Yesterday, in a rare burst of eloquence, I said "The realism dials of GURPS are adjustable; no sense complaining about the dial settings instead of just changing them."
I feel it's a mistake to try to ram a square peg into a round hole when you've got a square hole nearby. The martial artist from yesterday's post is a good example - the source material (D&D monks, every unarmed striker in a video game, countless kung fu movies that inspired both of those) says that people punch effectively against monsters. You could solve this through lots of fiddly rules adjustments. You could also just say "this is how it works in this game world, in this campaign."
The second isn't a cop out. It's a fundamental GURPS assumption that you will do this.
Don't believe me?
Check out Cinematic Combat.
Check out rules like "Bulletproof Nudity" and "Flesh Wounds."
Check out the "Harsh Realism" boxes in GURPS Martial Arts. Yes, that's a supplement, but it was being written when Basic Set was in production.
Heck, check out the three different detail levels of combat.
GURPS starts with a fairly realistic approach. Generous, yes. Punches to the skull don't hurt your hand, you get unlimited Dodges, you can suck up a lot of damage just short of crippling without problems. But not so generous it strains the bounds of reality or versimilitude. It then lets you set the dials to much harsher levels (punching the skull hurts, cumulative Dodge penalties, partial crippling) or much more generous ones (swords can't parry punches, access to Enhanced Dodge and even Cosmic defenses, Flesh Wound rules).
There are dials for silly vs. serious.
There are dials for realism.
There are dials for complexity.
There are dials for detail.
That's before you even adjust the rules, or adjust the point totals of the PCs/NPCs to figure out how they stack up against the world (a 500 point commando in a gritty harsh reality game is going to still be a total badass, maybe even more so, while a 50 point farmer in a wuxia game is going to be dead, rules to bend reality and run on treetops or not.)
I'm a big fan of spinning these dials around.
If possible, I prefer to solve problems without changing any rules - if an solution already exists, I do it that way. (Ex. the problem is people need X skill to do feat Y reliably, the solution is to get people who need to do Y feat X skill. Like perks to give limited access to cinematic skills or to access a genre rule just for one character - a solution exists already.)
If not, I prefer to change the genre settings - spinning the dials. (Ex. The problem is people in the source material do A without consequence B, so I just say consequence B doesn't happen. Like punching swordsmen without getting cut.)
If that's inappropriate, I'll jump through hoops to get the job done in a different way. (Ex. A specific character needs to do something that's neither amendable to a genre dial turn, nor an existing rule. This is how some of my new rulings and then rules are born, like turning into an Evil Cleric.)
I understand the approach that basically says "There doesn't seem to be a dial setting that matches the feel I want - help!" But I'm a little confused the approach that basically says "I turned the settings to X instead of Y - help me get to Y through new rules." It seems like it's a lot of effort spent doing something the various rules settings are there for in the first place. This is why my solutions for problems can sometimes seem very flip - "just don't do that." But it's a working philosophy of gaming. Why do hard work on a potentially finicky rule with potentially unforeseen consequences when something else is sitting right there and does it already, more easily?
This isn't to say doing it the hard way is wrong. But for my gaming, I'll try to stick with the easy way.