Difficulty curve in RPGs

[laclongquan]

A RPG without combat is either a visual novel or an adventure game, not RPG~

 

[Aoyagi]

A RPG without combat is either a visual novel or an adventure game, not RPG~

"Combat is not all that important" is far from "I don't want combat at all".

And no, visual novel implies linearity.

 

[worldsmith]

Decent post overall, but this i cant agree with.

Keep in mind that the 2nd part of what you quoted was just a quicky proof-of-existence example, not necessarily exactly how I would do this. (In fact, definitely not how I would do things as I do not wish to go the "static world" route at all.)

Especially putting a cap on character growth.

Technically what I suggested wasn't a cap - you could always improve your skill by sinking more skill points (or more training or whatever the game uses) into it. It's just that skill point #N+1 doesn't buy quite as much skill as skill point #N did. (Such functions can have a theoretical "limit" - but that's a limit which can't actually be reached - not without literally putting in an infinite number of skill points.)

Just as an example, you could make skill point #sp be worth an additional 0.99^sp amount of skill. That always gives some non-zero benefit per skill point. You can use other constants, e.g. 0.9 or 0.999, to make that amount diminish faster or slower. And of course there are entirely different functions that also result in diminishing returns. It all depends on exactly what you want the curve to look like.

In all real-world endeavors, skills don't go to infinity. In fact, they do pretty much exactly what I have suggested - it's usually not too hard to pick up the basics, progress remains noticeable through the medium levels, but as you start to get really good at something it gets harder and harder to get noticeably better - you tend to start to notice improvement over years rather than over hours.

Now I'm not making an argument from realism - that it must be that way because that's how the real world works. But why not? Why shouldn't it work like the real world works? What's the benefit to gameplay of flat skill curves? I certainly see the gameplay problems flat curves produce, but I'm not seeing the benefit.

Note that besides affecting issues of balance between PCs and NPCs, the suggested skill curve also encourages a player to learn and use more skills, because the lower part of the skill curve is easier to pick up - getting their first 5 pips in tracking may very well be worth more to the player than getting their next 0.237 pips in swordsmanship. If you don't do that, if you just have a flat skill curve, you end up quite frequently with lots of "dump skills", with almost all skill points being funneled into very few skills. This also means that with the non-flat curve the decision process, deciding where to spend your skill points, is much more interesting because there's an actual meaningful choice to be made that is very situation dependent.

But also coming up with statistically impossible to beat scenarios.

Why not?

If the game world has 3 main kingdoms, are you supposed to be able to just walk up to any of the capital castles and with your party of 4 characters completely wipe out the 300 troops defending it? (And no, they're not retarded and aren't going to attack you 2-6 at a time. They're going to have the main portcullis closed, you standing on the outside, and them hitting you with a 100 archers at once with a serious height advantage, with some ballistas thrown in for good measure.) If you want to steal something from that castle, you better find some way other than frontal assault.

What if the game has actual gods in it? Are you supposed to be able to defeat them too? If they decide to smite you because you piss them off, are you supposed to be able to handle that, maybe chug a health potion and brush the singed skin off your skeletal remains and just shrug it off?

Keep in mind, I'm describing an RPG where you aren't required by the game to do anything. There is no pre-written plot, so no scenarios you are required to beat. (If the game is generating objectives for the player rather than the player choosing their own, then yes, in that case the game should attempt to keep those objectives reasonable. But that's a "for dummies" mode of play and not what I'm really talking about.) If you the PC/player don't think you can beat a given scenario in the game world (or think you can but at far greater cost/risk than it's worth), then you should do something else. (And making that determination, and getting the information useful for making that determination, can and should be part of the gameplay.)

Plus, the whole point of this thread is the inability to balance something which is highly unknown (PC power combined with player skill) against some other thing which is not necessarily entirely predictable either (NPC opponent power combined with AI skill). You simply can't, under any circumstance besides astronomically impossible good luck, succeed in making NPC opponents that are "just powerful enough that the player barely beats them - i.e. are really challenging" without also having made some NPC opponents that are "too strong". It's just never going to happen. If you have to have all NPCs be beatable, there is no real option other than making the PCs overpowered. Therefore, if one doesn't want overpowered PCs, one shouldn't even try to make all NPCs beatable.

Also i dont believe npcs should "grind it up", sounds awfully close to level scaling.

Should not NPCs operate by the same rules as the PCs?

And it's absolutely nothing like level scaling. In a proper living world, just because the PC is level 1 does not mean NPCs are wimpy - some of them are younger than the PC, some of them are older (and more experienced). Knights of a kingdom are going to have had much training and probably a decent amount of experience and subsequently (if the game has "levels") are going to be much higher than level 1. And yes, they may continue to train and acquire experience while the PC exists, or they may get themselves injured, killed or dismissed. (You don't expect them to stand frozen like statues for years on end just waiting for the PC to come along do you?) But except for the effect of time (which limits what everyone can do by its finite nature), their level is not at all related to the PC's level. If the PC engages in much more intensive training than the knights, then the PC may very well catch up to and surpass them in fighting skill. If the player decides the PC is going to be a merchant or something else that perhaps doesn't involve much if any combat, then the PC may never be anywhere near as skilled at fighting as any of those knights.

In other words, once again, things work as they would in the real world - just as any player who hasn't been mentally corrupted by existing RPGs would expect them to work.

But remember, the party is the one tackling the biggest baddest obstacles in that place, they are the ones getting the shinnies

Huh? Who says?

There could be a config option for being "the chosen one" for those players who insist on it, for having some extra rules that apply just to the PC party to make them nearly guaranteed to be so out-of-whack powerful with the rest of the game world's characters that they and only they can stop all of the Big Bads, but why should that automatically be the case in every RPG? (And really, it makes no sense. If the Big Bads are so powerful that no one else can stand against them, they should just go ahead and do their thing before the PC levels up from zero to hero. I'm more interested in a game world where things make sense, not fantasy fiction that only a young child could listen to without doing a face-palm.)

The PC party should accomplish whatever the player decides to try and manages to accomplish. That's it. Destiny is just another type of pre-written plot which is exactly what I'm saying shouldn't be.

That said - players do have such an advantage, and that is their brains. The AI, even if made well enough to mostly avoid doing dumb stuff and to even do many things you do not see in current RPGs (like conducting strategic wars between kingdoms, resource management, and engaging in production and trade) is not going to be anywhere near as smart as a human for some time to come. Especially if you have one home PC powering lots and lots of NPC AIs. And isn't that what's supposed to put the PC/player on top - relying on their gamer smarts rather than special advantages granted to "the chosen ones" by the game?

and being in any actual risk of losing.

How are the NPCs at any less risk of losing? I mean, they can die, right? Isn't that "losing"? If there's some big bad enslaving characters, and they end up enslaved for life, isn't that losing?

I didn't say so in my previous post, but it does sort of logically follow from there being no pre-written plot - there need not be any predefined win/loss condition either. My current thinking is that, if the game is going to "judge" the player's progress at all, it would be much better for the player to select the criteria by which the player will be judged. (That could tie into roleplaying as well - by setting goals consistent with the type of character they wish to play.) And while that could include some discrete events (e.g., helping one kingdom conquer another, assassinating a certain evil lord, etc.), it could also include things that are on a more ongoing basis (e.g., number of undead exterminated, amount of gold acquired, favor curried with some guild or deity, etc.)

Bte, You do know you are basically describing Mount and blade across your entire post, dont you?

Yeah! Another game I need to try.

(I already bought the game, but from what I remember of the reviews I saw I wasn't really impressed. It's perhaps the right nature - a more free-form do-what-you-will while the world {to some extent} does as well type of game, but not super strong in execution. But, the proof's in the playing...)

 

[Lhynn]

I understand where you are going, and i think you are correct for the most part.

1. Character growth. This one is debatable, but as long as there isnt a cap and its rather flat im ok with it. See mount and blade for this.

2. Statistically impossible to beat scenarios should exist for the most part, i just dont believe they should be scenarios that should be presented to the player so much as the player should actively seek them for them to show. Again see M&B, mess with people more powerful than you and they will chase you, weaker armies tend to be faster, so you can usually run easily, especially if you get rid of units that are not mounted.

3. Making non player characters grow works well in open world rpgs, again see M&B, but it doesnt work in any other kind of rpg.

4. Chosen one. Its not about the player character being the chosen one, its about the players overcoming obstacles and growing out of it. challenges that would probably easily wipe your AI.

5. If the NPCs are at risk, then the NPCs cannot be said to evolve with time, the opposite is true in fact, they will get defeated eventually and lose everything, the player only has to reload to avoid that. Or are you suggesting perma death?

You really need to play M&B, its just what you want.
Your proposed design philosophy only works for open world games, which is fine. But not everyone wants to play the same game over and over again, heavily scripted ones can be fun, ones with branching and some freedom too, the ones with static open world too, theres room for all of them. A lot of good games had none of this stuff and still worked. And especially on this thread, where the idea is to figure out a decent difficulty curve that could be applied to most rpgs.

 

[worldsmith]

3. Making non player characters grow works well in open world rpgs, again see M&B, but it doesnt work in any other kind of rpg.

Not sure what you mean by "works" here. And if "grow" can just mean "gets stronger by training over time", I don't see why it can't be made to "work" in non-open-world RPGs (though again, I'm not sure what "work" means, so maybe there's a catch there). Legends of Eisenwald is an example of a game that has NPCs that get stronger over time (presumably via training while in castles). Yes, it means if the player just has the PC sit on their ass too much they may fall behind and ultimately fail - but I don't consider that to be a problem with the game.

4. Chosen one. Its not about the player character being the chosen one, its about the players overcoming obstacles and growing out of it. challenges that would probably easily wipe your AI.

Not sure what your point is here. I already said humans have the "human brain advantage". On the other hand, NPCs can have a rather large numerical advantage, e.g. if the player controls one adventure party, but across the game world there are 100 NPC adventure parties. They might not make good choices as consistently as the PC, but with such numbers, luck is likely to favor a few of them quite well. And that isn't just speculation - you can see that very effect happen in existing strategy games when played against many AIs. It's really easy to get into a situation where the player hits a bit of moderate to bad luck, e.g., with respect to location or resources or who's next to them or just the way some early battles go - leading to slow progress, while one of the AIs experiences much better luck - and subsequently quickly grows large/powerful. By the time the player's empire meets that AI's empire, the player can have a major challenge on their hands.

What you say still seems to be indicating some sort of destiny - that the PC is supposed to do well. I do not agree with that. Failure is an option, or it's not a game.

5. If the NPCs are at risk

"at risk" doesn't mean "100% automatic death before any significant advancement is achieved" - it just means "at risk". NPCs will have to manage their risk, just as the PC/player does. (Again I guess I should point to strategy games as an example - some of them have heroes that level up using the same mechanics as PC heroes, and at an appropriate difficulty level they will do well enough that they will eventually come and rape the PC heroes. And since I don't want to keep repeating this, just note that in addition to the replies I give below, you should keep in mind the fact that some strategy games already achieve these things you say can't be done in RPGs.)

then the NPCs cannot be said to evolve with time

Sure it can, and it can be true as well:

1. It can be said: In an RPG which doesn't simulate NPC C&C, the game could of course just apply some (made up) "statistical model" - taking various things into account, apply some % chance they've died, else they've gained some amount of experience. (While this is one way to go, it's not really an appealing way for me.)

2. It can be true: In an RPG which simulates NPC C&C, there's no need to just wing it like that. Instead, whatever actions NPCs perform, even when the PC is not around to see them, happens in the game world, directed by the NPC AI, and governed by the same game world mechanics that govern the PC. Find them in a tavern later and maybe they'll even tell you all about what they did. (I'm going the simulation route myself for various reasons, including that simulation provides a game world which is both consistent and dynamic.)

the opposite is true in fact, they will get defeated eventually and lose everything, the player only has to reload to avoid that. Or are you suggesting perma death?

No, it's not true. No, I am not suggesting perma death (though I do believe it should be an option for those who like that style of play).

As I said earlier in this thread:

There's still save-scumming to deal with as that can be used to allow the PC access to more dangerous grinding than NPCs can get away with, but there are solutions for that as well.

There are an infinite number of methods that may be used to allow NPCs to compete more fairly in the face of the player's power of save/load. Many have been used in the past. Many involve nerfing save/load: Perma death (one life) is one of them. A fixed number of lives, or the need to earn more lives is another that has been used in countless games. IIRC, at least one RPG made saving the game consume XP. You could score the game not on how well one specific PC (or fixed set of PCs in a party) does but instead let the player pick up new PCs as the old ones die/retire (which is how Darklands does it). Those (and quite a few more) are just the methods that have actually been used by existing games.

When designing a new game, the primary limitation on what you can do in this area is your own imagination. You could make it so if the player abuses save/load, NPCs get more XP from their actions. If the PC has access to multiple lives, give NPCs (or at least some subset of them) that same ability. Maybe every time the player abuses save/load to bring themselves back to life, all NPCs get an extra life... except for 100 NPCs selected by the game - e.g., higher level NPCs the PC subsequently kills - which instead get two extra lives. Those are just some quick examples I came up with just for this post - for my own game design I have come up with (what I think is) much better stuff (and I'm sure there's even better ways to do it than what I've come up with).

All of the above said, it should be noted that some "gamers" don't care much about actual gameplay/challenge and prefer to just "explore", use the game world as a sandbox, try different things just to see what happens, etc. So a non-nerfed save/load option should be provided as an option. But making use of that option (specifically, using save/reload to erase negative consequences) should even in that case be reflected in the difficulty rating. Saying a game is played in "super ultra-hard I'm-a-god" difficulty even while the player is compensating for their horrible gaming skills by save-scumming their way to victory is pure BS. (And that solution could be used all by itself without any of the previous suggestions or anything else - i.e., any time the player abuses save/load, downgrade the difficulty rating that their game shows up as accordingly - let them go ahead and post their game log that shows they managed to defeat some Big Bad... at the "extreme retardo" difficulty level.)

You really need to play M&B, its just what you want.

Yeah, yeah, yeah - I need to try the game, but not because it's "just what I want". Don't presume that I have described everything that (or even some significant fraction of what) I want in an RPG in this thread.

Your proposed design philosophy only works for open world games, which is fine. But not everyone wants to play the same game over and over again

So, two entirely different open world games, with different settings, different lore, different characters, different game mechanics, different developers, different nearly everything else - are "the same game". I don't think so. Furthermore, I think the potential meaningful variety between two open world games is just as large as between an open world game and any other type of RPG - with only one really significant exception - the pre-written "hand crafted" story. And that just leads back to the old question: "Do you want to read a book or play a game?"

heavily scripted ones can be fun, ones with branching and some freedom too, the ones with static open world too, theres room for all of them

I read that and all I see is "crappier games can be fun too!" It may be true, but it's not very convincing for me. (My strong aversion to all things scripted is spelled out here using words such as "hate", "retarded", and "disease".)

And especially on this thread, where the idea is to figure out a decent difficulty curve that could be applied to most rpgs.

I don't think "figure out a decent difficulty curve" is the real issue - that's easy for many here who believe the difficulty curve should simply be "challenging (perhaps not monotonously so, but frequently enough) the whole way through". The issue that most in this thread have been discussing is "how can one achieve such a difficulty curve (while not ruining the game)?"

(And if you're going to change the subject of the thread to be "how to do that for most RPGs without changing them", then I can state with a very high level of confidence that you can't do that.)

I think with a pre-written plot and significant character progression with meaningful player skill involvement (and no lame stuff like level scaling), balancing all of the individual battles is not really possible. (To see why, see my previous post, specifically the paragraph starting "Plus, the whole point of this thread".) However, like a 4x/RTS game, achieving an early-game balance between "player is eventually attritioned into a no-further-progress-possible failure state" and "player eventually gets irrevocably ahead of the designed power curve win state" can probably be achieved with a combination of difficulty setting and consumables (where "consumables" generalizes to "anything that is used up and has significant cost to replace"). The consumables are important in this case because they make it so individual battles don't have to be balanced - instead they switch the focus to minimizing attrition in every battle. (To really work, this requires that even easier battles tend to involve some consumable use - otherwise the attrition challenge goes AWOL for such battles.) Basically this simplifies the balancing problem -- instead of trying to thread dozens of different needles (balance all of the individual battles), you only have to thread one needle (eventual fail/win state), which then makes the problem simple enough to handle with a single difficulty setting. Like a 4x, this is still unstable, and eventually it just "breaks", but can at least make the early game challenging. I suspect that (other than adding the main-plot-bypass I talked about earlier), that's about the best that can be done with such a style of RPG (without modifying it to be some other style). That conclusion seems to be consistent with player experience as well.

The above style can be "fixed" (modified into a "whole game balanceable" style) by taking any one element out of it. Take out the pre-written plot and it's easily fixed. Take out "significant character progression" (leaving little or no character progression) and balance is much easier. Take out "meaningful player skill involvement", meaning character progression proceeds at the same pace (relative to advancement through the game/plot) regardless of what the player does, and again balance is easier. Take out the restriction against level scaling and balance becomes easy.

off topic:
Any of those options can be made to work with respect to the topic of this thread (fixing the difficulty curve), but different players are going to prefer different solutions. My problem with the pre-written plot is not that a game using it can't be made to have a suitable difficulty curve (because it can). Rather, it's that you are taking a double whammy: First you tend to have severely reduced gameplay options from the pre-written plot itself. (That's not necessarily true for any given game, but in practice game devs put what could be "mechanics" and NPC AI in as scripted elements of the plot, meaning the given mechanic is not available for general gameplay use and NPC AI outside of combat is essentially non-existent so the player has no-one to play with/against even if the mechanics did exist.) Second, you either have to accept a broken difficulty curve, or you have to accept one of the options listed in the previous paragraph which further reduces gameplay. While it's possible for a game to recover from this double whammy - e.g., by having a strong AI and really good tactical combat (and/or a really good espionage system and/or a really good merchant/production/trading system, etc.), most RPGs don't bother doing that either. And of course the game that didn't cripple itself with the pre-written plot can also do those things, in which case such a "recovery" still leaves the game with the pre-written plot behind in terms of gameplay. Of course, none of this may matter very much to a "gamer" that prefers to "play a book" rather than engage in actual gameplay (and/or have far more freedom to role-play).

 

[Lhynn]

Man, you really love walls of text to say nothing new. Just do yourself a favor, go play M&B before continuing this argument, otherwise its beyond pointless.

Also yes, the real issue is figuring out a good difficulty curve that not only remains challenging till the end, but that is actually is harder when the stakes get higher, this has to do with pacing, especially on more linear games. As i said, i believe your approach is good for a open world game, but not every game is that game.