Tuesday 21 July 2009

The problem with obstacles

Even though freedom is something of a buzz words these days in games, most games needs to restrict the player somehow. This is especially true for various types of adventure games where the player must be guided along a story path. In this blog post I will call these restrictions "obstacles" and will briefly discuss the various design problems connected with these.

First out I want to start out with an example of some obstacles from Penumbra. In Black Plague, after the player as managed to escape from his cell and get to the residential area, we wanted the player to search the area, find notes and solve puzzles. In order to do so we need to halt the player's progress and this was done by a door needing biometric input in order to open. To do this the player must collect some body parts and these are in turn blocked by other obstacles that need to be overcome (another locked door, corridor with gas, etc).

The big problem we have had when designing things like this is to make the obstacles seem well placed, fun and varied. Unfortunately it often boils to having some kind of locked door. And as we all know, while fitting, locked doors are not that exciting. If locked doors can not be used, what can? Below follows is a list of some different types of obstacles:

  • Object. This is things like doors, bridges or other man made things that are blocking the players path.
  • Environmental. There are obstacle that somehow limit the player movement and include holes, rivers, fires,etc.
  • Character. This is normally seen in old point and click games. Some character is blocking the players path and requires something to let the player through.
  • Enemy. A deadly creature of some sort that blocks the player path.
  • Motivation. The player character does not want to continue because of some personal issue. Perhaps the road up ahead is too dark.
  • Hidden. The path that the player needs to take is not yet visible. For example a portal that magically appears after the some condition has been satisfied.
I think these pretty much sums up all of the obstacles that are found in games. Also note that sum of these overlap, for example a robot guardian could be classified both as object or enemy.

For games like penumbra the game mechanic sets limit on what kind of obstacles that can be used. For example characters did not work because there where no real dialog system. Other games might have other kinds of limits. In some games it might not be a problem if the types of obstacles are not varied as it is part of the basic gameplay. In adventure games it is very important though, and only having one kind of obstacles (like always facing locked doors), makes the game feel repetitive and boring.

After releasing Penumbra Overture, we got some critique that the game contained too many open-locked-door obstacles and tried hard to fix this for Black Plague. The first thing we set out to do was to the skip simple locked door obstacles and if a door was needed we tried to make it interesting. In the example above we used a lock that required human parts to be opened and even though it was still a lock-and-key type of puzzles I think people considered it much more fun. The final game still contained obviously locked doors and we tried to limit this. However, we found it very hard and where not completely pleased at the end.

For Requiem we completely skipped trying to come up with interesting obstacles and instead focused on puzzles. In Requiem a portal always had to opened using some strange orbs in order to progress and this made the rest of the gameplay a lot simpler to design. At the same time it was apparent this was not good for an adventure game and player responses showed this. By using the same type of obstacle throughout the entire game a lot of the adventure feel was lost. This is especially true, like in Requiem, when obstacles are not part of the story either.

What are your thoughts on the obstacles in Penumbra? Know any game with really good or bad obstacles?


  1. Silent Hill Homecoming shows how to do it wrong.


    "The way is blocked. I can't go any further."

  2. I'm more familiar with RPGs, and these suffer from very similar problems, though their obstacles are mostly called "quests", it's the same principle - it's something you have to do to progress your game; and they consist of similar things like your categorization, like killing some enemies ("Enemy"), getting an NPC to do what you want ("Character"), finding some treasure in ancient ruins ("Object"), finding the way to some other city ("Environment") etc.

    So, even those games have not really found any more interesting obstacles than you have listed, despite the decades of experience and the enormous amount of people they have working on them. What they do instead to make it more appealing to the players is: Provide multiple ways to overcome an obstacle. If an NPC is blocking your way, you can kill him, talk him out of it, bribe him, find out who his enemies are and make them kill him, discover some secret of his and blackmail him, sneak behind him, steal the key to the door he's guarding from him, and so on, you get the idea.

    Instead of finding more interesting obstacles, they try to provide more ways to overcome them. In your example of the biometric door, besides of finding the required body parts, I could think of mixing and drinking some chemical stuff to make your body match the required one (Dr Jekyll/Mr Hide style), turning off the electricity to deactivate the door mechanism, find a way to apply brute force (explosives that are well hidden and difficult to find or to get), etc

    While certainly not all of that would match the style of your games, maybe the idea of multiple distinct solutions to an obstacle can help improve your games.

    Disclaimer: I have just recently started Overture, and not finished Black Plague and Requiem yet, so some things I said might already be there, and I have been writing lots o crap so far. Well, then just ignore it.

  3. While playing Black Plague I got a bit annoyed that there was so many doors you couldn't open, ever. There were many rooms you could never enter, or they existed just as doors and it bugged me because I wanted more places to explore. Empty rooms with just rubble in them would have been fine with me :P

    You might see now that games like Silent Hill that has hundreds of impassable doors like these annoy me to no end.

  4. I would suggest reading the following article:


    It's written by internet-known game designer/critic "Yahtzee" and it's about game obstacles in adventures and survival horror.

  5. The fifth one that you said, motivation, that seems interesting: making the player scared not to go into a place would be amazing.
    But I guess if you were to make a game like that then the saving/dying system needs to punish the player greatly (which I'm good with as long as it is understandable).

  6. Regarding doors that never open:
    I disliked this a lot in games like Silent Hill and it is annoying when a large part of the game world is cut off from you. On the other hand it also helps create the illusion of a large game world while still keeping the game focused, so I think it is not all bad if used correctly.

  7. I'll second the suggestion of Yahtzee's article, with the caveats that it's a little more specific than what you're talking about here (not every river in a game is a puzzle to be crossed, obviously; sometimes it's just the edge of the game world) and that I don't really like most of his suggestions for alternatives; introducing a time limit does very little to improve most puzzles.

    Likewise there's something to be gained from looking at the multiple solutions angle, but I would suggest that Deus Ex is a much better example of this than any RPG made in the last ten years; Mass Effect, for instance, is an excellent example of how not to do this. You could close your eyes and click your way through dialogue trees in that game, which is a pretty clear indication that they're not very good obstacles. Deus Ex was a good sort of proof of concept for this approach to design, but since then we've seen designers back off of the idea rather than expand on it because it's a lot of extra development time that doesn't really equate to extra content for most players. If you have the budget for it I advocate this from a design perspective but I recognize it could be tough to justify.

    Anyway, this all looks sort of tangential in that it's focused on solutions to obstacles rather than the obstacles themselves, but I think that's really sort of the point. It doesn't really matter what kind of obstacle you use, as long as you're following common sense rules to make sure it fits in your setting, etc. It's a good idea to vary the type of obstacle, but it's more important that you vary the type of solution. This is another potential downside to providing multiple ways around your obstacles; if it's always force/sneak/persuasion, it quickly becomes just as dull as a series of keys.

    I think there are, generally speaking, two different approaches that produce the ideal result; either it should be unclear how to bypass an obstacle until the area has been explored and the player has gathered more information, or it should be immediately intuitive and obvious what needs to happen and the player just needs to figure out how to arrange the pieces to bring about that result.

    The former, for instance, might be a situation like curing yourself in Black Plague; it is made clear early on that something is wrong, but it's not until you've spent some time gathering information in the game that you know what it is and still longer before you have the faintest idea how to address it. The latter would be more like the gas-filled hallway, where you very quickly ascertain that you need to clean the place out and it's just a matter of arranging the pieces right.

    It seems like the most common implementation is for the big mystery obstacles (if a game even has any) to be subdivided into a few of the more obvious sort, but I don't think it has to be that way; if you're willing to go for a less linear overall design, you can give the player multiple obstacles to be thinking about at any given time and the process of learning about the big mysteries is at once less of a roadblock and more natural, which in turn lets you use them as little mysteries too.

    Ultimately, I suspect this is a problem we're going to need better technology to solve in a really satisfying manner; having left behind the precision of text parsers in favor of graphical interfaces, we won't see make up the lost ground until we have much better physics than we do today.

    And I promise I'll be concise next time. Somehow.

  8. About Yahtzee's article:
    He mostly discusses the puzzles connected with obstacles, something I am going to discuss in another blog post. It is still a nice article though!

    Another thingie:
    I forgot to mention that an important aspect of an obstacle is that it must be obvious to a player that it is indeed is an obstacle. As an example a cave-in might simply be a blockage and not something that the player can overcome. It would then be bad if the player spends hours trying to figure out how to get past the cave-in when it is in fact impossible.

  9. That is absolutely true and was well done throughout the penumbra trilogy.

    About Thomas' comment concerning never opening doors:
    Unless the game world is not totally imaginary like a character's dream or vision such doors are a completely must in my opinion to keep up the immersion of a 'real' setting. As it was done with Half Life 1 - there were a lot of impassable doors but they led to the impression of a huge mysterious Black Mesa Research Facility which was quite fascinating. But as you said the player must clearly recognize such doors as definitely closed.

    About obstacles:
    In addition to the concept of different solutions to overcome an obstacle you can devise different consequences for the chosen solution. For example there could be a locked door with a pretty obvious possibility to blow it open. But if you choose this quick way then the blast could block other passages / informational content / savepoint. It the most complex form it could even lead to different directions of the game.
    This obviously takes a lot of conceptional time and work and is therefore only suited for central puzzles that are a crucial part of the story. But I found these kind of decisions always as most impressing in games and they led me to play them again to see the other ways.

    Another interesting possibility is to let the character lose mental stability for some actions like it was done in Call of Cthulhu - there you lose the grip of reality when using Morphine to often.

  10. I think the main problem with obstacles in adventure games is that they almost always seem artificial, thus removing immersion from the game.

    One really bad example of this is Half Life 2. Despite being a very good game, there were so many physics puzzles that were some variation of "this board just so happens to be placed on this particular rock so that if you put this conveniently placed barrel on the other side of the board you can walk over the wall".

    One way to stop this, of which I've seen very few horror games ever use, is having multiple ways of overcoming an obstacle as has been discussed above. I can't tell you how many times I've been playing a puzzle game and wondering "well, why doesn't this work?" and trying to figure out what the game wants me to do, not how I would actually solve the puzzle if it was presented to me in real life.

    So in order to quickly sum up my thoughts on the subject: while variety of the obstacles are important, the method(s) in which the obstacles are overcome (if they can be) is just as important, if not more so, to the sense of immersion in the game (which is really the ultimate goal of any horror game, isn't it?).

    P.S.- As for the somewhat unrelated topic of impassable doors, I'm not a fan of them in general, and they have the possibility to be greatly abused. Either provide a reason why that door is impassable, or don't have it in the first place. Otherwise the player tries to open them and are reminded "oh, wait. This is just a game; that door doesn't actually do anything."

  11. In my adventure game. I use the good old point and click "obstacles" with characters... - (cuts off there). I just got an idea!

    I never noticed anything like this in the Penumbra series, but maybe have a locked door that you really need to get through, but there is a monster inside preventing you from proceeding without dying. Then the player needs to get rid of the monster somehow, or find another way in to proceed.

    Not necessarily like that but you get the picture.

  12. I want to mention 2 things.

    First is my favorite obstacle in Penumbra Black Plaque. You may remember the door you need a head and a hand to pass it. After collecting the things the door disappeared and Clarence is talking to you. After you move ahead and returned the door is back. I really loved this moment.

    And second is an example for a "hidden" Obstacle. Soul Reaver 2 has an interessting way to pass them. You can switch you character between two levels (?). In one of them you can go "through" some obstacles. But watching is better then reading. In the YouTube-Video at the End of my post you will see at the very beginning the character moving through a door. At 0:50 he switches back into the "real" level.
    Haven't played the game in English so I don't know the proper words.


  13. One important thing that I realized I left out from my previous post was that the obstacles need to make sense. Now, with a locked door, this is pretty easy, but it can get more comlicated.

    Lets say you need to get through a door, and somethings blocking it. Was that item put there by someone on purpose to barricade the door? If so, why? Or, if not, why did the object happen to be in front of the door? Was there a struggle? Make up a little story about how and why that door came to be blocked. It takes very little time to do, and adds a whole lot to the game.

    If you really wanted to, place a note or diary entry somewhere close by and tell that little story to the player. Or, if you wanted to add to the effect, make a scene inside the room with the blocked door that makes the player realize what happened visually instead of reading a note.

    One game that used this visual method quite well was a game called Portal. You could sometimes find small, confined areas that had a bunch of empty food cans, writing all over the walls, sometimes even pictures, and you knew immediatly that someone had been cramped up in this space for a long time, slowly loosing their sanity. These spaces located throughout the game were often more disturbing than the most evil looking monsters ever seen in a horror game (which is ironic, considering that Portal wasn't meant to be much more than a simple puzzle game)

  14. I agree with the things said about obstacles in survival horror games concerning their story embedding i.e. letting the player guess what happened before he arrived at the scene. This is a important part of the fun.

    By the way I liked the physical puzzles presented in both Half Life 2 and Portal because they pointed at the speciality of these games: the physics engine. Since Half Life 2 is - besides the horror elements - also a playful approach of a shooter (jokes about the never speaking Gordon, speed racing, hitting weird aliens with radiators etc.) such obvious puzzles did not destroy some kind of special atmosphere.

    Speaking of the board and barrel puzzle: there was a similar puzzle in Black Plague in order to reach a scaffold and accessing the surface. It was there for the same reason: using the physics engine for a puzzle (although without a gravity gun) without any further explanation why exactly these puzzle parts were lying around. it did not destroy the immersion though - in fact I was glad about the puzzle because it was too complex to allow disturbing monsters showing up... :)

    Why not put the physics engine of Penumbra to further use for obstacles? Imagine a set of mirrors you must carefully align in order to guide a laser beam to a target?
    Or maybe you could use the pull/push mechanism for objects for this: you have to wait for a enemy that knows the combination for a locked door but he will kill you once he sees you. So you have to hide in a room nearby and watch him entereing the code. For this you must open the door to your room wide enough to see something but not too wide and not too fast since it creaks awfully.
    Or something like an instable floor. You cannot recognize which parts will collapse by sight but you can roll/throw some heavy objects over it so you disarm the "traps".


Note: only a member of this blog may post a comment.