Systems where attributes are separate from skills and super hard to raise are fucking dumb

[V_K]

Yeah, I also liked that system better, even though it did invite a ton of save-scumming. :D

 

[Damned Registrations]

Realism or whatever doesn't even matter to begin with. The entire point of having attributes be discrete from skills like this is to make for more interesting gameplay. If your stats were as easily changed as skills, then the solution to optimizing your character becomes far far more simple. You can easily calculate the value of a stat by comparing it to the number of skills it will improve, and conclude that a point of say, dex, is worth 5 times as much as a point of move silently because you have 2 other useful dex based skills and some combat benefits.

This shit doesn't apply when the stats can't be changed after creation, because the value of +1 to your dex skills can't be directly compared to the value of extra hitpoints or skillpoints per level and so forth.

That's why in many of these system attributes while raised in a similar way to skills, also cost more to raise in order to compensate for that. And if they costed around the same then raising dexterity wouldn't be nearly as helpful as sneaking as raising sneaking

Sure, but that's still a minor change to the equation. The cost was going to become higher anyways as you raise one or the other really high. So it doesn't really matter if raising dex to 6 costs twice as much as raising sneak to 6- it's still an incredibly simple equation with right and wrong answers. Good systems make such decisions more difficult by making the benefits less comparable. Take something like Crawl for example- you can raise attributes and skills in it, but attributes are linked much more to your race and you've less control over them. And they don't do comparable things much of the time- for example, intelligence has a high effect on spell hunger, while the individual skill schools do not. So if you have to choose between raising int or a spell school for some reason, the decision is far more interesting than one of them giving X amount of spell power and success and the other giving Y amount. And because attributes and skills are generally changed in very different ways, you generally don't have to make the choice between say, raising attack power via weapon skill or attributes, because your choices are instead divided into the two categories- you can raise weapon or magic skills, and you can raise str or dex or int. But you generally can't trade one for the other to keep them at the most efficient balance. So you end up with interesting choices between getting something benefitting your specialty more (like dex on a rogue) or something with better impact in a lesser sphere of influence, because of diminishing returns (like the effect of int on your minor spellcasting abilities.)

 

[Maxwell's Demon]

It's a bit of a cop-out to simply say 'it's an abstraction' when developers know full well that the abstraction is at odds with intuition. Not saying that abstractions are bad - again, I broadly agree with your comment. But unlike chess, it does matter to a crpg that an abstraction clashes badly with the way in which the mechanic is presented.

It's silly to talk about the intuitiveness of the system when even something as simple as Newtonian mechanics runs so against one's intuition. How would someone simply guess that a baseball follows a quadratic equation when tossed without first watching a baseball in the air?

And let's face it, no one can measure strength in totality without first running an experiment on each muscle group, then comparing each result to an average score, and then finally performing some kind of data analysis. Believing that strength can be quantified in one number is about as absurd as thinking that an IQ score measures mental acuity in any significant way. If you're not autistic, your intuition should scream at you saying that a single integer value lacks the nuance needed to describe something as complex as a person's strength.

 

[Shin]

And let's face it, no one can measure strength in totality without first running an experiment on each muscle group, then comparing each result to an average score, and then finally performing some kind of data analysis. Believing that strength can be quantified in one number is about as absurd as thinking that an IQ score measures mental acuity in any significant way. If you're not autistic, your intuition should scream at you saying that a single integer value lacks the nuance needed to describe something as complex as a person's strength.

Strongman lifts can be used a benchmark for how 'strong' someone generally is, not saying it is exactly super scientific or anything, but there's a reason some people are considered to be the 'strongest' (physical) man, even if they do retarded shit like pulling tires and trucks and whatever. Just like there certainly is a difference between someone who takes a (serious) IQ test and gets a score of 57 vs someone who gets a score of 120. Just like having a (very extended) trivia quiz type of shit to measure your wisdom, or trying to pick up members of the opposite sex or try to sell 2nd hand cars to measure your CHA or whatever. It's not science, sure, but these things can give you a general sense of how someone acts (or 'is' if you would keep doing those fucking tests over and over).

but I'm drunk at 7:47 on a tuesday in the morning so wtf do I know

 

[Maxwell's Demon]

Strongman lifts can be used a benchmark for how 'strong' someone generally is, not saying it is exactly super scientific or anything, but there's a reason some people are considered to be the 'strongest' (physical) man, even if they do retarded shit like pulling tires and trucks and whatever. Just like there certainly is a difference between someone who takes a (serious) IQ test and gets a score of 57 vs someone who gets a score of 120. Just like having a (very extended) trivia quiz type of shit to measure your wisdom, or trying to pick up members of the opposite sex or try to sell 2nd hand cars to measure your CHA or whatever. It's not science, sure, but these things can give you a general sense of how someone acts (or 'is' if you would keep doing those fucking tests over and over).

but I'm drunk at 7:47 on a tuesday in the morning so wtf do I know

The difference between a 120 IQ versus a 57 IQ is four days without sleep. Despite claims otherwise, one cannot generalize from a benchmark. A benchmark only shows how good someone is at that specific task, at that specific time, under those specific circumstances. This is not so much a science thing as a basic reasoning thing.