The three levels of choice-making in an RPG

2, of course. The only one to acknowledge the existence of a little something called "party".

 

This is a silly argument.

1) Not every decision needs a unique outcome
2) Not every decision need to take into account other decisions
3) Not every decision needs to be taken into account by other decisions
4) You can use indirect systems

I've played a game of civilization where I've made 1005 or 10,005 decisions. How did that fit in memory?

 

Yes, some player could come with some absolutely crazy ideas but a cRPG doesn't need to keep track of all the ideas players might have just the most obvious ones, this would be light years ahead of the today's systems of blue, red or green choices at the end.

 

A proper RPG will have several choice layers on micro- and macro-level of the world, which amount to the following types:

"Do I stealth this area. or kill everyone, or maybe even have an option in-between?"
"Do I pick the lock for this door, break it, or find a way to steal the keycard from the NPC, or use the hammer on this hole in the adjacent wall?"
"In which manner do I convince this NPC of something? Threaten/logic/seduce etc"

(note that not all choices above could be available, depending on character build)

"I have found the little girl, but she's become a homicidal maniac. Should I tell her parents that I found her? If not, should I leave her be, or kill her to stop her from killing others?"
"Do I help this settlement against the raiders, or do I help the raiders demolish the settlement, so I can move up the ranks and either continue to roll with them, or kill their leader?"

These are the basics. I believe there can be more - such as choices based on previous actions in the world, that give you new dialogue to push NPCs around with.

Not all choices have to have "real" consequences. Often it is enough that the choice itself is either mechanically entertaining, or simply stays with you because of a tough decision you had to make.

Ideally, a lot of choices shouldn't be even presented as "choices". I.E. killing the girl or not, going to her parents or not, those are simply actions player may choose to do within the existing gameplay context of the world.

Big settlement-level choices make a neat display that can be shown in slides at the end of the game. But the game should contextually integrate them gracefully, instead of just the player stumbling upon a random and obvious Choice Of Grandness(tm).

 

1) Not every decision needs a unique outcome
2) Not every decision need to take into account other decisions
3) Not every decision needs to be taken into account by other decisions

Excactly that I meant with "pseudo choices have to be implemented" -> many different choices have to lead to the same outcome

I've played a game of civilization where I've made 1005 or 10,005 decisions. How did that fit in memory?

Civilization, like many other games simply save the map with the current situation. Strategic games don't need to "remember" much.
Each concrete map situation can be the output of millions of prior decisions.
CRPG saves on the other hand are growing increasingly (look at Skyrim for example). They need a least some history info.

 

3) Seems like special case of 2). Theoretically you may want some checks to override others, but but this may still end up being 2).

I don't see any choice as being more monocled than the others - obviously, when you have mechanics handling something (like INT stat) you'll want to use it where relevant, but it will still leave a lot of potential choices where no mechanical check is applicable, plus making a game that plays itself is not and should never be a goal of cRPG design.

Other than that 1. is technically less RPGish, if only because it can and does occur in non-RPGs, while the rest is cRPG exclusive (at least as far as stats are involved - you might easily conceive of similar check in, say, an FPS based on inventory contents or health).

 

I think that 3 differs from 2 insofar as it is depending on class - a wizard will have different options from a thief due to his background (reputation of the class in society, education, class-limited skills, etc.), while both could e.g. learn some diplomacy or lockpicking (even though they might have a different aptitude for it).

What's more monocled? Probably 2>3>1, but ultimately a good RPG can and should offer all of them - you should be able to choose paths, use skills and have the game recognize your class in meaningful manner.
However, I'm coming more from the PnP background here, and putting all that into cRPGs will run against a wall of what is possible to simulate in the restrictions of cRPGs eventually. Although most games stay way below that anyway.

 

Save drowning kid or don't save drowning kid

 

If you can pick any option, then that means (1) some options just never really appeal and never get picked, and (2) there is no joy of discovering options on subsequent playthroughs - meaning it's not just that exclusivity provides replayability and varied experience, but that inclusivity can stop players from enjoying the options they have.

E.g. if you were to play through Torment, Fallouts, etc. with a cheated 10/10 character, you'd get the option to use your speech skill to manipulate your interlocutors every time; and you might pick it because it seems 'best', or it looks delicious and appeals to you personally, etc. - leaving all the other options in the dust. You might never even realise some of the options you had because you never took the time to look for it.

 

It's only shallow and pointless if
a) it's reasonably easy to max most important skills (esp. in party-based games this is a common threat) and
b) the outcomes are too arbitrary and similar:
e.g. when you can either lockpick a door, bribe/charm/bully the guard for the key, kill him or sneak around it by using a different route, but all for the same outcome, i.e. you pass the obstacle using different flavors, but they only differ superficially, there's no further difference.
In comparison, picking the lock might give you a good entry point if you manage to do it but you also need to evade the guard, the "social" options might be very situational or depend on finding additional info earlier (like what does he like/fear), killing him might rise an alarm, sneaking around the door will have you enter at a different place that has other (dis-)advantages, etc.

 

I don't like you not being able to do a quest at all just because you don't have the stats. In Fallout for example you could finish multiple quests thru stats check but you could also simply go and try killing everything or something.

For example, Some hostages are being kept in a bank, You could try and sneak in and try and save the hostages without being seen or you could negotiate and try and get the hostages released or go and kill everything. Of course depending on your choices, there is an ideal solution where no one dies or get hurts or a bad decision where everyone dies or something like that.

I think Josh said if you pick a stat check in PoE, that doesn't mean its a win option or something, but it may lead to a bad or a worse situation, I really love that and I hope it turns out great as it sounds.

 

I don't think there are many players who genuinely like that. It's the kind of thing that only people who replay a game several times would enjoy, because it suits their playstyle. What most people seem to want is just to have choices that have a consequence instead of simple cosmetic effect (i.e., not being asked if you want to embark on thy quest against evil and being told "oh, but thou must" if they refuse, not being faced with the choice to nuke a town just to have the only local questgiver survive and take up residence in front of the rubble, not being presented with two exclusive skilltrees that are differently-flavored copies of each other, etc.)

 

That's only true if most (or all) of the options are presented through dialog - which is just bad design in itself.

 

Regarding the complexity of making a choice and reacting: there is no need to enumerate, check and envoke all the possible game states. Only a single game state needs to be stored in memory. Each game state can / is / should be represented by a (large) set of variables. When a choice is made, a subset of variables changes values. When the consequence is invoked, a subset of variables is checked and the game 'reacts'. Total number of distinct subsets of variables that is altered and / or checked within the course of the game is finite and constrained by the development resources.

 

It's ideal versus reality. People (I think) want a game in which hard choices have to be made and there are hard consequences that make sense. Since no such game has been made, they just powergame the shit out of what is available and have montyhaul adventures.

 

They don't have to be psuedo choices. Imagine 5 decisions that can each give you +10 reputation with a faction. Are those psuedo choices?

Who plays an RPG multiple times and picks the exact same options every time?

 

I have never met anyone who has never maxed out Wisdom in Planescape Torment.

Show Spoiler Hide Spoiler

I could've mentioned some JRPGs but that would be too easy.

 

Wait, what.

No, seriously, I don't understand what you've said.

You mean like in a use-based system?


Anyway, you could broaden stats into state variables in general (which would make neither #1 nor #2 RPG exclusive) and simply divide choices into unconditional and conditional.

Isn't #3 a subset of #2 in this case as well?

Unless you distinguish between shared mechanical layer and metamechanics responsible for stuff like AI and decision routines that needs to differ between PCs and NPCs, and make #3 refer to the latter specifically.

That's what state variables are for - so that you don't have to deal with exponentially growing tree of outcomes.

Anyway, in terms of tracking consequences you basically have 4 levels of choice:

1. Biowarean choice AKA non-choice. You're given a choice but it's immediately ignored and doesn't affect further progression of events. Sometimes it's actually called for, but usually it's just a case of trying to pass fake choice as an actual one.
2. Local choice. You're given a choice and the outcome is tracked for some time and discarded afterwards. Choice has no permanent consequence, but may nevertheless impact gameplay or story in interesting ways. It's usually tracked like #4 but with limited scope.
3. State change. Effect of your choice alters state of the game and can be tracked indefinitely, but it does not necessarily produce unique consequences. It's a powerful yet manageable method, many choices can be implemented in this manner.
4. Individually tracked choice. Outcome of your choice has to be tracked individually. Only few choices need this and you can only implement few choices this way due to combinatorial explosion. Do note that if order of several such choices doesn't matter they are actually #3 and should be implemented as such even if it requires highly specific state variables. Otherwise you're just fucking yourself over.

That only means that you made some shitty options.

 

Infinitron I'm not sure what your armchair psychology is getting at. What does it mean to 'embrace it' here? Embrace it as in let everyone do everything? In which case the fun of building an A Team would be gone, because a B Team could easily do everything too. Or are you saying that, for example, a game where you have enough skill points to become good in everything (i.e. Bethesda Design) is actually better for players than a game which is more parsimonious and makes you specialise (let's say, AOD)? Sure, a balance must be struck, but anyone arguing for Skyrim balance over AOD balance would have to give me a really good argument beyond "hey we play games we're all low class scum we just want carnal pleasure anyway". There's other considerations, e.g. the length of a playthrough, that comes into play here, too, of course - it would be boring if your Skyrim guy had to go 100 hours on just bows.

DraQ Sure. There are good and bad ways to do it. I'm trying to address the point that even when it's not done so well, it's still better to have some exclusivity.

 

Actually yes that would work as well. Usually use-based skill systems just get really meta-gamey though. Which is a whole other can of worms.

 

Well, again- You add another layer of complexity, my friend.

Say, for instance, I know that every 'build' will be viable for different outcomes and experience through play. This is a 'monocled' version of this instance. I KNOW from the start that a charismatic character will be able to sometimes talk his way through an encounter while a very charismatic character can sway pinions, seduce and generally pass most things through guile and manipulation (for good or ill)- This would allow for diversity in not only the choices you will get ( A charismatic thief will have different options than a charismatic priest with different outcomes) but also 'degrees of success' further branching the system.

If 'build' really (in truth) only means physical powerr vs magic vs archery with a small smattering of the other skills being used in some instances but with similar outcomes- Its gay. Its the illusion of choice and leads to meta gaming. I know my Mage needs Int and probably some dex and whatever stat is pertinent to 'magic points'. I skip strength, wisdom or whatever they rename wisdom to (usually for priests) and charisma and often see constitution (or its renamed stat) as only for hit points.

IF Constitution actually played a bigger role than 'HPs' and Charisma means something a bit more and wisdom could even help a mage in a valid way- In other words, if all the skills you chose mattered in such a way as to make it very difficult to decide how to build your character to your specific tastes and this actually changed the game in a meaningful way- Its a win.

The problem is more in execution, I think. Games like Vampire the Masquerade: Bloodlines did pretty well with this concept but failed in other ways. Most games promote certain builds (only a few) that are viable and most of the choices in character spec are wasted unless they are used in those few viable builds.

 

In other words, you don't have a point, you want other people to make your point? I'm only being harsh because I don't think "exclusivity = better" is really such a dogmatic idol that needs to be brought down, I think significant exclusivity happens in very few games and people are often split on it.

If you have a break-in quest where you need climbing to jump the wall and speech to talk the guard, crudely speaking, you could set the checks at 3/3 so quite a few characters could choose either option; set them at 8/8 so that only dedicated specialists can pursue the options, or even have to give up on the quest altogether; set them at 6/4 so that som eoptions are more obscure and difficult, but there's at least an 'easier' and maybe less optimal way to progress. It's hard for me to think of situations where inclusivity adds to replayability, immersion, whatever. It's easy to think of situations where too much exclusivity might cause frustration, but that's not the same as inclusive choices.

The key problem that remains is that make the options too inclusive/easy and you dilute the consequences of each choice and also break down the general feeling of C&C over the long term, the feeling that your character building matters. Too much exclusivity might make you feel that your character can't do half the cool shit; too much inclusivity makes you feel that none of the work you put into the character over time actually enabled you to do any cool shit, it was always available in the first place.

DarKPenguiN The primary challenge to that is making all skills similarly useful without making them the same or having so many options for every challenge that it hardly seems to matter. If you can solve problem X with speech, combat, or stealth, fine, but if there are skill checks for survival, repair, science, charisma and endurance too, then we come back to the problem...