Putting the 'role' back in role-playing games since 2002.
Donate to Codex
Good Old Games
  • Welcome to rpgcodex.net, a site dedicated to discussing computer based role-playing games in a free and open fashion. We're less strict than other forums, but please refer to the rules.

    "This message is awaiting moderator approval": All new users must pass through our moderation queue before they will be able to post normally. Until your account has "passed" your posts will only be visible to yourself (and moderators) until they are approved. Give us a week to get around to approving / deleting / ignoring your mundane opinion on crap before hassling us about it. Once you have passed the moderation period (think of it as a test), you will be able to post normally, just like all the other retards.

An XP-less, class-less and level-less CRPG system

baturinsky

Arcane
Joined
Apr 21, 2013
Messages
5,608
Location
Russia
Limit the number of buffs? I never really liked the idea of them anyway. One or two good ones, sure, but I shouldn't have a checklist of shit I have to do before a fight.
Yes, D&D failed spectacularly here. M&M did right rolling a lot of buffs into one spell (Day of..., Hour off...), but then added wells with huge buff, that you will want to visit each morning.

I think best way to do buffs is sustains - you can have them for as long as you want, but they have some penalty attached, like reduced mana cap.
 

waywardOne

Arcane
Joined
Aug 28, 2010
Messages
2,318
Skills could have synergy. If you have 100 points to spend, you could use them in a 1:1 ratio if all the skills are related, so 20 points in 5 very similar skills like 1-handed melee weapons. But if you varied your skills more, say diplomacy, stealth, and climbing -- your ratio would drop based on how dissimilar the overall skill categories were. Increasing skill X would make all skills Y more expensive.
 

MicoSelva

backlog digger
Patron
Joined
Sep 10, 2010
Messages
7,512
Location
Arx
Codex 2012 Codex 2013 Codex 2014 PC RPG Website of the Year, 2015 Codex 2016 - The Age of Grimoire Make the Codex Great Again! Grab the Codex by the pussy Insert Title Here RPG Wokedex Strap Yourselves In Codex Year of the Donut Codex+ Now Streaming! Enjoy the Revolution! Another revolution around the sun that is. Serpent in the Staglands Divinity: Original Sin Project: Eternity Torment: Tides of Numenera Divinity: Original Sin 2 Bubbles In Memoria A Beautifully Desolate Campaign Pillars of Eternity 2: Deadfire Pathfinder: Kingmaker Steve gets a Kidney but I don't even get a tag. Pathfinder: Wrath I helped put crap in Monomyth
I don't really know how to avoid the problem of people ignoring the degradation because they for example want their mage/cleric to spam buffs all the time.
One possible way is to make buffs sustained abilities, with the total number limited by spellcaster's Constitution and/or Willpower or some equivalent.
 

baturinsky

Arcane
Joined
Apr 21, 2013
Messages
5,608
Location
Russia
Even for hitpoints there would be a counter that increases every time the character when he gets damaged, i.e. "successfully avoids death".
About this thing - is it actually good idea to make player get damaged to raise hp? It is usually not working well - better players get hit less often, and are "punished" with less hp. Same goes for healing magic.
 
Joined
Sep 8, 2008
Messages
11,313
Location
SPAAAAAAAAAACE...
Project: Eternity
Here's that test program I was working on: https://dl.dropboxusercontent.com/u/29099307/rpgcodex/rpgtest.7z
Sorry, it's 8 MB :eek: If there's a way to make compiled Python smaller then I don't know it yet. Just unpack and start the .exe in the directory.

rpgtest.png


What all this means:
* STR, DEX, HEALTH are strength, dexterity and hit points
* skill_clubs is the skill that governs using clubs, obviously :M
* The values ending in _prog means the current state of the progress counter for that skill
* The values ending in _proglim means the value that the corresponding progress counter must reach for the skill to increase by 1.
* "Real" skills as well as abilities use the same algorithm for progressing, just abilities start at a higher value while skills start at 0.

Ignoring the mechanics of the attack roll for the moment, basically every time the attacker succeeds his STR and DEX progress are increased by a small fractional value that depends on the weapon (strfactor and dexfactor here). If the attacker does not succeed, the defender's DEX progress is incremented by a small fraction instead.
 
Joined
Sep 8, 2008
Messages
11,313
Location
SPAAAAAAAAAACE...
Project: Eternity
Even for hitpoints there would be a counter that increases every time the character when he gets damaged, i.e. "successfully avoids death".
About this thing - is it actually good idea to make player get damaged to raise hp? It is usually not working well - better players get hit less often, and are "punished" with less hp. Same goes for healing magic.

I'm having second thoughts about this too, hence I didn't put anything to deal with HP in the test program yet. This needs a more "positive" approach, I agree. Maybe something connected to melee attacks.
 

baturinsky

Arcane
Joined
Apr 21, 2013
Messages
5,608
Location
Russia
I think CoffeeScript+HTMLl is perfect for prototypes. Lightweight, portable, making gui is a breeze.
 
Joined
Apr 2, 2010
Messages
7,428
Location
Villainville
MCA
What to prevent the sort of tactics we use to abuse the learn-by-doing system we have done in morrowind?

Put a "per Critter" increase quota. You can raise your skill only so much by killing rats. Also set up to-hit and defense/evasion ratios per Critter. Eg. optimal skill level to kill a rat: %15. Highest you can raise your skill by killing rats: %15. While better than unconditional learn-by-doing model, even that is not perfect because it still doesn't distinguish between fighting a rat vs a swordsman. You don't fight a rat the way you right a warrior. Gothic 3 did a sensible thing by introducing Big Game skill to account for fighting against wildlife but it wasn't the perfect way.
 
Joined
Sep 8, 2008
Messages
11,313
Location
SPAAAAAAAAAACE...
Project: Eternity
I think I'd just have a counter for how many of a certain creature you can kill before they stop giving you progress. Like 20 rats you can train on, afterwards it's just ballast to fight them.
 

baturinsky

Arcane
Joined
Apr 21, 2013
Messages
5,608
Location
Russia
Skill degradation over time can probably do same indirectly. So, you can spend days smacking rats, and gaining some relevant training, if you want, but if you are already have many high skills, you will forget more, then you'll learn. So, you'd better smack something more interesting.

Another idea - have skill per enemy type, that give big bonus vs that type, and small, but cumulative bonus vs everything. So, it's more effective to fight lot of different things, to accumulate bigger bonus.
 

J1M

Arcane
Joined
May 14, 2008
Messages
14,731
And in a hand-made world with a fixed number of enemies and no wandering monsters, you're golden. But in a dynamically generated world (such as the mentioned Morrowind), by this system, when playing I am instead incentivized to grab up my trusty cursed sword of hit-nothingness and run around the map in a huge circle over and over, striking a single time at every creature that appears in order to get my miss experience and then running away.

Hell, I could stick a cursed sword in one hand and a real sword in the other, and get experience for both hitting and purposefully missing at the same time.

When I am being rewarded more for missing, it becomes in my best interest to ensure that I miss.
I'm assuming the xp-on-success system will have a sword skill cap for killing rats. Why are you assuming that there wouldn't be said cap in a xp-on-failure system?

XP-on-failure is one of the few ways I can think of to prevent a player from save-scumming lockpicking and similar activities.
 

Akratus

Self-loathing fascist drunken misogynist asshole
Patron
Joined
May 7, 2013
Messages
0
Location
The Netherlands
Codex 2016 - The Age of Grimoire Make the Codex Great Again! Grab the Codex by the pussy Insert Title Here Pillars of Eternity 2: Deadfire Steve gets a Kidney but I don't even get a tag.
Or just not have saves. Just profile slots that automatically update.

By the way, is anyone else bothered that a stat can increase your survivability independant of armor or skill? I mean, you might have good endurance in real life but it won't help you resist a fucking hit with a bastard sword. I would just like to see all characters have 100 hp (with debuffs as it gets lower?), which would be representative of the state of their body, with only equipment, magic and skills being able to block/evade damage.

Also, skill increase on use is not a fundamentally broken idea. You could very easily stop people from leveling with an exploit by decreasing xp gains exponentially the lower level your enemy is. Added onto that could be a decrease in xp gain as your skill level gets higher. Perhaps you could have skills go up to 200, with 100 being the cap for skill trainer and book level training.

XP on failure is a nice idea. Perhaps you could prevent exploitation by limiting it to one xp gain per unique action, or possibly use diminishing gains?
 

Norfleet

Moderator
Joined
Jun 3, 2005
Messages
12,250
I have a better question: Why is it that the character's stats need to increase at all? You are functionaly creating "XP" and "Levels", just hiding it under a different label and obscuring it from the player. This never works. And in the end, the outcome is the same: A total schlub suddenly becomes the greatest whatever on the planet over the course of a few weeks to months because the player is driving.

Why not just have the player be whatever he picks out at the start, and...that's it? He gets better equipment, more toys, but in the end, he's fundamentally himself. Whatever skills he has learned and honed to competency that defines his class will not undergo any monumental change. He's competent at what he does, and this isn't going to change in the month or so that your game spans. It's up to the player to learn how to use these abilties at appropriate times to achieve good results. The character himself is a competent warrior, rogue, wizard, or what-have-you, neither some kind of Grandmaster nor a totally inept schlub fit only for killing rats, and whatever happens, he's going to remain so.
 

Hellraiser

Arcane
Joined
Apr 22, 2007
Messages
11,741
Location
Danzig, Potato-Hitman Commonwealth
While the rate at which skills improve may be in some cases too fast given the timespan it is not completely unrealistic. If you spend a month whacking rats with a club you are going to improve. The only question is how good were you at that task before, because in general spending more time to improve a skill has diminishing returns the more you master it. Going from not knowing how to fire and aim a gun to being able to kill something with it is fairly easy, going from regional level shooting champ to Olympics level shooter is a lot harder.
 

odrzut

Arcane
Joined
Apr 30, 2011
Messages
1,082
Location
Poland
Why not just have the player be whatever he picks out at the start, and...that's it? He gets better equipment, more toys, but in the end, he's fundamentally himself. Whatever skills he has learned and honed to competency that defines his class will not undergo any monumental change.


Because leveling up mechanic is addictive and easy to implement, and game designers want their games to be addictive.
 

DraQ

Arcane
Joined
Oct 24, 2007
Messages
32,828
Location
Chrząszczyżewoszyce, powiat Łękołody
Progress would be made by "successful use" of abilities, skills and even the hit point max. Basically for each value in the character sheet there is a progress counter, which is incremented with every successful use.
Every?

That's widely open to grind of worst possible kind.
If stabbing any legit target increases your skill with dagger, then you can just run around stabbing mudcrabs until you achieve dagger mastery.

To make such system workable you'd have to concentrate only on those uses during which you learn anything - formalize it. And not just successful uses, there is something like learning from your own mistakes.

For abilities successful use would have to be tied to whatever actions involve that specific ability, so skill use would also increase the progress counter for the relevant ability a bit. But stuff like attacks would increase the STR and/or DEX counters too, magic use would increase the INT counter etc.
That effectively mirrors the tree-like sturcture I proposed once in another thread. It's a good idea, but would need more than that to stand.

This makes quest rewards a bit more complicated of course, since you can not just dump a bunch of nonexistant XP on the party.
Actually, this would make player's and designer's approach to quests more rational, since there would have to be some expectation of actual reward fitting the situation to motivate the player, not just doing everyone's shit because XPs.

What do you think?
Too bare-bones in its current form. It's effectively bethesdian system minus level ups but with all problems left intact.
 
Joined
Sep 8, 2008
Messages
11,313
Location
SPAAAAAAAAAACE...
Project: Eternity
Ultimately I guess that all training involves *use* of a skill foremost, with successes and failures both contributing to the process of achieving mastery. So, you would have to reward any use of a skill really. Let's say for the thought experiment that *any* attempt for a melee strike or the casting of a spell will increase progress for the relevant skills, no matter if it did any good, i.e. caused damage or healed wounds or whatever.

The fear that many here have voiced is that the player will simply stab a million mudcrabs or rats or whatever and after a month of in-game time emerge from the tavern cellar with a sword skill that equals the greatest masters of legend. Even if there is no unlimited supply of puny easily squashed enemies, this obviously shouldn't be a legitimate way to get that skill that high and the system definitely needs a way to counter that. If you only train on rats, you should not be able to train your skill past a certain point.

At the same time, I think "grinding" should definitely work if the player works against increasingly higher challenges. It may be annoying if players choose to play that way, but if someone really is willing to invest the time, then they are playing the game at least and that is what the designer wants more than anything else really. So, the big question is really how you reward the player in a fair and hopefully transparent way for seeking greater and greater challenges. Unfortunately I can also see a flip side of the mudcrab grinder, which is if someone finds an enemy that counts as a high challenge but that isn't actually very lethal and will in effect just stand there while the player hacks away at it. I suppose though that is very hard to avoid for all possible builds and all possible gear selections.

So, it seems to me there is no way around a challenge rating which will make fighting an enemy pointless (at least from a skill progress point of view, gaining gold etc. is a different matter). I was hoping to avoid that really, because it's one more thing to design, but I guess it is too important to leave out. I'm not even sure if they are usually automatically calculated or preset by hand or a mix of both?
 

Johannes

Arcane
Joined
Nov 20, 2010
Messages
10,669
Location
casting coach
While the rate at which skills improve may be in some cases too fast given the timespan it is not completely unrealistic. If you spend a month whacking rats with a club you are going to improve. The only question is how good were you at that task before, because in general spending more time to improve a skill has diminishing returns the more you master it. Going from not knowing how to fire and aim a gun to being able to kill something with it is fairly easy, going from regional level shooting champ to Olympics level shooter is a lot harder.
Another thing is, if whacking rats with a club prepares you for fighting people or bigger animals at all.
 

DraQ

Arcane
Joined
Oct 24, 2007
Messages
32,828
Location
Chrząszczyżewoszyce, powiat Łękołody
Ultimately I guess that all training involves *use* of a skill foremost, with successes and failures both contributing to the process of achieving mastery. So, you would have to reward any use of a skill really. Let's say for the thought experiment that *any* attempt for a melee strike or the casting of a spell will increase progress for the relevant skills, no matter if it did any good, i.e. caused damage or healed wounds or whatever.

The fear that many here have voiced is that the player will simply stab a million mudcrabs or rats or whatever and after a month of in-game time emerge from the tavern cellar with a sword skill that equals the greatest masters of legend. Even if there is no unlimited supply of puny easily squashed enemies, this obviously shouldn't be a legitimate way to get that skill that high and the system definitely needs a way to counter that. If you only train on rats, you should not be able to train your skill past a certain point.

At the same time, I think "grinding" should definitely work if the player works against increasingly higher challenges. It may be annoying if players choose to play that way, but if someone really is willing to invest the time, then they are playing the game at least and that is what the designer wants more than anything else really. So, the big question is really how you reward the player in a fair and hopefully transparent way for seeking greater and greater challenges. Unfortunately I can also see a flip side of the mudcrab grinder, which is if someone finds an enemy that counts as a high challenge but that isn't actually very lethal and will in effect just stand there while the player hacks away at it. I suppose though that is very hard to avoid for all possible builds and all possible gear selections.

So, it seems to me there is no way around a challenge rating which will make fighting an enemy pointless (at least from a skill progress point of view, gaining gold etc. is a different matter). I was hoping to avoid that really, because it's one more thing to design, but I guess it is too important to leave out. I'm not even sure if they are usually automatically calculated or preset by hand or a mix of both?
I think this might be relevant to your interests (crosspost):
A proposition of how I would make a use based system:

First, the character, as far as use based is concerned is built of skills. Character may (and should) have things like other stats - including basic attributes - the attributes may be tied to skills - they may be built up by skill increases and/or modify the increases, intelligence may limit the character's potential for advancing in multiple skills as was proposed in this thread, it's all details. What matters the most are skills themselves. Not even how those skills look like - they might just as well be linear scales as perk trees or something in between.

The main question is "what makes you learn something new when mastering a skill?"
The answer can't be just success or failure. If any of you (assuming we don't have brian surgeons here) attempted brain surgery on someone, they would surely fail. They would also not have learned anything. If any of you were to write a post on the 'dex they would certainly succeed. They would also not have learned anything in regards to their forum usage skill.

Noticed anything those two cases have in common?
Their outcome is deterministic.

Now, I think it's reasonable bet that learning something involves absorbing some new information and amount information is defined by information entropy.

Entropy is defined as -sum( p(i) log(p(i))) over all possible outcomes with number of possible outcomes as base of logarithm and p(i) being probability of i-th outcome - notice that since all p-s sum to 1 maximum of this function is when all probabilities are equal - the less you know about outcome, the more you learn from it.
Our logarithm here would have base 2 (success or failure).

Now, Shannon's entropy isn't the only function that conserves this valuable property. You might as well use 1/2 sum( p(i) p(~i)) - it also maxes out in the same place and also drops to zero if one of the probabilities does - for our purposes it's close enough.

Now, this is overall function of expected information gain, how to find skill gain from individual use?
It's simple, take the probability of outcome that hasn't occured. It will give you average skill gain like the above.

Notice that attempting tasks at which you can't succeed or tasks that are trivial will not advance your skill no matter the time spent.

Of course, you'll have to adjust your skill gain for levelling rate - a summoning spell may be cast once per battle or even few battles, while stabbing people with sword for the same effect will have to be much more frequent.

If the challenges can't be just obtained on fly (finding new effects and formulae may be necessary for making new and more powerful spells, powerful enemies don't just roam the countryside, well locked chests are typically found in well guarded castles, etc.) this will force player to actually play the game in order to improve skills.

This limits what you can grind on.

Now, the second part is asking "does learning something is always of benefit"? You'd like to say "yes", but consider that:
-learning stuff usually has some cost, at best it's the time you could have spent learning something else
-people ultimately have limited potential, learning new unrelated stuff generally becomes harder and harder

We are no longer in abstract XP rewardland. Learning something isn't necessarily a reward, it's just something that happens. Whether it's a reward is entirely situational. I'm sure every one of you who ever had to memorize some shitty poem in school will agree with me here.

So:
Let's make unmodified difficulty of first level up the same for all skills regardless of their initial level
Let's make difficulty increase with each subsequent level up of any skill.

Increase may mean geometrical increse of skill gain needed for next level, subtractive penalty to effective probability used in calculation, whatever. It would be best if this increase of lavelling cost yielded a bounded function with actual amount of level ups over all skills tending to a fixed number as time spent on tackling relevant challenges tended to infinity.
You'd also want to make skill scale fine grained enough to let the player keep leveling up for reasonable number of hours played and don't really have it go visibly asimptotic on their ass until they should have by all signs finished this playthrough - but that's just math.

Of course, you can modify the actual skill-up rate with controlling attributes or number of associated skills or perks and their level - the best idea is probably to have math encourage specialization, but gameplay encourage at least some versatility. For example it may pay you off to do nothing but master your sword techniques for the entire game from purely mathematical pov, but then stone golems, and some environmental obstacles, and so on.
Making player unhappy make seem like a lousy strategy, but genuine dilemmas are fun and generally fondly remembered by players in spite of that.

This limits what you want to grind.

The last part is a bit more abstract and meta. It generally matters less the more you can press the player for time using other mechanics, such as supplies and upkeep in general, and the more you can limit opportunities or make them costly to exploit - for example by using spell components. It's also the only part Bethesda has been steadily improving in TES. I'm speaking of context.

You can make actions don't really raise your skills unless actually done for a reason. Protective spell will not raise your casting skill unless something tries to break this protection (may be based on probability of breaking or attack strength), pickpocket may not work when target has nothing interesting (value threshold, or marked important - like letters or keys) in their pockets or repeatedly with same item, offensive spell may require target that's legitimate and actively resists (so no self experimentation) and so on. Also make the gain scale with magnitude of effect caused so no shitty training weapons that do low damage so that you can bash something longer. This part is auxiliary but may help fine-tune your system.


This limits when you may grind.
 

J1M

Arcane
Joined
May 14, 2008
Messages
14,731
I don't understand why you guys are trying to patch holes in use-based systems in this thread too. It is a waste of space. In the other thread Telengard already went through the tedious and verbose process of explaining how these band-aids only serve to make a flawed system worse.

I know some of you are aware of the counter-points and refutations of what is being re-proposed here but choosing to ignore them based on personal preference.

I don't mind if you want to make some sort of use-based vs. planned training mega-thread, but I am tired of seeing this same discussion revert to point zero in so many different places.
 

baturinsky

Arcane
Joined
Apr 21, 2013
Messages
5,608
Location
Russia
"We know that it has no solution, too," said Junta, bristling immediately. "But we wish to learn how to solve it"
 

DraQ

Arcane
Joined
Oct 24, 2007
Messages
32,828
Location
Chrząszczyżewoszyce, powiat Łękołody
I don't understand why you guys are trying to patch holes in use-based systems in this thread too. It is a waste of space. In the other thread Telengard already went through the tedious and verbose process of explaining how these band-aids only serve to make a flawed system worse.
Do you have anything specific to say or need help GTFO?

For the record Telengard's points have been adressed earlier in the very same post I quoted.
 

J1M

Arcane
Joined
May 14, 2008
Messages
14,731
I don't understand why you guys are trying to patch holes in use-based systems in this thread too. It is a waste of space. In the other thread Telengard already went through the tedious and verbose process of explaining how these band-aids only serve to make a flawed system worse.
Do you have anything specific to say or need help GTFO?

For the record Telengard's points have been adressed earlier in the very same post I quoted.

Even if it was possible to patch up all of the abuses and problems related to grinding in a use-based system, it would still be a system that makes players want to grind and therefore flawed. Some might say more flawed than a naive system, since it would prevent the player who now wants to grind from being able to successfully do so.
 

As an Amazon Associate, rpgcodex.net earns from qualifying purchases.
Back
Top Bottom