This is my thoughts on one of the ways to implement nonlinear storyline in a single player RPG game.

There are many times in the RPG storyline that seems to happen JUST as the hero arrives, usually right before the final boss or some sort destroys the whole world (I think it's most prominent in JPRGs). And especially in linear storyline games, it's unavoidable because the story would not be able to continue without presenting such event to the player.

Now, if the game was nonlinear, wouldn't it be possible to implement an in-game time, where it is possible to miss such an event, and instead is given alternative event and has different ending based on that? Such in-game time can be determined by different actions that you do. Let's say you decide to camp out in the night to rest and recover HP and MP. However, camping and sleeping takes 7 hours which might stop you from getting to the next city that has an event in time. So player X decides to not to camp and keep moving on through the night to the city, and face the consequence of not having full HP and MP for a possible fight. However, player Y decides to be safe and camp. While camping, a group of bandits decide to carry you away to be sold as a slave, and you wake up in bandit's hideout and have to fight them.

Other things such as talking to an NPC or just walking around in general could take certain amount of time, and also there could be an ending where if the player decided to not do anything and just walk around and sleep, you could have an ending where the world just gets destroyed.

So this way, the storyline branches out, and decide what kind of ending you will have. Another thing that this Time factor can affect the game is that this sort of game would have a huge replayability value. You can set up different achievements that allows the player to try things in different ways (one achievement is to get to the event in the city in time. another achievement is to be kidnapped by bandits. So no character can achieve both achievements in same game).

It would be a realistic feature that's actually fun, and keep the player involved and make them think about the next move because now they can't just take the whole eternity, going from one city to another chatting and idling, because the quests are not going to wait for them.

I may be just presenting something that's already being widely used (because I haven't played wide range of RPGs yet) but the games that I have played so far didn't really have this sort of approach.

It also lets you put weight on spending time to progress your character. Crafting/making an item will take time (in which you could be doing something else, or pay someone to make it for y ou). Learning a new spell/skill would take time.

You'd want to give some leave way in the exact time that events occur in. Otherwise, it will be almost impossible to see any of them. You'd also want to give the player some knowledge/idea of when events will occur. Reports of enemy army movements would let them know about when they would have to arrive to help defend a location. Also if the player is in a location where an event is going to happen and decides to rest/sleep/wait, interrupt that when the event happens instead of skipping over it (and give an option to wait for it, when the player knows in advance when something is going to happen).

The main problem though, is the sheer amount of additional content you have to make. In a more linear game, if the city isn't sieged until you arrive to save it, then, you never need to create the destroyed/conquered version of it.

This will also easily lead to an "unwinnable" game, as odds are, the world will be destroyed without you. Course, for a shorter game, this wouldn't be that bad, especially if you want it to be replayable (you'll have a better chance of that, if restarting the game to try a different sequence of events isn't that huge of a time investment). You'd probably also be better served by a less "epic" scope than saving the world from destruction, where the consequences for not arriving aren't quite as dire.

You'd want to give some leave way in the exact time that events occur in. Otherwise, it will be almost impossible to see any of them. You'd also want to give the player some knowledge/idea of when events will occur. Reports of enemy army movements would let them know about when they would have to arrive to help defend a location. Also if the player is in a location where an event is going to happen and decides to rest/sleep/wait, interrupt that when the event happens instead of skipping over it (and give an option to wait for it, when the player knows in advance when something is going to happen).

That's what I have in mind. What would usually happen is that someone will most likely tell you to get to X under 3 days, so you have a set goal that you can either choose to follow or not. And if you're in general area, yes the event should happen. It can be made fluid by including a small cutscene or something where the character is 'woken' up by a loud noise and goes to investigate.

This will also easily lead to an "unwinnable" game, as odds are, the world will be destroyed without you. Course, for a shorter game, this wouldn't be that bad, especially if you want it to be replayable (you'll have a better chance of that, if restarting the game to try a different sequence of events isn't that huge of a time investment). You'd probably also be better served by a less "epic" scope than saving the world from destruction, where the consequences for not arriving aren't quite as dire.

Well I was thinking, at least for the game that I'm developing, it would take around 15~20 hours of playtime to get through to the ending, and there are 6~7 different endings. What would happen is that there are set events and the time it happens for all areas, and sequence the events that way. Another thing I was thinking of doing, is that for really crucial event, players are naturally 'led' to such an event. Using the bandit example, if you're not in an area that a crucial event is going to happen, the bandits kidnap you and bring you to the area that the crucial event is going to happen, because that area happens to be where the bandits hideout is.

It's a good idea, but I think you need a little AI Director stuff going on to make it fun. Imagine if the player wanders around aimlessly, manages to miss every major event in the game, and the game ends. Ouch, no fun. I think it's better to force an alternate event if they miss the main one, and I guess the bandit kidnapping could be such as case.

Do you think achievement system is good for directing players to actually know that certain event is there? The achievements will be shown (not hidden) so players will know under what condition you can get the achievement, and so try doing certain things that they might not have thought of doing. Other than achievement how else would you tell the players about different path choices? I guess NPC conversations could work, if written in good way.

And I think it would be fun to allow the player to purposely miss all the major events and purposely lose, and that could also be an achievement, adding to the replay value and something the player wants to achieve with one of his/her game. Game design wise, I would try to make sure that the players are not accidentally missing the major events, by hinting them about the events through NPCs, quests, etc.

I think manipulating the player would be the best way to direct them towards major plot events without directly forcing them down a particular path. Otherwise they might feel cheated out of their "open world" experience.

Take advantage of their curiosity - Smoke/fire off in the distance, explosions, a mysterious door... All sorts of effects you could use to draw the player in without making them feel like they were forced into it. Any good adventurer will have no choice but to explore to the top of the magical tower, or to the bottom of the deepest dungeon. Make the locations obvious through level design and drop subtle hints here and there to hurry them along if needed

Or take advantage of their emotions - Evil bandits burning down or stealing things from the house the player spent hours building up - they'll want revenge and charge into the bandit base. Or maybe have the player's favorite NPC be kidnapped and they have x amount of time to save them.

I think manipulating the player would be the best way to direct them towards major plot events without directly forcing them down a particular path. Otherwise they might feel cheated out of their "open world" experience.

Take advantage of their curiosity - Smoke/fire off in the distance, explosions, a mysterious door... All sorts of effects you could use to draw the player in without making them feel like they were forced into it. Any good adventurer will have no choice but to explore to the top of the magical tower, or to the bottom of the deepest dungeon. Make the locations obvious through level design and drop subtle hints here and there to hurry them along if needed

Or take advantage of their emotions - Evil bandits burning down or stealing things from the house the player spent hours building up - they'll want revenge and charge into the bandit base. Or maybe have the player's favorite NPC be kidnapped and they have x amount of time to save them.

Those are good incentives to drive the players into a major plot. But what happens if few minority players decide to just abandon the house they built? Just ignore the kidnapped NPC? If I give a goal to a player, but since the game is supposed to be 'open world' experience, they would feel cheated of their open worldness(?) if they can't just ignore those goals. That's why I thought I should include the 'game lost' ending.

Clearly there should be consequences to their inaction. Ranging from game ending if they let the bad guys win, to inconvenient if they let the bad guys defeat their allies, leaving the hero to face the enemies on their own - it all depends on how important that plot event was.

I think an outright "game lost" ending might leave a bad taste especially if the players don't perceive the plot event as having been so critical the world should end, but endings where a player can clearly think to themselves "Well... I dug myself into a hole so deep there's no way I could climb out" would be acceptable.

As an example of what isn't so fun - if you've played the Grand Theft Auto games (Which if you haven't, and you're discussing open world games... Well. That would just be weird), but they always have these segments which if they aren't completed successfully, its essentially game over. This isn't bad in and of its self, but it is annoying when the writers haven't sufficiently justified why failing that event is game ending. Its all a matter of compromise of course since you do have to tell your story, and you can't neatly lure the player into your plots every time.

A fairly common plot hook that they've used in most of the newer GTA games - you have your partner NPC who helps you out throughout the game. At some point they'll get kidnapped, caught, or in some other form of trouble and you have to save them. Failure to do so requires you to reload your save games until you're successful. What always bothered me is why do I care about that character? The writers never particularly established that character as being necessarily. Hence no desire to save them beyond the game taking away my open world toys until I cooperate.

On the other hand - if you setup the kidnapped NPC as being the tank to my mage, or a romantic partner to the hero in a love story, or having some unique special ability... Then I want to save them because squishy mages need their tank friend, or a love story isn't complete without both partners, or only my dragon friend can fly over the chasm. At that point if I have to reload from a previous save game, at least I can accept that "Well. There was only one dragon... And I went and got him killed".

I think the key thing is to always present the player a choice, even if that choice is between saving the world from the alien invaders or running off to go explore a cave while the aliens take over and destroy everything and eventually the hero too.

Yes I agree with you wholly. I think it all comes down to if the story and the game design has been made that the reasons behind the "game lost" is made clear to the players, with good supporting evidences. What could happen is that not long after the major supporting character that needs to be alive for major plot advance dies due to player's neglect, there can be an event that is pretty much impossible to win without the dead supporting character. And so now, player can reload the game knowing that they 'have' to save that character to win.
Yes, it would be a heavy burden on the writer's part, but when you look at those Interactive Fictions, they handle all these very nicely. And I think the same thing can be carried over to the graphical RPGs as well with some efforts.

And the main thing that my game would differ from other open-world/sandbox RPGs such as skyrim, GTA, and others like that, are the fact that since there is time limit, you will never get to finish all the quests in the game. Yes, skyrim also limits you in some way because choosing either Imperials or Stormcloaks limit your experience of the other faction's quests (which actually are very similar in design, so you end up not missing out too much), but all the major quests are usually independent of each other. For mine, player will be given multiple quests at a time, and each have time limit. If you complete one quest, you might not be able to complete other quests, so the player has to choose which quest they think is important to them, and game progresses that way.
This works at least for my game, because my plan is to make each game relatively short (around 15~20 hours), so it won't be a huge loss on the players to recreate new characters; in fact the path I'm taking encourages players to recreate new characters and play different way, giving them new experience each time.

I also thought of implementing a sort of endless arena/survival battles for those people who want to keep playing with their maxed out character, so that the characters don't go to a waste. The way I see it, it would be fun to make different kinds of characters and see which character can survive the most waves/battles. (Character building is similar to skyrim, where you begin without any specific "class", but can choose which kind of skills you want to pursue, and mix and match different skills. Also, I want to make a crafting system similar to diablo 3 with random stat affixes, so there's even more customization).

there can be an event that is pretty much impossible to win without the dead supporting character. And so now, player can reload the game knowing that they 'have' to save that character to win

No, no, no. This is not how it should be done.

If something goes wrong, like an "essential" character dies, the story should change not end. If the rightful heir to the kingdom dies, I don't expect a Game Over screen, I expect a twist where there is another heir rumoured to live elsewhere. If he dies, maybe I shall claim the throne myself ect ect.