Choices in CRPGs

[BlueSalamander]

What is a good choice in a CRPG (computer role playing game) ?

Here’s a paper I wrote some time ago about choice and consequence in CRPGs and how we can make better games.

In pen-and-paper RPGs, freedom is king. Anything that a character could do, the player can attempt. A player could focus more on one quest and less on another. Actions are only limited by the player’s imagination and the ability of the dungeon master to provide an experience suited to the player’s actions.

But in a CRPG, there is no dungeon master, so the player cannot attempt things that have not been made possible within the game. Choices are limited to what the game offers explicitly. Therefore, it is important for a CRPG module creator to answer the question: what is a good choice? What choices should my game offer?

As a general rule, good choices display the following traits: exclusiveness, fairness, substance and simplicity. Let us now review those traits.

Exclusiveness

Good choices are often mutually exclusive. You can choose one thing or one action, but not all of them simultaneously. When choices are mutually exclusive, the player must think before deciding. Conversely, when choices do not exclude each other, the choice is very bland and boring as the player does not need to think.

For example, the map designer could drop on a map two barrels, each containing 10 gold coins. Then, when the player explores the map, he sees two barrels and must decide which barrel to open first (if any). So the relevant alternatives are to open barrel A or to open barrel B.

Technically, it is a choice, but the player will not see it as a choice, simply because all alternatives can be explored in sequence. It does not matter which alternative is chosen as the player will explore all of them in sequence.

The rogue-like game ADOM contains a number of mutually-exclusive choices. For example, in the beginning of the game the player receives a quest to bring back the body of a bandit leader who is roaming the countryside.

The player is free to undertake the quest or to ignore it and find other things to do. But the bandit leader runs away and disappears forever when the player reaches level 5. As a result, this quest becomes a very interesting choice. The player must decide very early whether or not to undertake the quest, which presents a high level of difficulty and a good reward if it is successful.

Later on, the player is presented with two possible quests and only one of the two can be undertaken. He can either hunt down an evil druid or seek to rescue the village’s carpenter. Both quests are interesting and they offer different rewards. When the player accepts one of the quests, access to the other is lost forever to the player character.

The exclusiveness of a choice can also arise from the natural flow of the adventure, usually as a result of actions undertaken by the villain. For example, imagine that an assassin from a foreign land has fired a poisonous substance at a dignitary using an enchanted blowpipe. The player characters are immediately asked to go after the assassin and retrieve the antidote from him.

The assassin spots the party tracking him down and decides to wreck havoc in the city in order to stop them. He gives that order to his henchmen before running away. One henchman summons a huge fire elemental and orders it to kill everyone. Another henchman stabs a merchant’s horse, sending it on a wild dash down a crowded lane. Another henchman slashes the throat of a young child passing by.

What will the party do? It can decide to ignore the city’s chaos and focus on pursuing the assassin to bring him to justice and save the dignitary’s life. It can focus on battling the fire elemental in order to save the lives of innocent citizens. It can try to stop the stampeding horse, thereby preventing dozens of injuries. It can focus on the child and have the cleric provide emergency healing to save the life of a seven-year-old girl.

Maybe the party can divide into two groups, but it will not be able to do everything at the same time. In this situation, some of the possibilities are mutually exclusive, giving the player a chance to set his own priorities. The situation also creates an interesting challenge where success can be measured both in the number of enemies defeated and the number of lives saved.

Fairness

Fairness is the attribute of a choice that determines the validity of the alternatives. If a choice is unfair or heavily biased, there is no real choice because it is obvious that one particular option is the best one by far.

Conversely, if the various alternatives all present advantages and disadvantages that cannot be assessed and compared easily, then the choice forces the player to think and select an option based on his own playing style or based on his own sense of tactics and strategy.

As an example of an unfair choice, consider a situation where the player finds a treasure chest. Imagine that the game offers the following options:
1- Look for a trap and if you find one, disarm it, then open and take the contents.
2- Let the Barbarian open the chest and take the contents.

Is this a choice? Technically, yes. But in reality, why would anyone choose number 2 if number 1 is available? With option number 1, you avoid an explosion. With option number 2, you trigger an explosion. That is not an interesting choice.

Or consider the following situation. The King of Andawar has told the player that he can ask a single thing of him. The player can ask one of the following:
1- Provide information about our enemies.
2- Provide a magic weapon.

Now assume that the information the king provides is information that the player already has, and assume that the magic weapon is a very powerful one. Again, there is an absence of interesting choice. The player will first make use of a saved game in order to explore both possibilities, and then he will correctly think to himself: ‘What a stupid choice!’.

From this discussion it follows that each choice should be crafted carefully so that each possibility offered to the player is valid and acceptable.

If a choice of a reward is given, the rewards should all be roughly of the same value and all should provide unique benefits. If a choice of a companion is given, each possible companion should have his own unique strengths and weaknesses.

If a choice is given between factions or allies, then each possible faction or ally should bring something that may be of interest to the player. If a choice is given between several actions, then each action should offer unique risks and unique rewards, and the higher the risk, the higher the reward should be.

If there is a choice between quests, each quest should be interesting in its own way. And if there is a choice about ethics and morality, then all choices should be morally ambiguous.

Substance

High-budget CRPGs are chock-full of what is known as the ‘illusion of choice’. The illusion of choice is the situation that presents itself when the player is given several dialogue options that all mean the same thing, or nearly the same thing, and that all lead to the same answers, or nearly the same answers.

For example the options could be the following:
1- Dear fellow, would you be so kind as to provide directions to our befuddled party?
2- Well hello there! There’s a gold coin for you if you point us towards the merchant.
3- Ugh, you da man there, say where da shops are?

Or, when talking with an NPC, the options could be the following:
1- Please continue with your childhood’s story.
2- Oh, it must have been so tough for you to go through all this.
3- You were so right. Your father was a nasty man.
4- Such luck you had! I envy your family’s wealth.
5- I would have never done what you have done.

All these choices are essentially meaningless and serve no purpose beyond creating a shallow, fleeting sense of freedom. In addition, their presence can make it more difficult for the player to identify those choices that are actually valid ones.

None of the options above present any advantages or disadvantages, making them very similar to the choice between opening barrel A or barrel B first.

When it comes to the substance of choices, it is very much a matter of quality versus quantity. Do we want many dialogue choices, all pointless and unsatisfying, or do we want a small number of actually engrossing and rewarding choices?

Good choices are choices that affect the rest of the game in one way or another. They could affect the player’s equipment, the player’s knowledge, the reputation of the characters in the game, the next combat encounters, the next puzzle, the next dungeon, the composition of the party of player characters, the experience points of characters, the gold owned by the party, the spells known by spell casters, the strength of certain monsters that the party will soon face, the allies that the player will encounter, or anything else that has a real influence on the way the game is unfolding.

Simplicity

Simplicity is the element of choice that counterbalances the Substance trait. Simplicity is important for the module creator and map designer. The simpler a choice is, the less work for the designer.

For example, it may not make much sense to create a module that allows the player to play both good and evil characters. This is because the actions that good and evil characters would take are often diametrically opposed. In such a case, it may be better to create two modules, one for good characters and one for evil characters.

Likewise, it does not make much sense to create a large module where, in the beginning of the game, the player must decide whether all his remaining adventures are going to take place in the icy kingdom of Shatterpeake or in the desert of Pharaoh Tutmanster. Again, in such a case, why not make two separate modules?

On the other hand, many interesting choices can be implemented without a large increase in the amount of work required to complete the creation of the module. Let us make a list of some of them:

• Choice between item, weapon and spell rewards
• Choice between recruitable companions, allied factions, enemy factions
• Choice between quests, small dungeons, or stand-alone monster encounters
• Dialogue choices favouring one companion over another (choice of friends)
• Choice of whether or not to have a romance with an NPC, and which one
• Choice of a quest-resolution style (combat, diplomacy, bribery, ingenuity)
• Choice between using the front door or a secret passage (avoiding alarms)
• Choice between morally-ambiguous actions (who should be saved or helped)
• Choice of the sentencing of a criminal (especially, whether to execute)
• Choice of the new leader of a country or community (who should be king)
• Choice of what structure to build in a castle (chapel, mage guild, more walls)
• Choice between endgames (e.g. demon destroyed, banished or escaped)

Conclusion

In the above discussion, we have demonstrated the importance of four specific factors concerning choice and consequence in CRPGs. For a module to be recognised as offering an enjoyable non-linear experience with a lot of freedom, designers should concentrate on the exclusiveness, fairness, substance and simplicity of the choices offered by the game.

PS - Feel free to post comments. Thanks!

[Tiavals]

Well written. In general, I agree with everything, but there are a few important exceptions.

Hopefully the following isn't too confusing, I had some trouble turning my thoughts to words:



For exclusivity of choice, I believe that there are times when a hidden choice can be very powerful. A choice that is not obvious immediately, but has to be discovered by the player, whether using his own ingenuity or the abilities of his character. That is to say, a choice that's only available if the player was smart enough to realize it, or if the character has a given skill or such. This creates a sense of impact to the actions. A player feels satisfaction if he can do something that another player hasn't been able to do(that is to say, he isn't spoonfed the choice). Suppose a situation where you find a special gem in the lair of a dragon. The obvious choices given to the player might be to sell the gem, use it's powers for your own, or give it to a NPC who wants it. Now, a hidden choice might be one where the player has to use the gem in a specific manner at a certain place or time. The player has to realize to use the item in such a way, probably if he has paid attention to clues or bits of lore given inside the game. The main point has to be that it is hidden enough that most people who casually play the game don't realize that the option exists. If it's obvious that a hidden choice DOES exist, then it somewhat defeats the purpose(for instance, the gem description explicitly says that you might want to use the gem at the Temple of Gems or such).


This relates to fairness as well. The hidden choice can be better than a regular choice if certain conditions are met, that are related to exclusivity in a different sense. If the choice can only be made by a character with a high skill of a specific type, then it excludes the choice from other playthroughs, and thus can't be done with every character. This sort of allows the choice to be superior to the other options, since you can't choose it if you don't meet the prerequistes. This mainly works with a hardcoded "in-game" prerequisite, like a high skill, and won't work with an "out-game" prerequisite such as that the player has to realize that you can use an item in a certain way. This is because the in-game reason can prevent a player from choosing the superior choice in a given playthrough but the out-game reason can't if the player knows the choice exists. On a larger scale, such choices can only work if there are multiple such choices in the game that realistically prevent you from taking all of the choices(such as the choices having completely unrelated prerequisites you're unlikely to meet during the game, for instance that all your party members are Good aligned, or that none are Good aligned).

As for substance, what counts as substance? Your examples obviously are clearly pointless in the meaningfullness of the choice, but there may be situations that are somewhat different. For instance, you might pick a dialogue choice that adds a line or two to the dialogue, which may present the situation in a slightly(or even greatly) different light, perspective or flavor. Even if you're forced to listen to the story of an NPC, suppose you had these choices:

1. Please continue.
2. I heard that story before and it's a bunch of lies.
3. I thought that particular event was the other way around.

Now, regardless of your choice, the NPC will continue what he was saying, but if you chose 2 or 3, he will say a line before continuing the story, such as:

2. You are right, it's a bunch of lies, but interesting ones at that. Anyway, I'll tell you my version and see if it's the one you've heard before.
3. A common misconception, no doubt born of translation difficulties.

The actual gameplay remains the same, but choices 2 and 3 alter the perspective a bit. A tiny implication may give more depth to the setting and thus make it more interesting. But is this really a meaningful choice? It certainly can be in some situations, but might not be in others, since the feeling and the theme of the game can be an important part of the gameplay experience.
The choice is only meaningful, of course, if you get to choose it once. It may not confer an advantage or disadvantage from a gameplay perspective, but it alters the experience, which in my opinion counts. A single different line may affect the rest of the game, if only inside the head of the player. (although it only works in the modern "choose one thing to say" dialogue style instead of the classic "ask what you want" style)

As for simplicity, I disagree here the most. The more massive the choice, the more meaningful it is. A choice between a few items can be either very important, or very pointless, depending on how the player's character(s) function, but a choice that affects a large segment of the game is always important, since it sort of is a large amount of smaller choices at the same time. To use your example, the foes you face in Shatterspeake are likely wildly different from Tutmanster as are the items you find or the quests you can do. Now, the choice is meaningful only if some parts of the game are still shared. For instance, if 20% of the game is the same before the choice, then you play 60% of the game in the chosen area, but the final 20% is still the same. It provides a wildly different experience and gives double the replayability. It also provindes a contrast between the areas. Of course, if only 5% of the game in the beginning is the same, but after that the areas never "cross back" to a single line, then you might as well make two different modules. Thus, while simplicity might be better from an efficiency standpoint(how much work you put to the game vs how much more enjoyable it is to the player), I believe that ultimately a game with simple choices is inferior to a game with complex choices. Realistically, of course, it's probably better to use your resources to make a handful of small but meaningful choices than a few big ones, but if you had infinite resources, I'd prefer the big ones.
The better question is, are several medium sized choices better than many smaller ones? Depending on the game in question, one might be the better idea than the other. Likely the best combination is a handful of medium choices and many small ones, rather than a huge amount of small choices.


We'll see how my ideas pan out when you release KOTC2 and I try my hand at making my own modules.

[screeg]

The "flavour" dialogue choices are definitely one of my top three irritations with RPGs old and new. Even indie RPGs touted as heavily C&C driven still pad out their content with this crap.

Your take on simplicity I think points directly to the type of game you're making. More complexity definitely makes for an interesting game (or it could, almost no one actual tries anything complex anyway), but KotC is heavily concerned with tactical combat. I like that fine. Your rules and combat are so strong I'm not concerned about story and moral choices.

[BlueSalamander]

a hidden choice can be very powerful

Good point. Thanks Tiavals. It's true that somewhat non-obvious choices can make a game better and more exciting. But the designer has to be careful not to hide these choices too well. I mean, you don't want to create great content that will only be explored by 20% of players.

There is an example in Dark Sun Shattered Lands. One of the evil wizards in the game is making a sword or dagger called the terror blade. You might learn about it the first time in a completely tangential way: when you receive quests from a merchant asking you to look for the rare components that are necessary to create the terror blade. Later, you meet the wizard and defeat him. The secret option here is that you can create the terror blade for yourself using the components that the wizard had gathered. In my first few playthroughs, I totally missed that option. I only learnt about it much later when I read the online walkthrough and official guidebook. I think there should have been more in-game clues about the possibility.

If the choice can only be made by a character with a high skill of a specific type

That reminds me of Age of Decadence. Many of the choices there require a high level in specific skills. But when a game allows the player to control a party of characters, you can expect that all skills will be available to the party, because at least one character will be strong, one will be intelligent, one will be charismatic, etc.

for instance that all your party members are Good aligned, or that none are Good aligned

Yes, that's a good idea. For example a particular church may impose that requirement on the party before allowing access to supplies.

you might pick a dialogue choice that adds a line or two to the dialogue

You are right, but I see those extra dialogue choices more like an enquiry used to get a bit more information about a particular topic. You are probing the NPC to see what he thinks about a topic and confirm a truth or falsehood. So even if the responses are purely flavour, at least the intent may be meaningful.

if 20% of the game is the same before the choice, then you play 60% of the game in the chosen area, but the final 20% is still the same

If it makes sense within the story to have that branching, then I say great. But people who play your module only once will miss 30% of the content (half of the branching 60%). While working on the branching part, are you going to be happy thinking that any player who has chosen the other option will miss all the content you're working on, unless they replay? Anyway, I agree that branching makes a game better, it's more a question of how big it should be.

[Tiavals]

Wow, didn't know that about the terror blade, pretty cool if you ask me.

I suppose your point about having all skills is true in a game with as many party members and as few skills as KOTC. But there are other "easy" exlucsions. Like something that can only be done by a centaur/thrikreen/half-giant or such, due to their physical difference from other races. Or something only druids can do due to their connection with nature, or only psionic characters because of their mind powers, or only divine character, etc, since a party might not include all of those things. Suppose you find a mind crystal in the lair of a psionic monster. You can sell it for coin with any party, but if you have a psionic character, it could be used as some sort of item only psionic characters can use, or such. Arbitrary perhaps, but still makes the play through for the player feel special. And that's what it really is about, that the player feels unique, that his experience of the game isn't "mundane" but something only he has experienced. It becomes especially worthwhile if the player himself finds out a hidden choice by his own actions. At least for me it's really rewarding, which is why I like Arcanum so much, it's full of stuff you can only do if you realise you can do it, or have a certain type of character, and so on. But then, arcanum is a decidedly single character(with NPC party) game, and KOTC is not.

I'm the sort of person who loves playing the same game many times, if I realise it is a different experience with each play through, so I much prefer a game that lets me miss 30% of the content(but only if it makes sense, not without reason!) in a single play through. And that's meaningful content too, which in my book probably means an actually altered story in some sense, at least. Which might not apply to KOTC since it's mostly based on the concept of dungeon crawling, and in a game like that, missing a dungeon without a proper reason is just poor design. If the reason is good, that's different, for instance if the dungeon you're missing is the plundering of a holy temple or the king's castle, it makes sense only if the story involves it in some way.
On the other hand, you could include the dungeon but alter the encounters. Like, if the player chooses, he can plunder a holy temple, but if not, the he must protect it from monsters that attack. The dungeon would be the same, but the encounters different. Instead of encountering priests, you might encounter orcs, and so on. Yet the secret passages and traps might still be the same.

Anyway, I love the idea of "user unique experience" in games, since it's unique to gaming as a medium. If you read a book, it's the same no matter who reads it, so is watching a movie, but for a game, the experience can be wildly different for different people and for different times you play through it. So it seems like a huge waste to make a game that doesn't take full advantage of that difference. I guess it's largely in regards to story, since I'm obsessed with variable stories too, and the ability of a player to actually affect a story. Whether a key character dies or not in a certain part of the story, in an event that the player can affect, that results in an alteration in the story itself; that is the reason I'm so passionate about this. To see the story unfold in several different ways is a truly fascinating thing to me. All a lot of work for the person making the game, of course, but in my book it's truly worth it. Which is why I hope to try making such a module for the game someday.

[BlueSalamander]

there are other "easy" exclusions. Like something that can only be done by a centaur/thrikreen/half-giant or such

Yes, sure. But I wouldn't want to get too great a benefit from having a particular class or race in the party (unless it was a NPC I recruited). Because then you would feel like the game is forcing you, or encouraging you, to play a particular party and not the party you want.

Or something only druids can do due to their connection with nature, or only psionic characters because of their mind powers, or only divine character

Yes I totally agree and I would like to have some extra dialogue choices based on the presence of certain classes in the party. A Rogue could be the only character allowed to spot a tiny button opening a secret cache. He could also be the only character being offered dialogue options to pick a lock, use stealth or pick someone's pockets. A Ranger or Druid could spot medicinal herbs. A Paladin could elicit more trust from nobles. The healing classes could be offered dialogue options to heal NPCs. The caster classes could be offered dialogue options to decipher ancient texts. A Monk could detect if a substance is poisonous just by catching a whiff of it. A Bard could charm friendly NPCs of the opposite gender. A psionic class could detect a lie. A warrior class could get options to force a door open, force a chest open or intimidate people into revealing information.

Those options would provide a minor benefit to the party as well as some experience points.

if you have a psionic character, it could be used as some sort of item only psionic characters can use, or such.

Sure. For example, you could have the mind crystal provide a bonus in maximum power points when worn on the head slot, or you could have it be a single-use item that replenishes the power-point pool, just like a potion would refill hit points. That way only a psionic character would get the benefit. Likewise, when the party finds an arcane scroll, only an arcane class can learn it or activate it.

Instead of encountering priests, you might encounter orcs

Yes, that's another interesting possibility. It would be controlled by the dialogue script where the player makes that branching decision. In the first case, all the orcs would be hidden and all the priests would be shown. In the second case, all the orcs would be shown and all the priests would be hidden by the script.

I love the idea of "user unique experience" in games

For me the uniqueness of the experience comes in great part from the choice of party members, during character creation and through the selection of recruitable NPCs. When I played Avernum 1, I created four characters who could all cast offensive and healing spells, even though they were useless in melee combat. I'm sure it would have been a very different experience if I had used a more traditional setup of two fighters, one healer and one offensive caster. The way combat plays out with your own characters is a big part of the uniqueness and appeal of the experience.

the ability of a player to actually affect a story

Yeah. Sounds interesting. Taking Baldur's Gate 2 as an example, it would have been cool if in every chapter, there was a branching choice given to the player, even if the two branches would always join again at the end of each chapter. In a more traditional dungeon-crawl adventure, it may be difficult to alter the story, but you can at least allow the player to choose between enemies, allies, rewards and quests.

Whether a key character dies or not in a certain part of the story

Sure, good idea. For example, the party has caught a major villain who is also a very important person for the city's economy. So, will the party choose to be merciful, and what will be the consequences. It will be great to see you make that kind of module someday.

[Tiavals]

Nice to see you agree with these parts.

The good thing about making a module instead of a game is that there is no financial risk involved for me, so there is no need to allocate resources efficiently. Thus, if I want to make all kinds of foolish design decisions and experiment with branching stuff or a changing story or gameworld, then I can just do it if I think it's cool.

So I must thank you for making the module editor for the game, for others to use. I can only hope plenty of other people will make modules for the game once it's out.

[deathknight1728]

The biggest problems with having choices in CRPGs is that ALMOST all of them dont have choices in the game.

The only ones I can think of are Fallout series, Baldurs Gate series and Planescape. None of the other games have it in there. 2 of those games have a party present, the others are solo.

A good example of choices for a solo crpg would be morrowind. If you dont have as much choices as that game, then you dont have choices at all. Fallout was good too. Eschalon was bad and the Quest was good too.

So really party based games dont have as much freedom with regards to that. I cant think of a dungeon crawler with any choices. Dungeon Crawlers while turnbased-are nothing more than an Action-rpg.

Jade Empire, while an action/rpg-has more choices than mostly any of the other turnbased partybased roleplaying games. Pool of Radiance, Curse of the Azure Bonds, Gateway to Savage Frontier and all of the SSi's are not roleplaying games. They are tactical-rpgs with some rpg elements. They are not rpgs.

[BlueSalamander]

Been playing Gabriel Knight 2 recently and it struck me how games can either have a detailed story, or player freedom. In that game everything is detailed, you have videos and voices for everything you do and everyone you talk with, and there's an interesting storyline. But at the same time, the player is on rails constantly. It's a bit like a film that would frequently stop to check whether you're paying attention. Then it would ask 'what do you think the hero should do now? what item should he use?'. And for every situation there would always be one and only one solution. If you don't find that solution, you're stuck.

And then on the other end of the spectrum you have games like ADOM, Bard's Tale 2, Daggerfall, Might and Magic 6, Diablo. Where the scenario is paper-thin, but you're never stuck because you can go anywhere, you can skip dungeons, you can fight some monsters but not others, you can choose what kind of character or party to create, etc.

Finally, you have games in the middle of the spectrum like Baldur's Gate 2, Dark Sun, Dragon Age where there's an attempt to provide a good storyline as well as some player freedom. For example in Baldur's Gate 2 the game is divided into chapters, just like in Gabriel Knight 2. That shows some focus on scenario. But the scenario is not as well developed as in an adventure game, and the player's level of freedom is not as much as in games with free-form design.

[Tiavals]

Something I've thought about recently regarding story and player freedom is that usually CRPGs can be categorized to two types.
One where the plot moves the player(modern bioware games like kotor, jade empire), and one where the player moves the plot(Dark Sun, Fallout, Baldur's gate). Some games are harder to pinpoint(oldschool games like Might and Magic or Wizardry, or the elder scroll games in general).

"The plot moves the player":
-Games of this category tend to be heavy on story and cutscenes, with a lot of dialogue that's usually mostly related to the plot.
-Areas are strictly defined and the game decides where the player can visit at a given part/time of the game. The game offers reasons why the player goes to these areas so the player doesn't need to think about anything.
-Traveling is very limited in these games, and the only way the player can have agency in the story is to just get to the hotspot that makes the game go forward. Usually this means talking to a certain character, which activates a cutscene that may lead to a battle, and so on. The player usually can't decide to attack non-hostile characters, which means many NPCs cannot be killed unless the game allows it.
-The player is generally given many strictly defined objectives he must do in a certain order to get forward in the plot, and thus the game.
-New areas open only when the plot has progressed far enough.
-Any given challenges are balanced so they're fair at all times, since the player cannot go to an area that is too hard for him.
-The game decides when you need to go to a place and will make sure you're powerful enough to handle it.
-Areas are usually related to the plot with only some self-contained elements. They usually have few connections to other areas outside themself, and only to areas that are "forward" compared to it, due to linear progression of areas.

"The player moves the plot":
-Games of this category tend to have a loose story with few cutscenes, with most of the plot happening in the background that you can hear from other characters. The amount of dialogue is usually very high, but it's mostly related to the setting of the game, rather than the plot.
-Areas are loosely defined and the game rarely limits where the player can visit, it's up to the player to choose where to go and when, and usually why.
-Traveling is very free in these games, and the player decides when to further the plot. The player can often attack and kill any NPC in the game, as they aren't critical to beating the game since it isn't shackled by a strict plot.
-Usually the player is given a broad goal that they follow through most of the game(find water chip, unite slaves to a coalition, etc). What makes these games interesting is that the player is given freedom to choose how he pursues this goal, when he does it and where he goes in order to do so.
-You can visit any place at any time in the game, but you must be prepared for the consequences. Wandering to a super mutant hideout is a bad idea if you aren't powerful enough.
-Combat challenges usually aren't balanced, because they are a result of the setting itself, not of the plot, and therefore it's pretty much impossible to gauge the power of the player and to balance the encounters. The super mutant hideout is a dangerous place because super mutants live there, not because the plot says there needs to be powerful enemies.
-It's up to the player to decide when to go to a dangerous place and how to prepare for it.
-Areas are usually self-contained with some links to other areas and some links to the main plot.


The choices a player can do in a "plot moves player" game are usually related to how the story unfolds, and usually fairly limited at that. Often there are only two outcomes for a given choice, but the outcomes can be very meaningful in how the rest of the game goes. The choices usually result in different dialogue, cut scenes, and so on, but often don't affect the gameplay experience that much(mostly due to the balanced nature of the encounters you face).

On the other hand, in the "player moves plot" game the choices are mostly relevant to the gameplay experience itself and are connected to the setting, with very little impact on the main story. Choosing to help raiders instead of merchants, or whether to destroy a drug lab or to use it for your own ends usually have an effect that's limited to the area the "quest" is related to. Cheaper items from merchants, more money or exp, the people will react to you more positively, and so on. Sometimes they can have far reaching effects(Dark Sun does this pretty well, I think, the final battle can be quite different depending on what you do), but usually the effects are immediate.


For a party based game, I think it's very difficult to use the "plot moves player" type of game while still retaining the point of the party. After all, the advantages of the "plot moves player" type of game is that it can have plenty of strong, memorable characters that the player constantly interacts with, usually in the form of a party or otherwise recurring characters. In a party-rpg, it'd be strange since you already have plenty of characters, ones that you have designed. Of course, you could let the player design the class/mechanics of a story-type character(as is often done in japanese crpgs), but that only corrects one problem. Or you could give the player the choice of a "personality type" for the characters in their own party to make them more interesting, Wizardry 8 sort of does this, which is one of the more fascinating parts in the game.(it only affects dialogue though, but it's still very interesting).

All in all it's a very important thing to decide what sort of story you want the game to have, since a strong story usually limits the freedom of exploration in the game a great deal. Of course, if a game has a ruleset like KOTC, that already forces the scales toward the strong story, since the level of the characters means a huge amount in a d20 system(as opposed to Fallout where the ruleset is much freer and less combat oriented). This was pretty apparent in KOTC, with it's clear progression of areas based on encounter level. First hill giants, then frost giants, then fire giants. Combat is a big part of KOTC, so it has to be fair, and if you overlevel your enemies by going through the areas in a different manner than assumed, they'd be so easy it probably wouldn't be much fun. It's sort of strange, since a party-rpg is naturally more geared toward the free type of exploration, yet the mechanics of the ruleset force it to the "wrong" direction. I'm sure there are many things that could be done to avert that, of course, like the often hated "scaling enemies" system, but it's hard work to design combat encounters that can be challenging to both a level 5 party and a level 9 party. It worked in Dark Sun because the characters start at level 3 and end the game at level 9, so the range of levels is pretty small all in all. Besides, AD&D is a pretty different system than d20 anyway.