Tuesday 4 January 2011

Embracing Hardness

I am not very fond of new year's resolutions, but I will make one anyway: From here on, I promise myself to never take the simple way, but always take the hard one, when making game.

This might seem a bit weird, so let me explain myself.

When creating games in the past we have sometimes tried to take the easy way out, hoping to create a lot of "playtime" for little effort. This kind of thinking have always ended up being the worst parts of the game or the worst ideas. For example, in Penumbra Overture, I designed some of the maps to be maze-like and have roaming enemies, thinking it would be an easy way of adding engaging parts to the game. These levels turned out to be tedious and easily my least favorite parts of the game. Another example is from Amnesia: When coming up with the basic gameplay design we were set on creating some easy way of making levels. This ended up being a bad way to go about it and we pretty much discarded all of these features in the final game. Instead we went back to doing it the hard way - with much better results.

After releasing the Penumbra games, we actually felt a bit annoyed that it was so hard to make new maps for them. We saw that other games could put out a lot of map-packs and similar, but this was very hard for us, and would cost almost as much as making a new game. This feeling is not a new one, and I have personally felt like this many times. The ability games that could be completed quicker and that allowed for simple expansions.

This feeling has resided a bit after the release of Amnesia, but until very recently it was still there, nagging me. Then it suddenly occurred to me that I should not feel bad about having work that is hard to make. Instead I should feel proud and embrace it. I know this might sound a bit silly and self-evident, but it honestly came as a bit of a revelation to me. Not only should I feel good about any part of the game that was hard to make, I should actively strive for it and discard anything that is too easy. If a feature can easily create gameplay for a part of the game, it should be considered a bad idea and either scrapped it or a adjusted into a harder version. This not only because of personal motives, but because I am quite convinced that it will result in better games.

We should of course try and make the process of creating the game as simple and straightforward as possible. Just like we, to great success, improved and greatly simplified level and entity creation for Amnesia. Handling the tools of the trade is not what is meant to be hard, but the act of creation. I am also not implying that we should try and reinvent the wheel and try to come up with new solutions to already solved problems. What I am saying is that if any part of the game is too easy to design or implement, then we should be critically examine if it is really needed and if we really put enough thought into it. It should be considered if there are any ways to vary, expand or in any other way change it to make it harder.

What I am hoping will come out of this, are games that give a much richer experience. I think a good example of this at work is to compare a game like Braid to a "normal" puzzle game. The first few levels of Braid could easily have been expanded into a full game, but instead the hard route was taken. This resulted in a game where the gameplay is constantly fresh and provides a much deeper experience. You also see the same kind of forces at work when comparing the Super Mario games to contemporary platformer titles. There is a certain degree of quality to the Mario games, a large part of which I think comes from to doing things the hard way.

This will most probably result in more work for us, but as I now aim to embrace hard problems, that should only mean we are on the right track!


  1. Frictional Games' strive for perfection is exactly why I'm all in favour for Indie games.

    Although you guys really do take it to the next level. I wish you all the good luck for this year.

  2. I remember seeing videos of a level editor for Amnesia. But the final game levels seem to have a lot more variety than such an editor could allow for. Did you end up not using it?

  3. We used it. But we did use a lot of large objects to spice things up. For example, the entire cistern level is basically one large mesh that we imported into the map editor.

    Fallout 3 also use similar kind of edtior actually, but instead of having walls, etc as the basic build blocks they have entire room pieces (with walls, floors and ceiling), ie part of a corridor.

    This kind of editing is kind of slow started, as you just need a lot of pieces made. But once that is done, it is really nice since it is easier to change maps and to mess around (given that you do not use too many large set pieces, like the cistern).

  4. Great post!

    I totally agree.

    It’s not uncommon to see developers fall into the understandable trap of having to create large amounts of content and therefore finding the quickest and easiest way to do it, resulting in content that is less effective overall.

    So it’s good to know u guys are doing your best not to fall into the same trap.

    It’s funny how you would mention the Cistern room, as id easily say this was my favourite room in the whole game. No doubt putting it together as one whole mesh would have been a harder creative process in comparison to piecing it together in your level editor?


  5. Need any extra level designers?

    For free I might add. The experience would be worth more to me then money.

    I like challenges, the hard way sounds fun.


  6. my name is fabiano lima,i´m living in Brazil.First of all i´m a big fan of your games since the beguining. Íf you need 3d character models and animations ,i like to help some how.If you like my work off course.
    porti online: www.biano3d.wordpress.com
    e-mail: biano3d@yahoo.com.br

  7. I would say that your hard work has and will continue to pay off in the end. Your work is incredible and you've made greater strides in this industry than an Call of Duty game. Thank you stepping away from those publishers to create something that is completely your own and can truly speak to players in a way that the next Modern Warfare or Halo Bleach cannot.
    Frictional, you rock!

  8. I just visited Fabiano Lima's gallery (he posted as Anonymous above), and I must say it's a quality work, and that Frictional games could really use someone with zBrush skills like that, and if the guy knows how to animate right then IMO you should go for it, because frankly, animation is not where you shine (but, that can be said for some AAA games too...).

    Don't get me wrong, I'm a big fan since I've first laid eyes on Penumbra tech demo, even played Fear :).
    So, I'm glad to hear you are taking this game dev business seriously (LOL). Looking forward to the nextfrictionalgame.

    P.S. About the cistern... It was a really nice level. But... In the terms of the relation between the believable story and level design, what was the reasoning behind the decision to block the ladders with a pipe? If the pipe was there before the ladders, who in the right mind would putt the ladders in that place? If it was the other way around, then... who in the right mind would put the pipe there?
    A rushed puzzle design?

  9. A little late to the party, but here's my Amnesia review:


    Amusingly enough, I've had some annoying Internet issues on and off here, so I had to hoof it over and buy the game through a friend's connection, save it to a USB drive and install/play it on my own PC (Ah, for the days of just walking into a store). Of course, all that effort was well worth it (except for the part with me now sleeping in the closet under every blanket I own)...


  10. I agree with almost everything except this.

    "You also see the same kind of forces at work when comparing the Super Mario games to contemporary platformer titles."

    I always thought the Mario franchise was insanely over-rated.

    If you want deep platforming instead of dumbed down like in Mario, i suggest you try jump maps for the following FPS.
    Urban Terror
    Defrag(mod for quake 3)

    Its great that u guys have this approach.
    Amnesia is in my opinion the best survival horror ever and i hope your next game will surpass that.
    Good luck in future.

  11. Brandon Matthews5 January 2011 at 23:27

    Hi Thomas,

    Great article. Your post's are always the most thought provocative to me and I admire the way you treat the game's you make. Just remember when doing things the hard way that one day you will discover a way to make something easier and I think the proper implementation of that will be a great boon for everyone there at frictional. Much like the cotton gin was the textile industry. All things in moderation I say.

  12. I wish everyone had some kind of work ethic code. Too many dishonorables in Australia that don't do their job properly.

    It is good that you are doing this, because small problems on basic mechanics probably have already been solved by someone else and therefore, nothing is gained. Having this attitude amongst the programmers would also be very beneficial, because complicated and seemingly insurmountable overheads can usually be solved by some smart computer science practices and would teach them to think outside-the-box.

    By the way, HAPPY NEW YEAR!

  13. It's great to hear of such devotion to game making. I hope it turns out a good decision.

    But I cannot help but think that this entry is a big step against procedural development (in this case, of scenarios). You have limited resources as the small team you are, shouldn't you need to use "the easy way" eventually to make more content than you can do manually? I'm not saying that you should fill your games with unnecessary bits to make them bigger, but that, instead of totally discarding shortcuts, it may be more productive in the long term to create solid procedural techniques that produce quality "filler". That is, instead of discarding the easy path because it is not as wonderful as it looked like in the short term, expand it until it becomes wonderful in the long term.

    Besides, as developers of horror games, you may fall into the trap of trying to control every bit of the gameplay. This is something I noticed of Amnesia, it's a very linear game (Penumbra games too, but being the first using the "Frictional formula", they felt more fresh). This blog entry hints a little bit in that direction or, at least, that's the sensation I had.

  14. This is certainly an interesting article, but I do have some issues with your logic. It seems that you are essentially saying (correct me if I'm wrong) "simple = bad", which is true in some cases, but not in others. Yes, it can lead to crappy filler like mazes often do, but it could also lead to eloquent new ideas (the first thing coming to mind would be minecraft, which isn't necessarily the best comparison to a story-driven game, but I hope my point was still clear.)

    Really, the difficulty of implementing a feature shouldn't have much to do with its quality. Like Brandon's example of the cotton gin, sometimes simple can be good. And sometimes complicated features can be bad. It seems like it would be a better idea to ignore implementation and judge what a feature could add to the game, and what possible problems it might have.

  15. Dylan:
    Yeah, I agree that even if something is simple to implement it does not need to be bad. My main points are:

    1) One should not strive to do stuff in a simple way as this will often lead to bad decisions (eg the mazes). This case means "embracing hardness" is a help on how to attack problems and how to do think in general.

    - If an idea is simple to implement and provide a lot of "gametime", one should get suspicious. Creating gameplay is not like programming, where it is much easier to evaluate ideas (you can just ask: "does it do the job?" and it is much easier to try it out). Instead designing gameplay is something that you as a designer is instantly biased too. It is VERY easy to fool yourself (and others!), only to find out much later that it was a bad idea. In this case "embracing hardness" is a filter for ideas.

    Looking back at past choices I am confident that this will be a good idea, but I might be dead wrong :)

    Also want to add, that I had a lot of trouble writing this post as I was not sure exactly how to express it all. This means that actually writing this made me consider it more and to in someways evaluate this methodology.

    [side note rant]
    Using blog-posts as way of thinking about and solidifying an idea is quite common for me. Many times I start writing with just a small idea of what I want and then learn a lot in the process. As I know people will read it, I am try to check all my arguments careful so I do not look too foolish. This has been extremely helpful and I have actually abandoned some ideas / theories while writing a blog post about them (and then never publicizing).
    That means all feedback to you people write here is really useful, and I an grateful if anyone calls me out any bullshit I may write :) I am very aware that we humans are prone to various biases and I try hard to not fall into some line of bad thinking.
    [/side note rant]

  16. In response to previous posts:

    IMO, Thomas is not saying that
    (simple OR procedural OR quick) == bad,

    but that

    (simple OR procedural OR quick) + not_trying_to_be_creative == bad,

    or that

    (simple OR procedural OR quick) + not_trying_to_be_innovative == bad,

    or just

    (simple OR procedural OR quick) + not_trying_IN_GENERAL == BAD.

  17. MichaelPalin:
    Just wanted to comment on "controlling gameplay", this is actually something we moved away from in Amnesia. Instead of trying to set up exactly how solutions should be solved, we tried to create spaces where players where feel to explore and roleplay.

    An example of this is puzzles, where we tried to include any solution that testers tried doing, even if it made the game simpler. So we tried to make sure players where not stuck, trying to guess "what does the designer want me to do?".

    Another example is creature encounters. For example, in Daniel's room where a grunt appears, it is possible to avoid the encounter entirely (by not searching the room in a certain place) and hiding in a closet is not the only way to escape the monster. In other games, situations like this are often forced through quicktime-event or (worse) cut-scenes). Instead we try to leave it open and lure the player into reacting in an interesting way (through hints and environment design).

    However, I can understand that this might now show very well when playing and the choices are not always obvious (compared to something like GTA). I do think it created a more smooth experience and gave more freedom to roleplay.

    I know it is far from perfect in Amnesia and we will try and take it further in our upcoming games.

    In case you are interested, here is a post that cover more of this:


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