Some proposals to improve character XP allocation (combat skill training)

I thought I'd share some thoughts about improving the way the earned character XP is spent. You're welcome to share your own thoughts and ideas, please keep it civil, constructive and on topic.

Below I will list some proposals to improve the current situation, expecially focusing on combat related skills. Even though some things certainly depend on each other, they are presented here just as a bunch of loose ideas to encourage discussion; pick the ones you like, scrap the ones you don't, add your own.

Most folks know the current implementation well, so I'm putting these background details into a spoiler for the sake of readability.

Spoiler: Short description of the current implementation

The current implementation to my understanding is very simple; in short, the character performs some action that earns her XP, which in turn is put into one of the XP pools (adventurer or producer). Independently of that, character actions may also result in character skills drawing XP out of the XP pool relevant for each skill.

Some characteristics of the current implementation:

- The XP requirements to train skills rise exponentially with increasing skill levels.

- The amount of XP drawn from the pool depends at least 1) on the size of the pool and 2) the current level of the skill.

- The amount of XP drawn from the pool does NOT depend on the type of action; for example it does not matter whether:
--> one is attacking an easy mob already killed hundreds times;
--> one is attacking a gustball or a training dummy;
--> one is attacking a difficult monster never encountered before;
The end result, in regard to skills trained, is exactly the same.

The above also means that it is possible to spend one's (already earned) XP on a single skill by spamming that skill repeatedly. This is especially relevant with attack skills that can be trained by repeatedly attacking a gustball or a training dummy, all the way up to the highest levels using nothing but training tools. While this can be useful for example in the testing phase of the game, it is unrealistic to the point of breaking immersion, and can have some limited abuse potential.

Without further ado, here are my own proposals.

1. Make attack skill improvements dependent on the target.
It should matter whether or not one is targeting a training dummy or an obsidian bear. Both have the potential to improve attack skills, but training with a "real", difficult target should train the skills more than attacking an inanimate construct, or easy opponents. This is just common sense. It must be noted that one can certainly train effectively on the highest levels using training tools (no need to put a hard limit on certain fixed levels for example), but they should not be equivalent to the "real thing".

2. Make attack skill improvements dependent on the familiarity to the target.
Killed hundreds of wolves? One should still earn considerable XP from killing the next wolf, but maybe not quite as much as much improvement in regard to skill improvement. Nothing too dramatic though; the downward curve should be relatively flat, for example so that the first elder wolf, presuming a similar XP pool size, should train the skill by amount X, and the 1000th elder wolf maybe 70-80% of that amount. The familiarity should probably decay slowly over time, just like the skills do.

3. Make inanimate training gear that have a certain cost and durability.
As said before, artificial training gear can be a legitimate way to train one's combat skills. Make training gear -- dummies and other targets that can be hit and/or shot at, make them placeable as functional decorations. They provide certain, relatively constant level of skill improvement, but they never provide any XP. They have a durability so they break over time, and they have a non-trivial cost to craft/purchase. They make great, roleplay supporting decorations, for example I'd just love to have an archery target in by basement.

4. Make a small portion of the just-earned XP from kills go directly to the recently used skills.
This one is presented as a brainstorming-level idea only; I have not thought of that very thoroughly. The premise is that it is not really possible to totally "turn off" the learning of some skills. So regardless of the training status of the skill, if you actually use it, it will get trained to some extent regardless (for example if you kill a wolf, the skill(s) that were used recently while combating that wolf would get exercised a little bit, directly from the XP reward from killing that wolf, regardless of the training status of those skills. Now if you have deliberately set your mind to actually improve certain skill (in game terms, set it to "train"), the improvements would be much greater of course, but it feels unrealistic to set any skill learning rate at exactly zero. The motivation for this proposal is also that it might to some extent alleviate some concerns relating to the micro-management of skills.


Some aspects of these proposals are perhaps too "organic" (complex systems on top of complex systems) to the mindset of software developers, who generally prefer having everything controlled by a single formula in a deterministic way. I gladly admit that, but present the ideas as such regardless, in a kind of brainstorming spirit -- often excellent ideas spring out as modifications to even wacky-sounding ones.

Thoughts, comments? Yes please.

Edit - added numbering

 

The first three are all related to a Viable target and killing the mindless trainng on Gust balls, and other no viable targets - yes , YES, Y E S.

They need to really just kill these types of training.

As for killing the same thing over and over - - sorry I cannot get on board with this, since we have such limited places and critters to fight this can only lead to killing training.

The problem with the last is if I have a skill at GM and do not want to use any more points except to maintain it that is MY choice. Their future changes to combat macroing are going along these lines, once they hit we will see how badly they attack skill management. I only see bad things from that concept, but will wait.

 

I'm not sure what you mean by non-viable targets here, but as you elaborate it in another thread like so:

I will therefore presume that by non-viable targets you mean inanimate targets that do not fight back.

I would rather prefer calling them just inanimate targets, because their non-viability, even way past level 50, is far from clear.

As a real life example, would anyone in his/her right mind tell a world Top 50 boxer (or the practitioner of just about any similar art) that he should abandon all inanimate targets from his training regime because they are not "viable"? Just because the target does not fight back does not mean that practising with it is mindless. Far from it. All it can mean, in the context of the game, is that it does not provide any gains to the character's exp pool. Otherwise I'd consider it quite viable.

Spoiler: Present-day real-life gustball training

 

After carefully considering the offered proposals, I disagree with all of them, and don't want to see any of them in the game.
Why? They're all less fun and more punitive than what we have now.

It's going to be hard enough to attract new players, as it is, with the amount of punitive mechanics in SotA. Personally, I'm not in favor of adding any more.

 

Limit skill gain to Target Level +20. Meaning, even if you are Adventurer Level 100, if you want to train up your Skill Level 60 Polearms, you have to fight creatures that are at least level 41+.

Make Gust Balls and Training Dummies Level 1.

 

Thank you for the careful consideration. Would you care to elaborate why you find them punitive? I can not really see punitiveness in any of them, actually I though of this aspect pretty carefully myself. More specifically, in the order of OP appearance (I'll be adding numbers to the OP as well).

1. This should actually be helpful rather than punitive - as it would be easier to level a skill by picking battles that are more challenging (considering your current level, whatever it is). Less grinding, more interesting battles.
2. This is meant to encourage variety, by discouraging grind-farming the one and the same thing over and over and over. Admittedly it uses sticks rather than carrots as incentive, but the proposed amount is very modest: imagine the same thing applied to XP gain; if your 1st elder wolf gave you 1150xp, and your 1000th elder wolf "only" about 900xp, would it sound so bad? The same kind of ratio only apply to the speed of leveling , nothing else. I believe it would be almost not noticeable except by hardcore min-maxers and power-levelers.
   
3. This one is an attempt to have inanimate training objects in the game, useful without becoming objects of abuse and macroing, as something supporting roleplay and immersion instead of breaking it. An alternative is not to have them at all (which would feel more than a bit weird, considering that they are widely used in the game as scene decorations).
   
4. Yes, this one might indeed be regarded as slightly punitive. Possibly there are better alternatives. As said, this was included in brainstorming spirit. Again the effect would be very modest, maybe 10-20% of the total XP rewarded would be allocated this way. I did not give any explicit figures in the OP, perhaps I should have, given that folks are used to rather hard-handed adjustments, so "a little bit" might be interpreted very differently than it would in other contexts

I wonder, where would one find enough of those level 80+ creatures for the GMs?

This would in practice mean that one could train, using inanimate training equipment, about 0.04% of the XP required for GM level. Or, less than 0.3% of level 80 XP total requirements. Or, about 15% level 40 (which is regarded the level of "basic proficiency") total XP requirements. With such low limits there would be no reason to have inanimate training equipment even implemented in the game.

 

Really? A boxer hitting a speed bag ??? not even close and you are reaching for the moon. A fencer will never train on a target who does not fight back, there is NO point.

This is not like shooting where basic fundamentals take the major factor these are complex moves that can be countered . . .

So partice at that level while keep the level of training stable, even a boxer SPARES to get better. A speed ball is for BASIC training skills.

 

Easy fix. Scale the margin to be larger as the Target Level goes up. I just threw out some fast numbers. It's still a super easy algebra problem.

Scaling based off XP needed to level is not the point, unless you're just trying make leveling skills easier (like everyone whacking gustballs). Skill Level is what provides the benefit in combat. Who cares if it take a billion quadrillion XP to get to Skill Level XYZ, the inanimate objects should not provide "basic proficiency" (Read: Skill Level 40), At all.

 

Inanimate objects are used quite a lot in various fighting arts, even at highest levels. Please think for a moment what kind of proficiency your proposed "level 20" translates to (level 40 is generally regarded as "basic proficiency"). The counterexamples look a little bit too much like straw men to me - for example I have said nothing about fencers (and even they practice moves without sparring partners). And yes, various kinds of hitting bags do belong to boxers' training regime at all levels, I have never claimed that is all there is, just that they certainly are part of the regime. Please try suggesting any boxer that from now on his only training with gloves on can happen with a live sparring partner, hitting back. Because if I understood your point about inanimate training objects correctly, that is basically what you are proposing for the game, correct?