The Illusion of Complex Character Systems

[Big Nose George]

Or let's assume that a quest requires your character to have a Science Skill of say 200 to hack a computer, and that you are now at 100. How will you as a player know that you will need to reach 200 to hack this PC?

He wouldnt you dimwit. He would just fucking create a character stereotype and just go along with the flow of the game. Cant complete a quest, though luck, move on!

Let's say you want a specific set of skills, but whenever you encounter a quest that can be solved using them, youre always a point or two short, so you'll never get the feeling those skills do anything for you. Some games let you know if your skill was too low, which is frustrating, but not knowing it all makes you wonder if they are ever used at all.
The best solution would be success at different grades. Ex: Low lockpicking might require more picks, at really low levels you'll break the lock open, but the sound would alert guards or something.

Yeah, feedback. How much is too much and so on.

 

[Sartar__]

I agree with the thread starter.

You could imagine a system where you commonly use skills ranked from 1-1000, and rolled d1000s for all checks. It would be very unnatural particularly in a fantasy setting and perhaps discourages making decisions on instinct... having access to very exact numbers and easily rounded success chances encouraging calculation.

There is a trade-off between realism and an easier task game balancing (for the designer) due to the the greater choice of variables vs greater simplicity and immersion (for the player).

I think 1-100 can have it's uses. Greater attribute ranges I don't think are desirable.

 

[Mattresses]

High level skill ranges (1-100, 1-1000, etc.), I think, make more sense in use-based systems. Level-based arrangements need a feeling of considerable payoff at each levelup to maintain player interest in the game system (which seems integral to such tier-based rpg's). In an ideal use-based system, we would be dealing with numbers in the tens of thousands, allowing for every use of a skill, successful or not, to contribute to an increase in chance of success at said skill. The argument that randomised skill tests lead to continual reloads is ridiculous, whether the player chooses to or not use meta-game tactics is irrelevant, continual failure at a task will lead some to cheating themselves higher skill levels regardless of the game system, real men play ironman anyway.

 

[laclongquan]

Fallout series use 300 system and it doesnt bother me any. Arcanum use the 5(?) system and it also doesnt bother me any.

 

[DraQ]

Strength, Dexterity, Hacking, etc. from games such as Fallout and Arcanum got me thinking: What's the point of a skill or attribute running from 0 to 100?

Smoother stat progression?

Just because you have 20 skills running from 0 to 100, it doesn't make the character system more complex, does it?

Technically it does - compare a hypothetical system with two different values for an attribute with the one with ten. Other things equal, the finer grained stat progression is, the more possible builds does the game allow, and this is pure mathematics, no "if"s or "but"s.

However, the more fine-grained the stat progression already is, the less it will benefit from further refinement, so in case of most actual systems the difference is indeed cosmetic.

Still, for cRPGs where stat system isn't generally the most computationally expensive part, where computer handles all calculations and where variables tend to have minimal lengths and few programmers today work around this because it messes the code immensely, why not?

Think about it. Let's say I was to play Fallout and I decided to invest in the science skill. At what point will I realize (without having read a walk-through of some sort) that investing any further is a waste of time?

That's equally relevant whether there are two or two thousand possible values for a stat. If the system is well designed and the game is fine tuned in regard to the system - at no point at all. If the system and the game are sloppily tuned to each other - you can't tell and it will vary wildly from skill to skill, from build to build.

Or let's assume that a quest requires your character to have a Science Skill of say 200 to hack a computer, and that you are now at 100. How will you as a player know that you will need to reach 200 to hack this PC?

That's pretty retarded question. Usually, you should just assume that minimum value of a skill is a noobtard, maximum value is a 1337 ub3r pr0, and compare relative difficulty of the task to this scale. Besides, what kind of faggot reasons this way? Usually you chose solutions according to the build, not grind your build for specific quest solutions - go play an MMO or something.

That system suck, it just make you use your repair/lockpick skill over, and over till you get good roll, you would have to loose something signification with each try to make it a real choice. You could for example loose expensive parts that may be of limited supply first your character would asses difficulty, and then you would decide if you want to risk the part, of lockpick.

Resource loss is good mechanics, but not always makes sense.
And reloading in general is a problem as it creates what amounts to improbable lucky streak from character's POV and throws all the tension out of the window from player's. On the other hand, save points are retarded consolish mechanics that usually amounts to a lot of backtracking and goes against the purpose of saves in the first place, while the ironman is unreasonably draconian in any structured game (one with actual plot) as it forces doing the same thing over and over again if you die.

Since I do not want to repeat myself, go here if you want to read my musings on the possible alternatives or participate in the discuss!!ion.

 

[Kraszu]

See, its depended on the developer, not the size of the number...

No it depends on both of those things. System with 5 skill lvls is easier to balance then system with 200 points (that is practically impossible to balance), and system with 200 don't give any more meaningful options for the player over the system with 5.

You could say that. But I dont usually plan something. I just pump whatever I think might be interesting.

So having 1-200 instead of 1-5 don't really limit any meaningful choices for you.

 

[DraQ]

How about a system with 2?

 

[laclongquan]

System with 5 skill lvls is easier to balance then system with 200 points BUT it's also more boring. We dont even talk about the rigidity of the system: you cant experiment with different kinds of build.

 

[Kraszu]

How about a system with 2?

Nice straw man, how about a system with 10^999? Even more complex = better right? It must be balanced if there are no checks on most points then it should get smaller to the point where all would be useful.

System with 5 skill lvls is easier to balance then system with 200 points BUT it's also more boring. We dont even talk about the rigidity of the system: you cant experiment with different kinds of build.

Wrong, system with 5 could have programed more checks that would be useful for all parts of the game even in end game picklock skill with 2 can be better then 0 or 1. That would be much harder to balance with 200. It would still give as much meaningful choices you can have 2 in pick lock 3 in science etc. The skills would cost adequately, you could make skill lvl 5 more costly to better balance jack-of-all-trades characters with powergamers.

 

[DraQ]

How about a system with 2?

Nice straw man, how about a system with 10^999? Even more complex = better right?

How droll:

However, the more fine-grained the stat progression already is, the less it will benefit from further refinement, so in case of most actual systems the difference is indeed cosmetic.

It must be balanced if there are no checks on most points then it should get smaller to the point where all would be useful.

How about, you know, using skills in actual mechanics?

 

[Overweight Manatee]

Unless you intend to use each individual level 1-100 in a skill system, putting in 100 levels is just meaningless filler.

The only reason I see to use skill systems with 100 levels or more is to make it fine grained enough for a use-based skill improvement system.

 

[Kraszu]

How about, you know, using skills in actual mechanics?

That depends for speech checks I rather not have rolls (if they are dialogue tree like hard coded, not bartering for example). In practice the difference will still be slim even with rolls for locks etc having 200 seem rather pointless, and you will skill have some gaps for some number of point it will till be almost impossible to open/impossible to open.

 

[DraQ]

How about, you know, using skills in actual mechanics?

That depends for speech checks I rather not have rolls (if they are dialogue tree like hard coded, not bartering for example). In practice the difference will still be slim even with rolls for locks etc having 200 seem rather pointless, and you will skill have some gaps for some number of point it will till be almost impossible to open/impossible to open.

Lockpicking and speechcraft are hardly the only skills in the system and most RPGs tend to feature fairly complex combat mechanics, where stat system that is as smooth as possible makes a lot of sense.

 

[Pliskin]

How about a system with 2?

Nice straw man, how about a system with 10^999? Even more complex = better right? It must be balanced if there are no checks on most points then it should get smaller to the point where all would be useful.

What's wrong with a binary system? Isn't that more realistic? Either Yes You Can, or No You Can't. Is the long, drawn out crawl to meet the "this task requires Skill at 80%" task requirement really of any significance? Once you get there, it's an automatic Task Accomplished. Why not just create a skill system that is Yes / No, giving a choice of some number of Skills / number of levels.

Mechanics wise, if it makes you feel better, you can always add a random fail / success chance into the Skill Use. Or add items that you can use in liu of a particular Skill.

1's and 0's --- that's all you need.

 

[Sartar__]

Amber uses a kind of binary system. Attributes can be anything but only relative rank matters. It only suitable if you like a role-play heavy game.


"In any given fair conflict between two characters, the character with the higher score in the relevant attribute will eventually win. The key words here are fair and eventually - if characters' ranks are close, and the weaker character has obtained some advantage, then the weaker character can escape defeat or perhaps prevail. Close ranks result in longer contests while greater difference between ranks result in fast resolution. This concept has been developed further in John Wick's Advantage system. Alternatively, if characters' attribute ranks are close, the weaker character can try to change the relevant attribute by changing the nature of the conflict. For example, if two characters are wrestling the relevant attribute is Strength; a character could reveal a weapon, changing it to Warfare; they could try to overcome the other character's mind using a power, changing it to Psyche; or they could concentrate their strength on defense, changing it to Endurance. This concept is similar to the concept of escalation in Dogs in the Vineyard. If there is a substantial difference between characters' ranks, the conflict is generally over before the weaker character can react."

 

[DraQ]

What's wrong with a binary system? Isn't that more realistic? Either Yes You Can, or No You Can't.

It only works if a system is extremely detailed andconsists of a large number of subskills/perks/knowledge modules. Then it can be quite good, but should still use a finer, more continuous attribute system to manage the performance of those individual building blocks. In fact, skills as we know it are to a large degree an abstraction of such large collections of non-interchangeable building blocks.

Otherwise, there is whole range of skill between an absolute master and complete noob, so no, unless the system is extremely detailed, such approach is extremely unrealistic.

 

[Pliskin]

What's wrong with a binary system? Isn't that more realistic? Either Yes You Can, or No You Can't.

It only works if a system is extremely detailed andconsists of a large number of subskills/perks/knowledge modules. Then it can be quite good, but should still use a finer, more continuous attribute system to manage the performance of those individual building blocks. In fact, skills as we know it are to a large degree an abstraction of such large collections of non-interchangeable building blocks.

Otherwise, there is whole range of skill between an absolute master and complete noob, so no, unless the system is extremely detailed, such approach is extremely unrealistic.

Because yr going with a sliding scale. Say, on a scale of 1 to 100, Noob = 20% proficient and Master = 100%. They can both accomplish the same task, just one of them will be better at it. You're not talking about whether the task can be accomplished, but how well it can be accomplished. Isn't that what leveling is all about? Level 1 (noob) you can use a weapon --- level 20 (master) still using the same weapon, only now you do more damage / chance to hit is higher. Point being, it doesn't stop you from using that weapon, you are just better at it.

Gradient scales only matter if you've created a system where shades of grey have some effect. Those sort of systems usually only come into play in combat situations (accuracy and damage), or crafting items (usually relative worth). OTOH, most other situations call for a cut-and-dried Yes / No outcome. Did I open the lock? Did I sway that NPC? Did I repair that computer? Yes / No.

Yes, it would require a game system with an ungodly amount of Skills to choose from, but so what? Isn't that part of the appeal of RPGs?

The only thing we seem to be arguing about here is Base Chance of Success, and what the odds should be.

It would be easy enough to do by having a base set of Skills, and then having yr BCS derived from player Level. Would certainly make leveling up actually mean something.

What's really idiotic about most game systems is time scale: Your character goes from noob to Grand Master in anywhere from hours to days --- maybe weeks. You wanna talk about unrealistic? Most characters should start out with an ingrained skill set, and maybe --- maybe --- the ability to pick up one or two more during the time frame of the game.

But where's the fun in that?

 

[DraQ]

OTOH, most other situations call for a cut-and-dried Yes / No outcome. Did I open the lock? Did I sway that NPC? Did I repair that computer? Yes / No.

Of course, without proper amount of component skills it would be limiting. Also, if the repair, for example, required one to jerry-rig a solution rather than simply replace some parts, sliding scale would make more sense - the repaired mechanism might fail with different severity after different amounts of time depending on the skill level.

Yes, it would require a game system with an ungodly amount of Skills to choose from, but so what? Isn't that part of the appeal of RPGs?

At least partly yes. Though someone would have to code and implement those skills.

Personally, I'd be happy to see a hybrid system - sliding scale system plus good amount of yes/no bits (heh) of knowledge working as plugins for those sliding scale skills.
Except, learning this kind of knowledge would usually require player to pick them on chargen, learn them in-game from specific sources or, rarely, with stringent requirement and limited chance, invent them on their own.

It would be easy enough to do by having a base set of Skills, and then having yr BCS derived from player Level. Would certainly make leveling up actually mean something.

Personally I loathe levels. They are completely abstract and force a lot of unreasonable stuff in game.

The character should be perfectly describable with stats, besides, mixing high level of abstraction (levels) with low level of abstraction (myriad of binary abilities) is just wrong.

What's really idiotic about most game systems is time scale: Your character goes from noob to Grand Master in anywhere from hours to days --- maybe weeks. You wanna talk about unrealistic? Most characters should start out with an ingrained skill set, and maybe --- maybe --- the ability to pick up one or two more during the time frame of the game.

But where's the fun in that?

TBH I'd love to see such a mostly static system, with bits of learning from in-game sources, maybe some use based thrown in.

Also, swaying an NPC just calls for sliding scale - social skills just can't be divided into neat, base components - there is just too much subconscious shit going on on both sides.

 

[Big Nose George]

See, its depended on the developer, not the size of the number...

No it depends on both of those things. System with 5 skill lvls is easier to balance then system with 200 points (that is practically impossible to balance),

God dammit man, just fucking divide it by 5 and you've got yourself the same scale. Fuck.
But I give you that it can be tricky. For example, game has 20 skills, skill goes to 100, character gets 20 points per level. Now the player decides to put 1 point in each skill, each level. It would be impossible to make this build playable. But! dead end builds are ok in my book. And! you have the same problem with 1 skillpoint per level and an even distirbution by the player. Not as severe but the same problem.

and system with 200 don't give any more meaningful options for the player over the system with 5.

Define meaningful. As I said before, a clear cut 1-5 scale can be boring. But if you have something along the lines of a quest with a check for a max value in "Outdoorsman" skill in a game where the max value is hard to obtain, and you get for pumping that skill something special, it feels good, it adds fluff and as Zomg said - fluff is important.

You could say that. But I dont usually plan something. I just pump whatever I think might be interesting.

So having 1-200 instead of 1-5 don't really limit any meaningful choices for you.

Yeah, I dont care it its 1-200 or 1-5. Its the developers task to make something interesting out of their character system. But a 1-XXX scale has absolutely no logical reason to be inferior to a 1-5 scale. Except maybe that some people just cant do basic math and that concern is valid in the day and age of quest compass and "rpg is riding a horse and killing things"...

 

[MetalCraze]

I agree with the thread starter.

You could imagine a system where you commonly use skills ranked from 1-1000, and rolled d1000s for all checks.

You really fail at pulling strawman

What's wrong with a binary system? Isn't that more realistic? .

No it isn't. There is always a chance of success depending on your skill as well as a chance of failure depending on your skill. That's what dice rolls do - chance. And that's why in, say, Fallout even when you had high skills in something the chance of success was 95% because even pros can fail. However who says that if you are weaker than a bull you won't bash the rusty lock out provided you strike at a correct place (again chance) - like in, say, JA2.
Binary system is always a predictable boredom.

 

[bhlaab]

Unless you have threshholds for skill use (Such as Fallout 3's 25-50-75-100 Lockpicking) then any raise will at the very least help a die roll, even if it's 51 to 52.

So what this really is is a question of economy.

If your skill goes from 1-100 then that just means that skill points have less value and you will likely collect more of them more often.

If it's 1-10 then that means skill points are rare and have much more value. Levelling up "means" more, but it happens less often so it should all even out, hypothetically.

 

[GarfunkeL]

OTOH, most other situations call for a cut-and-dried Yes / No outcome. Did I open the lock? Did I sway that NPC? Did I repair that computer? Yes / No.

That's just laziness. Consider this - Did I repair that computer?

No, I caused it to explode, killing myself and setting the house on fire
No, I caused it to explode, hurting myself
No, I caused some serious mechanical damage and it's complete wreck now
No, I cause some slight mechanical damange but maybe there's something to salvage here
No, I just can't find the problem, everything seems to be allright
Yes, but so badly that it will overheat very shortly
Yes, but somewhat badly so that it will break down in few hours
Yes, but it's still a crappy piece of shit and will break down any day now
Yes and it'll probably last a good few months
Yes and thanks to the nice spare parts it runs like a brand new one

Might be clunky in a P&P-system but with a cRPG, rolling through various tables is not a big deal for the CPU, as long as developer is imaginative enough to create enough variables.

And make it always a roll, no automatic successes and suddenly your 0-100 (or 0-300) system makes sense!

 

[Krash]

And make it always a roll, no automatic successes and suddenly your 0-100 (or 0-300) system makes sense!

So, reload until you get what you want? Rolls for skillchecks outside of combat - I'm very sceptical.

 

[GarfunkeL]

Hey, there is no way that a game dev can make player exploiting or cheating disappear. If nothing else, there are the +1 - +N trainers or save-game editing. I don't understand this obsession that devs should make their games impossible to exploit/cheat. Sure, remove really obvious exploit-bugs (like multiplying items etc) but if your game allows saving/loading it's going to be exploited. Don't lose sleep over it.

 

[Tails]

Fallout 1/2 and Arcanum are just games from the Dark Ages of gaming, where each game tried its hardest to be as user-unfriendly as possible, so you had to read a walkthrough to get through some parts. I mean, did anyone of you guys discover that you can blow up the radscorpion cave on your own?

Yes, I discovered it at first play-trough. Also you are wrong.