Crafting shouldn't be a scalar. I hate how you have to make X swords in order to make an axe, and then X axes to be able to make a big war mace. It's stupid; I think if your character is strong enough and smart enough, he should be able to read some instructions and make a fucking hammer. Of course, practice would play a part in this, and by blacksmithing, a user would gain some strength and most likely some intelligence too. For other things, for example alchemy, a user would gain a lot of intelligence due to the extensive knowledge required to brew potions correctly, and the chance to succeed in making said potion would be modified by the user's intelligence attribute. (just like in D&D, which actually did things right for an RPG)
I never said anything about requiring 10 swords to make an axe... I'm not talking about components yet, that's a whole other story. We're talking about skill requirements.
I said that my skill in smiting would likely have to be measured in levels too. You shouldn't expect someone to be able to make a "Sword of the Champions + 9001 dmg" when he's never touched a smiting hammer or seen a blacksmith in his life, just because he happens to have a strong arm. It takes more than strength to do something, it takes skill. Skill is, in the end, acquired through experience.
It's not even a balancing thing because anyone skilled could still give the ubersword to a noob. Let me give a very easy example: bicycling:
You can be as agile as you want, if you've never been on a bike before you'll undoubtedly have trouble riding it. After a while, you'll get used to bicycling, and you'll be able to do it on yourself, but even the best bicyclist has once in his life not been able to ride a bike. Not because he wasn't agile or strong enough, but simply because he didn't have the technique to keep his balance on a 2-wheel thing. Granted, everyone here will think the technique is easy, but it's still something
everyone needs to learn.
With something like writing or smiting it gets a lot more complex because, really, you need to have a lot of experience in writing before you can be called a poet. And not all intelligent people are poets. At the contrary. A "tradeskill" is mainly measured by talents and experience, not by attributes.
That, ofcourse, doesn't mean attributes can't contribute. To extend the bike example: it's not because you can drive a bike that you can win the Tour de France. It takes strength and stamina on top of much biking experience to win a tour.
In real-life, things are even more complex. You can't really "scale" skills. You can only say "I'm better at it". You can't say Lance Armstrong has 1231 skill in biking while Oscar Pereiro only has 1228 in biking. But let's face it: it's a game and you got to use some sort of measuring system. The discussion wasn't really about how to measure it, but what to base the measurement on. Void suggested to measure it through attributes, and I disagree for the reasons I've summed up.
They both do play an important role in skills. However, I think:
- learning skills heavily depends on attributes.
- using skills heavily depends on your "skill level".
It's a fact that intelligent people simply understand mathematics faster. It's a fact that strong and agile people have more chance to become a GOOD bicyclist. But it's also a fact that you don't have to be strong and agile to bicycle. You don't have to be intelligent to understand mathematics, as long as you spend enough time in doing so to get experienced in it. Being intelligent just helps.
My overall suggestion:
Skills should keep track of a separate experience rate. This rate is influenced by the attributes (the more agility the faster you learn how to bicycle) but can only be increased by using the skill. You only get better at doing something by doing it, and you get good at it faster if you have good attributes.
"What" you can do with the skill should be mainly dependant on the level of the skill. I'd say the chance you successfully make something should be 95% dependant on the skill level. That means you're still able to smite a sword of leetness at a low smiting skill, but you're gonna need a LOT of luck to do so. But as long as you have ingredients available you can keep attempting at making the sword, and each failed attempt would increase your skill experience. If you're strong and agile, this would mean you'll learn faster from your mistakes.
Something on ingredients: it might only require a hot fire, a hammer, some iron and an enchanted gem to make a sword of leetness. If you fail at making it, you've lost the ingredients but you increase your smiting skills. Therefor, if you have 1 hammer, 10 iron parts and 10 enchanted gems, and you're low skilled but have high strength, you'll probably have a higher skill and therefor a higher chance at success by the time you're using the 10th piece of iron...
EDIT: although in this specific example it's probably more logical to split the process in:
- Create a sword
Requires:
- Hammer (static)
- Hot Fire (static)
- Iron (variable)
Skill: Blacksmiting [Strength]
Result: Iron Longsword
- Create enchanted sword
Requires:
- Iron Longsword (variable)
- Enchanted Gem (static)
Skill: Enchanting [Intelligence]
Result: Sword of the Champions
Variable requirements would be consumed upon using the skill (thus: smiting a sword will use only iron, but require a hammer and fire too. If it fails, the iron is no longer useable)
I've split this example in 2 parts because enchanting a sword to a magical sword of champions wouldn't make much sense if it was dependant on your smiting skill