That is a problem of poor scaling (or rather, no scaling as leveling exists in modern games, where HPs start at 20 and end at 2000). If system is carefully topped after particular levels, and everything is scaled to a medium, the world design would not be harmed, characters will still feel growth, yet even weak creatures, properly used in encounters, would remain a threat.
I'm not sure I read it correctly - are you seriously proposing level scaling as good, fairly universal solution to cRPGs' woes?
tl;dr version: I don't think that there is a single right way of making a RPG game.
That still doesn't explain why including kill XP is one of the desirable ones.
And no, I wouldn't say it worked particularly well, not after cleaning some wilderness area of bears with a low level druid, just because they were several hundred XP a pop, so it was really making a lot of difference in a really short time.
By this logic Bethseda's games are mechanically watertight masterpieces too.
Meanwhile, simplicity and elegance alone is perfectly sufficient reason for dropping kill XP.
At least with my pet use based systems you have an excuse of simulationism for creating and trying to work out a convoluted mess, plus they really are the only sane way to have character growth in games focused around derping in wilderness with multiple possible playstyles and builds.
With XP you don't have such excuse so a simpler, cleaner system that effectively does the same job with much less work and no idiotic loopholes to be patched should be automatically favored.
It is up to the designer how he rewards (or doesn't reward) XP for killing a quest giver. You are acting like it is really complicated and unsolvable problem, when it isn't.
No, I'm acting like it shouldn't even *be* a problem (as in something the dev has to consciously handle) with sane progression mechanics.
It is different, because it only applies to quest givers and only after you help or spare them.
Why?
Because that is what a DM would do. I wouldn't reward a player for XP double dipping.
It's not double dipping. One chunk of XP is for fighting a (potentially challenging) character. The other one is for doing quest(s) which might not even involve killing anyone.
This was marketed to fans as a return to IE gaming, but instead such fans are being called "grognards" and there doesn't even seem to be an attempt at replicating the mechanics that made those games fun to play.
Put a clock near your computer if you want to see numbers go up as you kill shit.
Because that's what XP mechanics was in IE games most of the time - inconsequential numbers going up.
The level ups were few and far in between.
Who the hell are these perverse designers to decide what players should or shouldn't do?
Precisely. So why make a system that specifically rewards narrow group of playstyles?
WRONG! Combat XP provides an alternative means of developing your character aside from doing tasks that Sawyer wants you to be doing. It is not about incentives, it is about actually having the choice to PLAY not DO.
Yes, it provides alternative in form of single playstyle not really supported by content and with at best tenuous support from any sort of logic, at the expense of multiple playstyles that both make sense and have content created for them.
What's next, complaining that you can't take over a kingdom and command vast armies instead of derping around as murderhobo?
Maybe being able to spam daggers at the forge in Skyrim was also good mechanics because, hey, alternative! Those are good, no?
Besides, can't you play by doing?
(hurr...)
You are still completely missing the point. It isn't about being able to kill or not. It is about being able to progress by doing what you enjoy doing as opposed to what the designer would like to force you to do.
Like killing shit?
What if I'd rather progress by ghosting around shit?
Thats a problem that comes with finite exp.
Meanwhile the ability to go to arbitrary level by grinding the same shit over and over is a problem that comes with infinite XP.
If you need to do A, but B stands in your way, doing A implies that you must have overcome B in some way, therefore rewarding A also rewards B. Simple.
Wouldn't have a problem with that (group B being positioned in such a way that you have to go through it to reach your goal) but what if say you achieve pyrric victory over group B, don't get reward for it( whether in loot or XP), reload and find out you can just circle around group B to your actual goal (group A). In that case what would be the purpose of group B? To showcase how dangerous wilderness is (immersion)?
If you can get around B, and don't have any reason to fight it (loot, quest, rep, influencing some sort of dynamical mechanics by changing world state), why fight it?
Because you failed to scout out the alternative route? Because you were too stupid to think strategically? Why should failure or stupidity be specifically rewarded?
Because you wanted to burn some supplies getting in a big fight for lulz or to prove yourself? You have, you've got your reward, why do you want an extra for it?
It almost mimics the logic of modern 'tards who refuse to play old games because they have no achievements.
Do you really need something resembling achievements keeping your dopamine pumping?
I think it is. If it was more of a TES style game, I would disagree.
I've been meaning to ask you, how would you go about designing TES style system (improve skills by use) without it being open to exploits (hopping all the time, letting rat hit you for an hour to build up your block/armor skill etc.)?
It would take a longish answer because use-based systems are a mess, that only has simulationism and ability to work without explicit support by content, as their saving graces (that's not particularly damning, actually, because those are huge upsides, but would probably be bad enough when designing a quest centric game like IE ones).
I don't really want to derail the thread (again), and I have already went over this more than once but I'd:
- Make the system involve skill based success failure for *all* skills (can't have meaningful fail state? can't be a skill, although it might still be a stat - or part of it) and have probabilities just cut off at 1 and 0.
- Make the skill gain (or probability of skill gain, as I'd probably make it nondeterministic) depend on unlikelihood of outcome - (1 - p) as multiplier. That way skill gain would be tied to difficulty and guaranteed success or failure would never increase skill.*
- Preferably make information about skill gains delayed to nearest rest or similar - like in Daggerfall.
- Make upkeep cost in some way to make derping around unproductively non-viable.
- Make all skills that can be increased contribute to character's growth (to eliminate ability to grind non-class skills like in MW and OB).
- Make skill increase 'cost' independent of individual skill level, but instead a function of sum of all skills - meaning that each subsequent skill increase would get harder independent on skill being increased and its level. This would remove the comparative advantage of increasing low level skills and would introduce opportunity cost to developing your character.
- Preferably make a hierarchical structure of numerous skills and sub-skills, and let them "trickle down"
- Either make those skills shallow or try to ensure infinite, but capped growth (asymptotic) that would nevertheless remain relevant in skill contests - for example n+1 sword skill would give you a good chance at besting n sword skill guy in sword combat despite having negligible mechanical effect otherwise (because asympthotic).**
*) This is probably the trickiest part. Straightforward enough with traditional RPG rolls (because you could just take probability) but with more detailed simulationist mechanics it would require estimates. For example let's say you're shooting a ranged weapon. The game would have to take your dispersion cone, select all targets within it (preferably try to select hostiles and environmental triggers only, if no hostiles do so with neutrals, otherwise take all potential targets) sum up their hitbox sizes at given distance and divide by dispersion size. This would sadly drop the lower cutoff, but it would keep the upper one.
**) For example, take our sword skill. Being asympthotic, its effects on hitting, swing speed, penetration % and whatever parameters it could influence would effectively cap off relatively early, but whenever for example parrying or locking blades or performing any blade vs blade maneuver with another sword user it would compare the scores directly (with some influence from other stats, but let's not overcomplicate now) - large difference would guarantee success/failure while small difference would sharply alter probability of it so while at level 1000 you wouldn't be perceptibly better at swordswinging in terms of speed, precision or whatever, than at level, say, 50, a level 1000 swordsman would probably end you in seconds if he were to cross blades with you at level 50. It may sound like terrible waste of points but it depends on how often will you cross blades with someone - it might end up being primary reason to pump your weapon skill if mostly fighting other human(oid)s. For different weapon types there might be a table of pairwise effective level conversions, or you might default to more general skill in hierarchy.
If you want to pursue the topic just quote the relevant part of my post (starting from last quote - to keep the ideas in one place) in the relevant thread, preferably this one:
http://www.rpgcodex.net/forums/index.php?threads/oiling-the-cogs-of-use-based-systems.34792/
Other threads of interest on the topics from use based to PoE current system:
http://www.rpgcodex.net/forums/inde...fic-knowledge-based-system.35973/#post-886816
http://www.rpgcodex.net/forums/inde...by-use-skill-system.83200/page-3#post-2694004
http://www.rpgcodex.net/forums/inde...el-less-crpg-system.83249/page-2#post-2693329
http://www.rpgcodex.net/forums/inde...still-proper-design.83251/page-5#post-2693785
http://www.rpgcodex.net/forums/inde...-rewards-for-quests.65766/page-6#post-1931133