a nearby quarry. I have 2 pieces of granite, one with a low aspect score and another with a very high aspect. I put the low aspect granite into a slot first.
At that point, the computer looks up the possible qualities for granite and dagger and comes up with the following:
(0.5%) Stoneskin: +1 Toughness
(0.5%) Stalagmite: +1 Earth Damage
(0.1%) Statuesque: +2 Strength, +2 Toughness, -2 Speed
(0.3%) Blockhead: +2 Toughness, -1 Intelligence
(0.1%) Piercer: +1% Critical Armor Penetration
From this list, it randomly picks (0.5%) Stalagmite. The 0.5% is the chance that this quality will actually stick to the crafted item. The dismal chances are because of the item's crappy aspect score. My high aspect piece of granite has the same potential qualities, but with much better chances of them occurring. When, I put that piece in, the computer picks (1.5%) Blockhead from the list. Now, when I hit "OK" and craft the item, I have a 1 in 200 chance of doing extra damage, a 1 in 67 chance of getting a toughness bonus and intelligence penalty, and a 1 in 13333 chance of both.
Now, a couple of possibilities here. One is that the character's material skill could influence the chance of a quality sticking to an item. In the above example, a granite master might have his percentages of producing a magic item increase by fivefold, whereas a novice might have half the normal chance.
Another factor would be the slot machine concept where if you put two identical items into aspect slots and the computer chose the same quality from each, both would have a higher chance to "proc". So, in the above example, let's say both granite pieces landed on Stalagmite. Because there were two of the same quality, the chance of each sticking would be doubled.
Another possibility would be that each time the computer rolled a different quality, the crafter would have a chance (based on his material skill?) to "learn" that quality. Once a quality is learned, the crafter could choose it in the crafting screen and it would have a higher chance of appearing in the slot machine. This would allow a crafter to influence what kind of stats appeared and if the matching qualities == higher chance idea was in place, would increase the chance of a magic item being produced.
Keep in mind, the numbers are totally out of thin air and a lot of tweaking would have to be done to get something that keeps magic items rare, but makes them worthwhile to pursue.
Now, here is the final piece of wackiness. Once a magic item was produced by using the aspect slots, the crafter would have the option of purchasing a blueprint for that item. If he held the blueprint and used the same materials in the same slots, he would be certain of producing the same magical item with the same qualities -- no random roll involved. Once a crafter has "discovered" a magic item, he can purchase the means to replicate it over and over.
So, going back to my example, let's say I get fortunate and create an iron dagger with the Stalagmite quality. I now have the option to purchase the blueprint for an iron dagger with bonus earth damage. When I craft again, I can start from scratch, or use the plan, put in 2 iron for the blade, leather for the handle and 2 pieces of granite for the aspect slots and I am guaranteed to create another iron dagger with +1 earth damage.
Now, the catch is that there are a limited number of each unique blueprint (in my case it is iron dagger with +1 earth damage) on any given server... let's say 100. So, the first 100 crafters to discover this item (and have the cash to lay down for the plan) get the means to replicate it. They now have a unique item to sell on the market. Everyone else has to stumble upon that particular item, and will never be able to reliably create it. Blueprints are items like any other and could be traded, sold, or stolen (in games where that is possible).
That's all for now. If there is any interest, we can expand on this idea a little more and talk about teaching, or copying blueprints, or even reverse-engineering blueprints from items. In the meantime, post a comment and tell me what you think.