Since Notch was tweeting on the subject of linear games today, I decided it was as good a topic as any to address today.
Linearity in a game is not nessecarily a bad thing. However, it has to be handled well. A perfectly linear game gives the designer a high degree of control, and if utilized well can have some really interesting results. The advantage is that you know precisely where the player will go, what resources they have accessed, and can utilize that to your advantage. You can throw in scripted events as you wish, and know when they will triggered. This offers a lot of power from a storytelling perspective. Challenges can be specifically crafted, and its a much more straightforward experience.
A Mostly linear game shares a lot of properties with the completly linear game. The main difference is there is some minor branching. The ability to take a side path and discover a bonus area, for instance. your course through the game is still pre-determined, but you have some freedom you can turn to your advantage. This allows some elements of exploration while still providing a rigid structure for the designer to work from.
A branched game offers actual choices about how you get through the level. you could take the high road, or the low road. Both offer a fresh experience, and they can offer different choices to make on different play-throughs. Sonic (the genesis ones) was a branched game. You very frequently had multiple ways you could get through a level. Taking a high path could drop you to a lower path if you messed up. There were hidden rooms you could only access from certain paths, paths that required more skill but had greater rewards. This can create a more interesting experience, even though the overall game flow is still linear. Repeated playthroughs will see different choices.
A multi-linear game is a large step forward. A multilinear game has many linear paths through it, that intertwine, allowing you to move between them. People often consider this an open-world, even though there are specific paths you will be following through the level. Deux ex is multilinear. People think it is open, becuase you can approach the level in so many different ways. You could sneak through the center of the compound, you could climb up to the roof, you could take the vent, you could find the side entrance, etc. However, each of these approaches was specifically added by the designers. You have lots of choices, but they are from a specific set of options. The designer specifically put the code to the door in that gaurds pocket so if you kock him out, you can access the door. Certain paths are locked if you do not have the proper upgrades. Paths very frequently split and merge, with multiple ways to access or leave it.
Such a high degree of freedom is great, while the underlying network of potential paths still allows the designer to engineer a satisfying path through it. Sidepaths that are harder to access can lead to bonus areas, or skirt aroudn danger. Paths can lead off to get something to open up a new path, such as a key card. Different paths could be more supportive of different approaches, so one path is good if you want to just kill everyone, another isa good choice for stealth. Not every path needs to be immediately obvious; discovering new paths is a rewarding experience. On subsequent play-throughs, you can make entirely different choices on how to approach it, and get a very different expereince out of it.
A completly open world has no pre-defined paths. Minecraft is a great example. You can go anywhere you want, and get there is whatever way you please. A truly open game is rare. There is complete freedom to the player to go wheerever they feel like, making it very difficult as a designer to create a flow. It is possible, at least in a borad sense. Minecraft accomplishes this with resources. The progresssion from wood to diamond, then through the nether, potions, and enchantments, and culminating in the end gives the game a distinct flow, whilst leaving the player with their freedom.
Most "open world" games do not truly have a pure, open world. They really have a hybrid setup. Certain areas are open, with some pre-defined paths between them. Deux ex is really a hybrid. There are open-world areas connected and interwoven into the multilinear web. How you approach the open sections is free, but you still end up using it as a node to reach other paths. Also popular is the "pearl necklace" model, where you have discrete areas that are open, with specific transitions between them. This retains an overall linear structure of the game to hang your plot on, while offering the player moment to moment freedom. Portal(both 1 and 2) is a pearl necklace model. Each individual test chamber is open, and you have great freedom. But you are still moving from one to another in a linear fashion.
Then you have games with a disparity between the world design and the plot design. The world may nominally be open, but the places you need to go for plot are still strung out in a linear fashion. Sometimes the plot itself is branching, or has a loose ordering. A loose ordering is more common. That is where you have a batch of missions you could do, but the order you do them is not important. These missions can have their own plot chains, allowing the plot to have multiple facets running in parrallel. They can all converge on important plot points, dividing the structure up into acts, which have their own internal mission structure. This is similar to the pearl necklace model, in that you have specific choke points between sections that you must pass through, while still offering freedom within the nodes.
Which of these models is most appropriate depends on the game. There is no singular best answer, it depends on what you want to acheive. However, the advantage a game has over other media is the interactivity. Interaction is about making choices and exterting control. More freedom is a good thing in general, assuming you know what to do with it.
Another thing to consider is that video games are a different medium than books or movies. As such, how you tell a story with a video game is different that with a book, just as a book is different from a movie, and a movie is different from a webcomic. Different mediums have different properties. Video games have the extra element of player control, and that can be harnessed. They don't have to be linear, which can also be utilized. The plot of a game is not like the plot of a movie. It does not have to unfold in perfect sequence. It does not have to be a static thing simply presented to the player. The player should be able to live the story. You can use the gameplay to reinforce the plot. The story can be interactive, engaging the player to participate in it. The player's choices can be woven into the story. You cna have a much looser sructure. Going to the pearl necklace model of plot I mentioned earlier, you can have multiple subplots going at once, which merge and interact into the overall story.
Jak II followed a complex model for this. There is a central, open-world hub you are free to explore and utilize. it is connected to other level areas, each of which has their own structure. Some are very linear, and you simply progress through them. Others are more open, with multiple paths to explore, but the order is not important. There are often hidden areas you can find and utilize. Some levels are open, and let you tackle them as you will. The plot operates in distinct acts, but within the acts you can follow serveral plot treads in whatever order you want. You are presented with several possible missions, and are free to tackle them in whichever order you choose. This makes it feel much more like you are in a living world, taking place in a setting, instead of simply following a pre-scripted plot.
The key is to create a rich enviroment that serves your needs. Deux ex needed multilinear levels, offering the player choices on how to approach it, opening up paths based on the upgrades they chose, letting them choose how stealthyor unstealth) they wanted to be. Uncharted 2 offers a very linear game flow, with hidden pockets to explore, and uses it to create a very fine-tuned, designed experience that worked wonderfully. It depends on what your game needs to accomplish.
Shadow of the Colossus is an interesting thing. The entire gameworld (save for some blocked-off areas) can be explored from the beginning. Riding around and taking in the environment is half the game. But there is always only one thing to do in the world, going to a specific point and fighting the colossus. There is no way to change the order in which you fight them. So in effect it's an open world game where the player can't branch the scenario.
Shadow of the colossus is indeed interesting. The battles themselves are great, very varied and interesting, and defintely give you the sense of fighting something that is mind-boggling huge. That aspect of the game was great.
The open world was kinda pointless. Its not that there is no control over the order you fight the colossi. They had a leveling scheme and a scaling difficulty,so allowing you to face them in an arbitrary order would not work. Its that there isn't really anything to do. You can ride around in it, and it looks pretty. Thats pretty much it. Sure, you could catch lizards or something pointless like that, but it was mainly open space. It served as a nice spacer and to give you a sense that all of this is in a world, and hence there is a context to these fights. From an artistic standpoint, it may have served a purpose, but from a game design standpoint it was simply travel time.
Greetings, nice to meet ya! if i helped you in any kind of way be sure to press the lil' arrow button there, and you will help me too! that's cool isn't it???
Glad you like them. I will continue to make posts like these. I am open for which topic I discuss, if you have anything in particular you would like to read about.
Fallout 3/NV is pretty interesting. They never force you to do anything, in fact, you could screw around and kill stuff until you ran out of ammo or died, but they do lay out a few definite paths you can follow, but you don't have to. In fact, there are a huge number of ways you can go about beating the game. There only "linear" parts are quests themselves, and even then, there are a huge number of ways you can go about doing the task. Some they tell you about, but some are more hidden. Like when you have to find the igniting agent for the dudes at repconn in New Vegas, you could either go to Smith's field, like they told you, or you can go to Clif Brusco's shop in Novac, and steal/buy the model rockets from them, that happen to contain the igniting agent, which was never mentioned, except in guides, or if you happened to have the rockets on you when you were assigned the task.
Interesting topics... but I'm not sure how much you leave open for discussion. :smile.gif:
But on the topic of linearity in games... I personally prefer the more open-ended games.
I think open-endedness allows a game to be good. The more open it is, the better it can be.
But there's a large difference between "open world" and "open gameplay". An open world game, say for example Grand Theft Auto IV, can be quite shallow in actual gameplay options. In the GTA 3 series, you could buy properties, customize your cars, do tons of side missions, and etc.... In GTA IV, they took all that and opted instead for a better game engine, and essentially better graphics. There is still a some things you can do... but simply just not as many as the previous games in the series.
Does this mean it was a worse game? No, not objectively. It was simply just "different"... not what was expected from the series which typically added MORE and MORE options, rather than taking away options.
Did this mean it was a worse game in my opinion? Yes, absolutely. I did not have as much fun with GTA 4 as I did with the entire GTA 3 series.
And yet, all games really are ultimately linear in some respects... Even entirely open-ended games like Minecraft have linearity in them. And you do a good of pointing this out, but I want to emphasize how 'PLAYER OPTIONS' are what create the linearity and open-ness of a game.
Even in a game like Minecraft, linearity can be seen in that you have the that all you really do is begin the game with low-level tools, mine, get better tools, mine more, and then get the best tools... Which you then typically will use to mine a lot of what blocks you want so you can make cool structures. And in the midst of it all, you fight bad guys, too.... But that's all you ever do.
It's linear in that there are not any options other than 'Mine - Build - Fight'... The open-ness of the game only comes in with the fact that you can 'Build' nearly anything imaginable.
I think these player options are what truly make the open-ness of a game... and as well as what makes a game "fun", at least for me.
I don't really have much "fun" with entirely linear games... What's the point? You run through once, then you're DONE. The end.
I like my games to have replay value, personally.
As for game design in general, as I said already, I love the more open-ended games. The ones with more options, the better they can be.
And personally, all of my plans for game design try to encompass and put emphasis on these gameplay-options, and as many of them as possible.
And while I won't say what I'm working on right now, I give most of my thought into 'how a player might want to play it'. I try to think, as a player, what options I would want to see in the game... I try to use my imagination to think up even outrageous stuff like "What if I could just pick up that chair and hit someone with it?" and while a lot of that is unlikely to be practical in a game, there are still ways to implement the "fun" of such ideas, while not taking too much time on actually implementing the details of it.
An example is a game idea I've had for a long time where it is a traditional-looking "RPG" game, yet the player has the ability to open up, or even destroy, towns all over the game-world. And to even determine how those towns are run, or if you just want to start massacring a village because it's fun. Those are things the player imagines they want to do when they play an RPG... so why not let them do it!? Of course, that game idea will have to wait for a while, when I hopefully get a little bit of income to work on games like that...
I think, for example, the 'Fable' series did this relatively well at the time, at first... You can basically, slightly determine the outcome of certain events, and your actions have real repercussions on the game-world. You could even get married to random villagers, or kill them. You could wander around being a trader, trying to make profit, or you could go around being the 'good guy' or the 'bad guy'. It was your CHOICE... because the games gave you those options.
Although I feel the latest, Fable 3, discouraged this point of view and did not stay true to the roots from which the first two games were designed. I thought Fable 3 basically steered you directly down a single path, and it was pretty annoying... it didn't really give you many options... Kind of like GTA 4. They replaced all the game options that made the game good, and replaced it with better graphics (which weren't really that much better, in Fable 3's case).
Anyway, just thought I'd throw in a few of my own thoughts about linearity, because it's a topic I'm very passionate about in gaming and game design.
Interesting topics... but I'm not sure how much you leave open for discussion. :smile.gif:
<snip>
Anyway, just thought I'd throw in a few of my own thoughts about linearity, because it's a topic I'm very passionate about in gaming and game design.
I agree, the more open-ended designs tend to make more interesting games. But being linear isn't nessecarily a bad thing, if done well. Thinking about it further, uncharted 2 is a great example of a linear game done well. It is designed to feel like you are in a movie. The plot if very fixed, many sequences are heavily scripted(this is apparant the first time you fail to climb the train dangling from the cliff, and everything breaks and slides as you ge to the exact same parts). This works for them, because they use it to finely tune the expereince. They can engineer the fight that will occur while you are hanging from a street sign. They can script the train that is sliding off the cliff to make it feel urgent and that you just barely made it out alive. In that case, the linearity allows them to craft the experience very precisely, which worked out remarkably well.
In generaly though, I find that more player choice is a good think. My game designs tend to revolve around it. For instance, the game I am working on now will allow you to customize every cybernetic component you have, so you can aechieve a side variety fo playstyles. Another design that is on the back burner features a extremly open-ended magic system to allow you to create whatever spells you want. With the phrase "Go anywhere, do anything", the "do anything" half is the more important part of gameplay. That is what is so great about minecraft. Yes, you can go anywhere; people generally don't. There is no point. They do cover the "do anything" part extremely well. The sheer variety of things people accomplish is great.
Open-endedness in games is great. But it is not the answer for every game, and it is a choice that requires serious consideration about what you want to acheive with the game, not simply picking the singular right answer.
Yeah, there's definitely a market even for 'movie-games' that are completely linear.
I don't think linearity is bad at all... but personally, it's not the type of game that I like.
I'd personally rather go to the movies, because that costs less than video games, gives me the same linear-experience, and is also fun.
But even 'movie-games' can have some non-linearity to them, like Heavy Rain, which is actually done quite well. It is a cinematic game, yet it allows the player the ability to choose how the movie turns out, and what the characters do and don't do.
I'd agree that there is no "right answer" in overall game design.
Although, there is definitely a "right answer" for me and the games that I like. But ignoring my personal opinion, there is a market out there for practically anything... and if you do something right, you'll hit at least one of those markets.
Yeah, there's definitely a market even for 'movie-games' that are completely linear.
I don't think linearity is bad at all... but personally, it's not the type of game that I like.
I'd personally rather go to the movies, because that costs less than video games, gives me the same linear-experience, and is also fun.
But even 'movie-games' can have some non-linearity to them, like Heavy Rain, which is actually done quite well. It is a cinematic game, yet it allows the player the ability to choose how the movie turns out, and what the characters do and don't do.
I'd agree that there is no "right answer" in overall game design.
Although, there is definitely a "right answer" for me and the games that I like. But ignoring my personal opinion, there is a market out there for practically anything... and if you do something right, you'll hit at least one of those markets.
Yes, personal taste is a big factor in what games someone will like. Even with a flawless design, not everybody will like a game. There are people who don't like portal, because it is not their type of game. Though a solid game design is more likely to transcend people's normal prefrences. Personally, I don't care much for first person shooters in general. Nothing wrong with it as a genre(though I could argue against many instances of the genre), just not what I prefer. Yet I love the half life series. It doesn't do anything to transcend the genre, but it is so solidly designed I love it anyways.
Its also harder to design an open-ended game. Its not so much that its hard to create those options, but its harder to design a game around a body of choices. With half-life, they know you will have a certain set of weapons. The entire game situation is known, and they can design it with that in mind. When you add options, you need to make it so those options are meaningful and impact waht you are doiung, otherwise they are pointless. Yet you also need the options to be equally viable, otherwise its not an option so much a test as to the right choice to make. And while not every combination of options needs to be, nor should be, equal, there should be a clearly defined power level that good choices will result in. No one combinaton should be notably stronger than the others, nor shouldany intelligently designed build be particuarly poor. Of course, it will frequently be possible to make bad choices that result in poor builds, but if you are training in archery and weilding a sword, that is your own fault. However, training to be an archer and training to be a swordsman should both be equally viable. optimising your choices is an interesting aspect of choice, and needs to be considered by the designers. Optimization should be possible. However, optimized should mean effective, not game-breaking. Its also helpful to include a method to recover from poor choices. Otherwise you make the wrong choice, and kick yourself about it for the rest of the game. You also have to consider how different builds will function in every part of the game. It is perfectly fine for certain builds to excell in certain situations and have a hard time with others. As long as "hard time" doesn't mean "Can't get past". Don't mandate that the character must wrestle a lion. The big, hulking characters may be able to tie it into knots, but the scrawny archer is goign to be cat-food. All of this is possible to design around, but it is significantly harder than designing a linear game.
Since Notch was tweeting on the subject of linear games today, I decided it was as good a topic as any to address today.
Linearity in a game is not nessecarily a bad thing. However, it has to be handled well. A perfectly linear game gives the designer a high degree of control, and if utilized well can have some really interesting results. The advantage is that you know precisely where the player will go, what resources they have accessed, and can utilize that to your advantage. You can throw in scripted events as you wish, and know when they will triggered. This offers a lot of power from a storytelling perspective. Challenges can be specifically crafted, and its a much more straightforward experience.
A Mostly linear game shares a lot of properties with the completly linear game. The main difference is there is some minor branching. The ability to take a side path and discover a bonus area, for instance. your course through the game is still pre-determined, but you have some freedom you can turn to your advantage. This allows some elements of exploration while still providing a rigid structure for the designer to work from.
A branched game offers actual choices about how you get through the level. you could take the high road, or the low road. Both offer a fresh experience, and they can offer different choices to make on different play-throughs. Sonic (the genesis ones) was a branched game. You very frequently had multiple ways you could get through a level. Taking a high path could drop you to a lower path if you messed up. There were hidden rooms you could only access from certain paths, paths that required more skill but had greater rewards. This can create a more interesting experience, even though the overall game flow is still linear. Repeated playthroughs will see different choices.
A multi-linear game is a large step forward. A multilinear game has many linear paths through it, that intertwine, allowing you to move between them. People often consider this an open-world, even though there are specific paths you will be following through the level. Deux ex is multilinear. People think it is open, becuase you can approach the level in so many different ways. You could sneak through the center of the compound, you could climb up to the roof, you could take the vent, you could find the side entrance, etc. However, each of these approaches was specifically added by the designers. You have lots of choices, but they are from a specific set of options. The designer specifically put the code to the door in that gaurds pocket so if you kock him out, you can access the door. Certain paths are locked if you do not have the proper upgrades. Paths very frequently split and merge, with multiple ways to access or leave it.
Such a high degree of freedom is great, while the underlying network of potential paths still allows the designer to engineer a satisfying path through it. Sidepaths that are harder to access can lead to bonus areas, or skirt aroudn danger. Paths can lead off to get something to open up a new path, such as a key card. Different paths could be more supportive of different approaches, so one path is good if you want to just kill everyone, another isa good choice for stealth. Not every path needs to be immediately obvious; discovering new paths is a rewarding experience. On subsequent play-throughs, you can make entirely different choices on how to approach it, and get a very different expereince out of it.
A completly open world has no pre-defined paths. Minecraft is a great example. You can go anywhere you want, and get there is whatever way you please. A truly open game is rare. There is complete freedom to the player to go wheerever they feel like, making it very difficult as a designer to create a flow. It is possible, at least in a borad sense. Minecraft accomplishes this with resources. The progresssion from wood to diamond, then through the nether, potions, and enchantments, and culminating in the end gives the game a distinct flow, whilst leaving the player with their freedom.
Most "open world" games do not truly have a pure, open world. They really have a hybrid setup. Certain areas are open, with some pre-defined paths between them. Deux ex is really a hybrid. There are open-world areas connected and interwoven into the multilinear web. How you approach the open sections is free, but you still end up using it as a node to reach other paths. Also popular is the "pearl necklace" model, where you have discrete areas that are open, with specific transitions between them. This retains an overall linear structure of the game to hang your plot on, while offering the player moment to moment freedom. Portal(both 1 and 2) is a pearl necklace model. Each individual test chamber is open, and you have great freedom. But you are still moving from one to another in a linear fashion.
Then you have games with a disparity between the world design and the plot design. The world may nominally be open, but the places you need to go for plot are still strung out in a linear fashion. Sometimes the plot itself is branching, or has a loose ordering. A loose ordering is more common. That is where you have a batch of missions you could do, but the order you do them is not important. These missions can have their own plot chains, allowing the plot to have multiple facets running in parrallel. They can all converge on important plot points, dividing the structure up into acts, which have their own internal mission structure. This is similar to the pearl necklace model, in that you have specific choke points between sections that you must pass through, while still offering freedom within the nodes.
Which of these models is most appropriate depends on the game. There is no singular best answer, it depends on what you want to acheive. However, the advantage a game has over other media is the interactivity. Interaction is about making choices and exterting control. More freedom is a good thing in general, assuming you know what to do with it.
Another thing to consider is that video games are a different medium than books or movies. As such, how you tell a story with a video game is different that with a book, just as a book is different from a movie, and a movie is different from a webcomic. Different mediums have different properties. Video games have the extra element of player control, and that can be harnessed. They don't have to be linear, which can also be utilized. The plot of a game is not like the plot of a movie. It does not have to unfold in perfect sequence. It does not have to be a static thing simply presented to the player. The player should be able to live the story. You can use the gameplay to reinforce the plot. The story can be interactive, engaging the player to participate in it. The player's choices can be woven into the story. You cna have a much looser sructure. Going to the pearl necklace model of plot I mentioned earlier, you can have multiple subplots going at once, which merge and interact into the overall story.
Jak II followed a complex model for this. There is a central, open-world hub you are free to explore and utilize. it is connected to other level areas, each of which has their own structure. Some are very linear, and you simply progress through them. Others are more open, with multiple paths to explore, but the order is not important. There are often hidden areas you can find and utilize. Some levels are open, and let you tackle them as you will. The plot operates in distinct acts, but within the acts you can follow serveral plot treads in whatever order you want. You are presented with several possible missions, and are free to tackle them in whichever order you choose. This makes it feel much more like you are in a living world, taking place in a setting, instead of simply following a pre-scripted plot.
The key is to create a rich enviroment that serves your needs. Deux ex needed multilinear levels, offering the player choices on how to approach it, opening up paths based on the upgrades they chose, letting them choose how stealthyor unstealth) they wanted to be. Uncharted 2 offers a very linear game flow, with hidden pockets to explore, and uses it to create a very fine-tuned, designed experience that worked wonderfully. It depends on what your game needs to accomplish.
Game Design Theory 3- Levelign and grinding
At first I was like this is toooo long, but damn you know how to make a post interesting.
I'll go check out the other one now.
Live with the Creed.
Shadow of the colossus is indeed interesting. The battles themselves are great, very varied and interesting, and defintely give you the sense of fighting something that is mind-boggling huge. That aspect of the game was great.
The open world was kinda pointless. Its not that there is no control over the order you fight the colossi. They had a leveling scheme and a scaling difficulty,so allowing you to face them in an arbitrary order would not work. Its that there isn't really anything to do. You can ride around in it, and it looks pretty. Thats pretty much it. Sure, you could catch lizards or something pointless like that, but it was mainly open space. It served as a nice spacer and to give you a sense that all of this is in a world, and hence there is a context to these fights. From an artistic standpoint, it may have served a purpose, but from a game design standpoint it was simply travel time.
we need MOAR!
Glad you like them. I will continue to make posts like these. I am open for which topic I discuss, if you have anything in particular you would like to read about.
But on the topic of linearity in games... I personally prefer the more open-ended games.
I think open-endedness allows a game to be good. The more open it is, the better it can be.
But there's a large difference between "open world" and "open gameplay". An open world game, say for example Grand Theft Auto IV, can be quite shallow in actual gameplay options. In the GTA 3 series, you could buy properties, customize your cars, do tons of side missions, and etc.... In GTA IV, they took all that and opted instead for a better game engine, and essentially better graphics. There is still a some things you can do... but simply just not as many as the previous games in the series.
Does this mean it was a worse game? No, not objectively. It was simply just "different"... not what was expected from the series which typically added MORE and MORE options, rather than taking away options.
Did this mean it was a worse game in my opinion? Yes, absolutely. I did not have as much fun with GTA 4 as I did with the entire GTA 3 series.
And yet, all games really are ultimately linear in some respects... Even entirely open-ended games like Minecraft have linearity in them. And you do a good of pointing this out, but I want to emphasize how 'PLAYER OPTIONS' are what create the linearity and open-ness of a game.
Even in a game like Minecraft, linearity can be seen in that you have the that all you really do is begin the game with low-level tools, mine, get better tools, mine more, and then get the best tools... Which you then typically will use to mine a lot of what blocks you want so you can make cool structures. And in the midst of it all, you fight bad guys, too.... But that's all you ever do.
It's linear in that there are not any options other than 'Mine - Build - Fight'... The open-ness of the game only comes in with the fact that you can 'Build' nearly anything imaginable.
I think these player options are what truly make the open-ness of a game... and as well as what makes a game "fun", at least for me.
I don't really have much "fun" with entirely linear games... What's the point? You run through once, then you're DONE. The end.
I like my games to have replay value, personally.
As for game design in general, as I said already, I love the more open-ended games. The ones with more options, the better they can be.
And personally, all of my plans for game design try to encompass and put emphasis on these gameplay-options, and as many of them as possible.
And while I won't say what I'm working on right now, I give most of my thought into 'how a player might want to play it'. I try to think, as a player, what options I would want to see in the game... I try to use my imagination to think up even outrageous stuff like "What if I could just pick up that chair and hit someone with it?" and while a lot of that is unlikely to be practical in a game, there are still ways to implement the "fun" of such ideas, while not taking too much time on actually implementing the details of it.
An example is a game idea I've had for a long time where it is a traditional-looking "RPG" game, yet the player has the ability to open up, or even destroy, towns all over the game-world. And to even determine how those towns are run, or if you just want to start massacring a village because it's fun. Those are things the player imagines they want to do when they play an RPG... so why not let them do it!? Of course, that game idea will have to wait for a while, when I hopefully get a little bit of income to work on games like that...
I think, for example, the 'Fable' series did this relatively well at the time, at first... You can basically, slightly determine the outcome of certain events, and your actions have real repercussions on the game-world. You could even get married to random villagers, or kill them. You could wander around being a trader, trying to make profit, or you could go around being the 'good guy' or the 'bad guy'. It was your CHOICE... because the games gave you those options.
Although I feel the latest, Fable 3, discouraged this point of view and did not stay true to the roots from which the first two games were designed. I thought Fable 3 basically steered you directly down a single path, and it was pretty annoying... it didn't really give you many options... Kind of like GTA 4. They replaced all the game options that made the game good, and replaced it with better graphics (which weren't really that much better, in Fable 3's case).
Anyway, just thought I'd throw in a few of my own thoughts about linearity, because it's a topic I'm very passionate about in gaming and game design.
I agree, the more open-ended designs tend to make more interesting games. But being linear isn't nessecarily a bad thing, if done well. Thinking about it further, uncharted 2 is a great example of a linear game done well. It is designed to feel like you are in a movie. The plot if very fixed, many sequences are heavily scripted(this is apparant the first time you fail to climb the train dangling from the cliff, and everything breaks and slides as you ge to the exact same parts). This works for them, because they use it to finely tune the expereince. They can engineer the fight that will occur while you are hanging from a street sign. They can script the train that is sliding off the cliff to make it feel urgent and that you just barely made it out alive. In that case, the linearity allows them to craft the experience very precisely, which worked out remarkably well.
In generaly though, I find that more player choice is a good think. My game designs tend to revolve around it. For instance, the game I am working on now will allow you to customize every cybernetic component you have, so you can aechieve a side variety fo playstyles. Another design that is on the back burner features a extremly open-ended magic system to allow you to create whatever spells you want. With the phrase "Go anywhere, do anything", the "do anything" half is the more important part of gameplay. That is what is so great about minecraft. Yes, you can go anywhere; people generally don't. There is no point. They do cover the "do anything" part extremely well. The sheer variety of things people accomplish is great.
Open-endedness in games is great. But it is not the answer for every game, and it is a choice that requires serious consideration about what you want to acheive with the game, not simply picking the singular right answer.
I don't think linearity is bad at all... but personally, it's not the type of game that I like.
I'd personally rather go to the movies, because that costs less than video games, gives me the same linear-experience, and is also fun.
But even 'movie-games' can have some non-linearity to them, like Heavy Rain, which is actually done quite well. It is a cinematic game, yet it allows the player the ability to choose how the movie turns out, and what the characters do and don't do.
I'd agree that there is no "right answer" in overall game design.
Although, there is definitely a "right answer" for me and the games that I like. But ignoring my personal opinion, there is a market out there for practically anything... and if you do something right, you'll hit at least one of those markets.
Yes, personal taste is a big factor in what games someone will like. Even with a flawless design, not everybody will like a game. There are people who don't like portal, because it is not their type of game. Though a solid game design is more likely to transcend people's normal prefrences. Personally, I don't care much for first person shooters in general. Nothing wrong with it as a genre(though I could argue against many instances of the genre), just not what I prefer. Yet I love the half life series. It doesn't do anything to transcend the genre, but it is so solidly designed I love it anyways.
Its also harder to design an open-ended game. Its not so much that its hard to create those options, but its harder to design a game around a body of choices. With half-life, they know you will have a certain set of weapons. The entire game situation is known, and they can design it with that in mind. When you add options, you need to make it so those options are meaningful and impact waht you are doiung, otherwise they are pointless. Yet you also need the options to be equally viable, otherwise its not an option so much a test as to the right choice to make. And while not every combination of options needs to be, nor should be, equal, there should be a clearly defined power level that good choices will result in. No one combinaton should be notably stronger than the others, nor shouldany intelligently designed build be particuarly poor. Of course, it will frequently be possible to make bad choices that result in poor builds, but if you are training in archery and weilding a sword, that is your own fault. However, training to be an archer and training to be a swordsman should both be equally viable. optimising your choices is an interesting aspect of choice, and needs to be considered by the designers. Optimization should be possible. However, optimized should mean effective, not game-breaking. Its also helpful to include a method to recover from poor choices. Otherwise you make the wrong choice, and kick yourself about it for the rest of the game. You also have to consider how different builds will function in every part of the game. It is perfectly fine for certain builds to excell in certain situations and have a hard time with others. As long as "hard time" doesn't mean "Can't get past". Don't mandate that the character must wrestle a lion. The big, hulking characters may be able to tie it into knots, but the scrawny archer is goign to be cat-food. All of this is possible to design around, but it is significantly harder than designing a linear game.