Osmos Rage, part 3: The Perversity of Inanimate Objects

Posted by on Feb 2, 2010 in Osmos, Rage | 6 Comments

Welcome to part 3/4 of our “Osmos Rage” series. We introduced the topic in part 1, announced a video contest in part 2, and today we explain why the game gets “so bloody hard” in the end.

Believe it or not, our primary motives were a) truth and b) respect.

  • Truth: to the game concept and the scientific principles that underlie it. We could have easily made an easier game. In fact, I am 100% certain Osmos would have received better reviews and a better Metacritic rating if we had done so. For the most part, reviewers and gamers alike really appreciated the concept, execution and ambiance of the game, but a common complaint was the difficulty in the late stages. We could have simply removed those late stages, but there are some beautiful mathematical and physical principles that only emerge in them: subtleties of momentum transfer, efficiency, and in particular orbital mechanics. Players can “brute force” their way through most of the easier levels, but not these. In these, the player is forced to really understand the systems at work, and how to be efficient in them. These are some of my favorite levels in the game, and I feel we would have done a disservice to the concepts and to players by leaving them out.
  • Respect: for players and their time. Levels in Osmos are procedurally generated, so it would have been trivial for us to add more. We could have stretched the game out with a much smoother difficulty curve; It would have forced players to spend more time building their skills, but the game would have also felt more repetitive. We wanted each level to feel distinct, in concept or at least in difficulty. As an “adult gamer”, I don’t have the time nor inclination for grinding anymore. One of the things I really appreciated about Braid is its lack of filler — it had been reduced to the essential experience. I wanted Osmos to do the same.

And when I see reviews like this, it all feels worth it.

Somewhere between completely loathing the difficulty of the last level and finishing the game, I finally “got” Osmos. You see, Osmos was able to do what no other zen (or ambient) game has actually made me do—chill out. Osmos has made me contemplate the very essence of human existence, evolution, consumption, and so much more. The difficulty of Osmos’ later levels was very frustrating at first, and often caused me to quit the game rather angrily; I was approaching Osmos like I would any other game, swiftly trying to complete all given goals in the pursuit of completion. Once I started approaching Osmos as not a game, but as an experience, I was able to fully appreciate the genius behind it. – Dan Carew, Blast Magazine

We even got this praise from Jonathan Blow, creator of Braid.

The result, by the time you’ve finished, is not merely a satisfying game. It rings with that faint and distant sound of truth: because the game is based around laws of physics, it immerses you in these and you learn something about them. Perhaps not anything you didn’t already know in an abstract intellectual way, if you took physics classes in school; but here, you get a feel for them, so they become more real, more tangible. This game can change your perspective.

All that said, people may still criticize our presentation of Osmos as an “ambient” game. My honest opinion is: it’s mostly ambient, but not entirely. Our focus throughout was on gameplay. Yes, we put a lot of love, time and importance into the mood, visuals and sound; but when we discovered level types that really worked well with the fundamental mechanics of Osmos, we included them, even if it meant breaking the ambient rule somewhat. In the end, Osmos is a tricky game to summarily describe, and since it’s mostly relaxed, we kept the tag. For thoroughness, here’s a detailed breakdown by level type:

  • One third of the levels are in the game’s “Ambient” branch, and I believe these easily qualify.
  • The “Sentient” levels, however, are not. They’re fun, and they’re many people’s favorite level type; but given the time pressure imposed by the AI opponents, they aren’t really relaxed — the player needs to click too quickly and too often. (Though slowing down time makes things much less frenetic.)
  • The “Force” branch is a mixed bag. I find that once one understands how to navigate in these systems, the “Zen Attractor” and “Epicycle” levels truly are zen; patience and judicious clicking is the way to play and win these levels. “Warped Chaos” levels on the other hand are not relaxed; players need to move quickly at the start to gather as much mass as possible. However, once these levels stabilize, it becomes more like a slow, warped billiards game than anything else, which is quite ambient.

Finally, we included a number of features in Osmos that give the player a great deal of control over the difficulty and progression.

  • Time Warping: This allows the player to control the speed of the game, giving them enough time to think and click wisely. It’s a form of user-controlled DDA (Dynamic Difficulty Adjustment). It also allows players to speed up time to avoid long waits, or to increase the difficulty for the hardcore.
  • Randomization: Instead of including more levels and potentially creating too shallow a difficulty curve, we included this feature. Players who are looking for more in-between levels need look no further. If you’re frustrated on a level you can’t beat, go back one level and play some random variants. You’ll enjoy it a lot more while still building your skills. Try to do those variants as efficiently as possible; before long, you’ll be able to return to that “impossible” level, and probably beat it.

Hopefully this helps people understand the rationale behind the difficulty of Osmos’s later levels; and perhaps you’ve found this view on our design decisions interesting. We stand by them, though we’re always open to constructive criticism. Actually, no… wait… screw “constructive” and hold on to that rage! We need videos!! Tortured, infuriated, death-metal videos! We’re bad, bad people. Throw rocks at us. We deserve it. ;-)


  1. zenpunk
    February 9, 2010

    you did the right thing. The difficulty of the later levels was 90% of the enjoyment of the game for me.

    ambiance schmambiance … I love the puzzle aspect of the game. I finished the game, but I’m sure I will return every so often just to have fun flying orbits in the zen levels, to see if I can train myself to swim around efficiently and smoothly without having to thing too hard.

    Beautiful game.

  2. zenpunk
    February 9, 2010

    *think too hard

  3. Yowza
    November 11, 2010

    What you guys seem to forget is that completionists like me and probably a ton of xbox users like to have that 100% completion stat…and making a game marketed as “relaxing” insanely hard makes me a little annoyed to say the least…and the definition of relaxing to me is a nice easy breezy game and not too challenging…will i beat the game?…of course…what i dont like is i have to waste time “mastering” it for a few levels…which is not fun IMO…

  4. player347
    February 2, 2011

    “essential experience” – great idea

    “[..]there are some beautiful mathematical and physical principles that only emerge in them: subtleties of momentum transfer, efficiency, and in particular orbital mechanics.” – making a game out of those physical laws was another great idea. I especially liked that you did not ignore Kepler’s 2nd law.

    What about a game based on quantum physics?
    the force there is not (cst.)*m1*m2/r^2 but (cst)/r^7 where r is the distance between masses and atoms respectively

  5. Omar Lopez
    March 6, 2011

    Can anyone give me some hints on epycicles. That’s the only level I cannot master yet. Awesome game, btw, Omar

  6. trlkly
    July 7, 2013

    And now we know the problem. An ambient game is SUPPOSED to feel repetitive. That’s what’s lets you have that zen-like feeling. You’re no longer thinking about the gameplay, but just doing it. Instead, you admit that you are trying to make the player think. That’s not ambiance.

    It is not respectful to the player to make them have to grind in order to get good. It’s the reason older games are so hated.

    And it’s not just in the final levels. You do some stupid things. For example, the first Nemocyte level is already extremely hard. You’re playing against essentially a copy of yourself. The only difference is the Nemocyte doesn’t have the goal of absorbing you. You can’t win this by thinking. You just have to move quickly and get lucky.

    So then why in the world is the next level 3 Nemocytes? Why in the world don’t you go back to the much easier enemies when you start including more than one? Even 2 nemocytes would be better. Instead, you’re just going along, slowly getting bigger, and, all of the sudden, one of the nemocytes will come barrelling toward you, and you don’t have enough mass to get out of the way.

    And, no, randomizing the older levels isn’t a replacement for that. They don’t get any harder when you do that. And it’s not as if it’s easy to get to all the levels anyways, since you put multiple levels per map item.

    And that gets me to what I see most–a complete inability to admit to mistakes. The bad stuff is all about us and how we don’t understand your game. The good stuff is all about you and your goals when making the game, instead of thinking about who is playing it. You even admit that making the game easier would make people like it more. But you instead decide to be selfish instead of caring about these other people.

    Even your methods for fixing the problem are selfish. You can’t be bothered to make a new trailer that would give a more honest impression of your game. You want to farm off the work on everyone else. Why should someone who is frustrated with your game want to help you advertize it?

    I really liked this game when it started, and just by the fourth level it got frustrating. Frustration is incompatible with ambient gaming. You were clearly going for ambiance. So that means you messed up. Admit you made a mistake instead of pawning it off on us, having us try to fix your mistakes.


Leave a Reply