• Felix Brendel
  • Jonas Helms
  • Van Minh Pham

Milestone 1: Game Idea


Milestone 2: Prototype


Milestone 3: Interim Demo

Milestone 4: Alpha Release

alpha release presentation.zipdemo.mp4

Milestone 5: Playtesting

Milestone 6: Final Release


The built executables can be found here:

We mainly developed the game on Windows, but we also built and tested it on Linux.
Since we wrote the whole engine and game ourseves  and since Linux comes in may different
variations, it would be too much to certainly claim that it will work on 100% of them.
Please, if you have any trouble getting gemji to work on your platform, let us know.
We would be happy to help!

  • No labels


  1. I like the idea of simple rules leading to chaotic outcomes and making a puzzle where it becomes difficult to plan ahead. However, there are two things to watch out for: First, if the game is too chaotic, players might start taking random actions to see if the puzzle progresses and second, if you indeed procedurally generate levels, there might not be a proper progression of difficulty from level to level.

  2. G'Day Folks!

    1.) What I like:

    I think this really has potential as some kind of simple time-eating puzzle game probably working great on mobile platforms. The simplicity of the core gameplay is probably one of the major strengths of your pitch as it is easy to get into the game but still quite complex in its execution (As an indirect reference to your pitch: 'Something deterministic causing chaotic behavior'). I really like this concept and it is a nice almost mathematical interpretation of the course theme (which also fits the gems as most of them are geometric shapes).

    2.) What I find problematic:

    I personally see some problems with the generated levels in your high goal. I guess this could be hard to incorporate as it may be too random and not managable from a difficulty stand-point... I would think that pre-designed levels are usually better but maybe the generated ones will be great anyways...

    3.) Improvements:

    I think you can easily improve your game by creating nice visuals for each of the gems and associations you get from their appearences. If there are a lot of gems, it should be easy to remember what each one does. If you have an idea about how to achieve this, I think the result will be a fun game with a well extendable scope for the entire semester!

  3. This idea is straight up money if it turns out as planned. Port it to mobile, add microtransactions for gem skins and special animations and send me a check to once you are millionaires. On a serious note though, difficulty balancing with procedural generation sounds difficult, but not impossible, give it a shot!

  4. What is your favorite aspect of the proposed game? Why?
    I like that the game only has one central mechanic (move one gem, activate other gems) that you try to build upon. Puzzle try to challenge the mind and will therefore become rather complex by nature, so balancing said complexity out by having with simplicity (one central mechanic) seems like a good approach, as the player can focus on the puzzle itself instead of being overwhelmed with options.

    What is your least favorite aspect? Why?
    The chaotic nature of the chain reactions might get out of control pretty easily, so it's important to not make the game a guessing game. Especially considering you want to add an "undo" button, players might temp to just trial and error their way through. I think a "reset level" button is essential if you can't avoid deadlocks, but "undo" makes it so that the player is never punished so they never need to think.

    Which single change or addition would you suggest to most improve the game?
    Depending on the number of different gems you will have, it might be really hard to imagine chain reactions in the players head. That's why I suggest you have a constantly available UI element with a list of all gems and their reactions. You could divide the screen into a "board area" as the main focus and a smaller "rule area" or something similar on the side. An alternative would be that the game board covers the whole screen and the UI element can be shown/hidden by the press of a button. It could even be extended by some functionality that helps the player plan their moves better, maybe with a smaller version of the board where they can draw or move pieces themselves.

  5. Your design seems quite elegant to me. You kept the basic elements simple, but I imagine that the game can still be very complex when the individual elements interact with each other.

    What is currently unclear to me (maybe I just missed the information somewhere) is whether only the colors relate to effects or whether the different gem shapes have their own effects as well, and if so, can the effects of both be combined? If they can be combined, then the game might be very overwhelming at some point. It may be better to have only a very limited set of different effects but have those interact in very interesting ways.

  6. I like the idea of making a game, which is easy and quick to play. Also the mechanics seem kind of interesting and new to me.
    However i also see a problem, that the game might get in a state where its just not solveable anymore, which might cause frustration for the player.
    The 2 possibilities i think of, to avoid this are:

    1. Somehow only allow levels that cannot get in such a state, which is probably veryyyy difficult, especially when the levels are procedurally generated.
    2. Detect those states and notify the player. Probably also difficult to do, but necessary in my opinion. 
  7. What do I like the most?
    What I really like about this project is the combination of a very simple gameplay principle, as well as, a potential for a lot of actually interesting puzzles.

    What do I like the least?
    I think, if not carefully crafted, limiting the game area to a 5x5 grid, could possibly influence the level variety negatively leading to repetitive tasks.
    On the other hand, increasing the grid size could probably lead to very chaotic levels, with a lot of not-easily predictable behavior, which isn't that great in a puzzle game such as this.

    What change would I suggest?
    Not really a change, but I have to say that I think procedural generation is both an intriguing and very risky subject for a puzzle game. I wish you the best of luck implementing this properly!
    Also I think the approach of making a rather "simple" game, but in your own engine is pretty cool!

  8. What is your favorite aspect of the proposed game? Why?
    I like your idea because it combines simple actions with a complex effect. This gives the players the feeling that
    they always know what they are doing but at the same time makes it quite challenging to solve the puzzle.
    I'm also very interested in how you will manage the procedural generation (if you have enough time to implement it).
    And you have my respect for using your own game engine (big grin)

    Possible problem:
    Especially when using procedural generation, your puzzles might contain possible infinite loops.
    I don't know if you plan to limit the number of activations for each gem, but if you don't this might become a problem. 

    Which single change or addition would you suggest to most improve the game?
    It would be great if you could add a "undo" action to revert one step (just in case one did just make a single mistake
    and doesn't want to restart the whole puzzle).

  9. I like the basic concept of creating a game with simple mechanics, because they are usually easy to learn but hard to master. Because of this players can be entertained for an extended period of time which is further enhanced by the procedural generation. 

    However, I think that it will be pretty hard to generate those levels and keep them interesting, because they are either too restrained or too loose and hitting the perfect middle can be quite challenging. 

    I think it would be nice to have a reset level button so that you can replay the level if your stuck somewhere and want to take a new approach. 

  10. I think the game idea in itself is not bad. But I think it's going to be quite challenging to create a puzzle of appropriate difficulty. They end up being too easy or too difficult very fast.

    I really like the design and the fact that you are working with your own engine for it.

    1. A puzzle games with a simple core concept but with a possible complex combination when it is well balanced sounds great.
    2. Generated level are awsome in concept but the execution would need to be perfect to have a good player experience. Maybe focussing on well handmade level is better.
    3. I know you wrote it in your extras: mobile platform. This would be the perfect casual mobile game.
  11. I LIKE that you create your own engine. It is definitely a lot of work, but experience you gain with it will be invaluable. Your sketches already look appealing!

    I FEAR the PC version won’t reach your target group. I know it is just a university project, not a real product, but I think such puzzle games are much better off on mobile devices. I hope you’ll have time to port it.

    MAYBE you should consider giving different forms to gems of different colour. White and red are the same on a Figure 1. You shouldn’t underestimate how many colour blind or colour weak people surround us xD Purple, red and brown are really difficult to distinguish if you wanted to introduce them.

  12. Like: The general mechanic and interaction between gems.
    Fear: Balancing levels, especially when randomly generated
    Suggestion: Puzzle games don't need to be played in one go, so people may not remember what effect each gem has when coming back to the game. You could provide some visual aids for the player so they can look up the effects easily any time.

  13. I think it's cool that you decided to use your own game engine.
    If you manage to port this on mobile, it could be quite successful.
    Maybe maximize the visuals and animations of the game to make it more exciting to the eye.
    Also, the play testing time to see what levels are especially easy / weak to solve or find unsolvable puzzles might be longer than with other projects.

    1. I think it is really cool, that you plan to implement the game using your own game engine and the concept seems as it would fit perfectly on a mobile device.
    2. In my opinion the level/puzzle design is a very important part of your game and you should put a lot of effort into it to keep the puzzles interesting but not too hard.
    3. It should be clearly described how the different gems work. So I would say a tutorial or menu that explain these functionalities would be very important.
  14. Favourite Aspects: I like the abstract puzzly take that you have on the order and chaos theme, and envision a lot of high satisfying ripple effect combos that a player could trigger in the game. I'm also looking forward to seeing how your custom engine endeavour works. 

    Fear: The custom engine overhead on top of the gameplay logic may be a rather large scope for a semester. Since you mentioned that you've already worked on the engine in the previous semester though, it sounds like you have your bases covered.

    Suggestion: In general, abstract puzzles can be off putting to some people because they are hard to grasp. I've been in this position with some of my designs as well. Looking through your gems and their abilities, the user is currently required to memorize the rules of six different abilities, albeit with potential visual cues as you mentioned. My suggestion is to maybe explore a thematic relation for your puzzle instead of just gems? i.e. keep the same core gameplay, but just call your gems something else that players can intuitively relate to? I would think that this would reduce a lot of the rules overhead for the players.

  15. I was positively surprised to hear that you are using your own game engine. I think building this game will thus be a great learning experience and I respect your choice,

    I hope the random generation does not interfere too much witth the difficulty of the puzzle itself, As people mentioned, finding an appropiate difficulty/learning curve is key in a successful puzzle game.

  16. I love that youre using your own game engine! Also the game already looks super clean and neat, i dig it. However, make sure to invest enough time for level design and later the generation of levels.

  17. Favorite aspect: Rules engine. A game that builds on the interaction of a solid rules engine already allows for fun by just messing arround with it. 

    Least favorite aspect: Long interactionchains prohibit player agency. If the interactions become to numerous the game may start to feel more like a random number generator then a playable game.

    Single cange: Limit number of gem types per level. Too many rules interacting with each other is cumbersome to play with.

  18. Although I'm not a big fan of puzzle games, I have to admit that I do like your particular game design and I can definitely imagine myself playing the game on a mobile device. I really admire the fact that you are developing the whole thing on your own game engine, but unfortunately I can't give any concrete suggestions on that since I haven't worked that much with custom engines. Independent from the engine, I still assume that you will have to invest a significant time for balancing, similarly to other projects incorporating procedural generation. Also, when it comes to puzzle games, the users may have a hard time keeping track of the rules/mechanics during the initial phases of the game. So I would suggest focusing heavily on the initial tutorials and maybe even integrating persistent cheat sheets as UI elements.

  19. What is your favorite aspect of the proposed game? Why?
    Not exactly about the game, but I like that you work with your own engine! Talking about the game itself, I think your focus on visual clarity, so that players can quickly figure out what will happen, is very good because...

    What is your least favorite aspect? Why?
    ...I'm a bit afraid that the chain reactions can get out of hand very quick and be nearly endless (or feel like it anyway). I'm not a huge fan of puzzle games like this, so I am a bit biased, but I think you really have to take care that this won't destroy the flow

    Which single change or addition would you suggest to most improve the game?
    Give the player an option to speed up the game. Especially for more difficult levels they will probably think quite a while about how the chain reactions will play out and don't want to watch the 20 chain reactions they thought about for the past 3min play out over the next 3min.

  20. Love the idea! I've got two feedback points:

    1. Some basic rules could be simplified perhaps, not all rules seem as simple and straight forward,
    2. Since the point of "chaotic" end-result of simple rules is, by definition, "unpredictability", if the player is supposed to merely watch this "unpredictable" sequence of events play out without any agency, that might be a bit frustrating. I would allow players to interact with the system "while" these chains are being played out, but providing some bonus for players who do that less often.