Thoughts on Auralux - Minimalist Real-Time Strategy and the Importance of Opponents

on Thursday, August 23, 2012


    • E. McNeill
    • Real time strategy
    • Release: Unknown, out now.
    • Platforms: PC (version reviewed), Android



    Auralux is a study of sorts. It is an experiment, like Eufloria before it, in the most basic mechanics that define a real-time strategy game; it attempts to boil the genre down and show us how it works. There are no tech trees here, and there is certainly no grand historical context for a setting. There are no control groups, and you will not need to memorize scores of hotkeys. Auralux simplifies in the extreme: one unit, one command, plain primary colors against black.

    The one unit at your command is a simple dot, and the one command they accept is a move command with the cursor. These are produced at capturable, sun-like spheres (your production "buildings") laid out in geometric patterns to comprise a game map. Red, green, and the player's own blue each begin with one sphere on every map. When a dot encounters a dot of another color, they annihilate each other, and if many are tasked on an enemy or untaken sphere, they enter it to capture it. With more spheres, more dots are produced. Capturing all of the spheres on map means victory, and there are 24 maps of varying sizes and difficulties waiting to challenge the player. The complications are few and far between: Sometimes spheres can be upgraded to produce more dots, and the player has the option of playing in a high-speed mode. There also seems to be a population cap on the larger maps; I found this slightly disappointing because it is an extraneous rule in a highly simplified system. It is likely a necessary feature, though, to ensure smooth functioning on a variety of systems.

    It's easy to see how this setup relates to more standard RTSs. Micromanagement has been (mostly) eliminated, so all decisions are of the strategic, macromanagement variety. At any given moment, you have only the decision of where to commit your units. They can stay to defend or move to attack an enemy, or they can be invested in an unoccupied or upgradable sphere. The former actions comprise the combat aspect of RTSs, the latter is a simplified economy. The game becomes interesting because this reductive version of the RTS formula still leads to familiar, if smaller, versions of common RTS tactics, strategies, and battles. It calls attention to the types of balance that make such games successful, and it allows us to see how they function in great detail. For example, difficulty in Auralux is controlled largely by the starting positions of the three colors. In developing a successful tactic for a given map, the player must develop a way to balance the organization of the map before the other colors gain too strong a foothold. It quickly teaches the player the different ways that positioning can affect game balance.

    As much as this minimalist approach amplifies the strengths and subtleties of RTS games, it also, perhaps to Auralux's detriment, highlights weaknesses. People have always railed against strategy game AIs, and nowhere is AI weakness more apparent than here. In a way, perhaps by accident, this is an interesting commentary on the weakness of RTS games compared to other genres. Unfortunately, however, Auralux fails to adequately explore the impacts of AI opponents because it has only one type, and that type is not sophisticated enough. An AI that fails to implement or defend against flanking moves is hardly an AI worth discussing; an attack away from the front lines always catches the AI off-guard here. This severely undermines the emergent complexity of the game, and reduces the game's power to illuminate strategic mechanics. It means that difficulty and complexity of planning are functions only of uneven starting positions and army sizes - important aspects, but rather uninteresting. I would have loved to see some way of tweaking the AIs or even the ability to pit different AIs against each other.

    Auralux has another major omission with regards to opponents. The game is single-player only, despite the importance of multiplayer to games of this ilk. I'm sure this was a practical decision, arising out of the constraints of making a low-budget, one-developer game, but it nonetheless remains a glaring omission. Without it, the game is small, short, and limited in its ability to explore the genre. We can only wonder about the possibilities it could have offered.

    In a way, Auralux thus becomes a puzzle game. With such constancy in the opponents and their flaws, the player must sleuth out the correct set of moves that the AI can't deal with without destroying each other. Perhaps any sufficiently simplified game system becomes something that we would identify with a puzzle game.

    So where does this all leave us with Auralux? There is no doubt that it is a polished game with a very interesting take on familiar mechanics. And I'd be remiss if I failed to mention it's lovely sound design, with spheres pulsing with the music as a battle creates a randomized melody, all emphasizing the rhythm of the gameplay. As a whole, Auralux is enough of an exploration of the genre to leave me interested in further experimentation, but the game's inability to delve deeper because of its lone, primitive AI and lack of multiplayer means that it is inherently flawed. In a game that searches for how complex strategy emerges from simple gameplay rules, an inability to provide an opponent with which to create that complexity is an unforgivable limitation, even given the game's considerable strengths.

    Update: See the comments section of this post for a response from the developer of Auralux.

    9 comments:

    E McNeill said...

    Hi! I made Auralux. :)
    (I Google-stalk mentions of my game.)

    I agree that the game reaches its limitations primarily in the AI. When the mechanics are so simplified, the AI algorithm becomes the only source of variety and challenge. Much of the expert-level game just becomes a matter of managing and exploiting this algorithm so that the two enemies fight each other instead of you. The latter levels still require excellent knowledge and performance to win, but simpler levels (and the game as a whole) turn into, essentially, a puzzle.

    Auralux was an illuminating attempt to make a completely essentialized RTS. I'm currently prototyping a new game with the same goal in mind, using a flow network as the game's "container". I hope to resolve the inelegant parts of Auralux and get a step closer to real-time Go. Wish me luck!

    Evan said...

    Hi, thanks for taking the time to comment. You're quick with the Google stalking!

    I totally agree with your assessment of your game's limitations. You're right that the later levels become quite difficult, but the maps don't really grow in strategic complexity as they grow in difficulty. Timimg simply becomes more important as you flank them to pit them against each other. In fact, it sometimes seemed to come down to luck as to what the two opponents did as their first moves. On one map, I can easily win if they make certain first moves, but if one takes the opposite direction first, it becomes too hard to pit them against each other, and I am overwhelmed by shear numbers.) The comparison to puzzles is interesting; perhaps any sufficiently essentialized game type becomes reduced to what we traditionally think of as a puzzle game. I think I'll add a few lines on that above.

    Did you ever experiment with multiple AIs (defensive, aggressive, cooperative, etc)? It would be interesting to play the same maps with different types of opponents. I also found myself wondering what the game would be like if I could tweak different AI parameters on the same map to see how it affected outcomes more generally. It would almost be a whole other game if you had an AI editing interface (something like sliders for characteristics) and had to achieve certain outcomes in games with just three AIs instead of a human combatant.

    Nonetheless, despite my criticisms of its failings, I really was impressed with the game. I'd rather have a really interesting but flawed game than a rehashing of the same old things any day. It left me very interested in the "essentialized" concept, and I'm excited to hear that you're working on a new one. Care to elaborate on what you mean by "using a flow network as the game's 'container'"? (I'm also just really impressed that you produced such a slick looking game with only one person - great work!)

    E McNeill said...

    I never considered doing multiple or configurable AIs for the original game, but that's only because I wanted to keep the project small. I later worked on improving the AI as my college thesis project, though I haven't yet gotten around to coding it in full. If I ever make a sequel, I'm pretty certain it will have multiple AIs, though I'm not sure whether they'll be differentiated by difficulty, or by "personality".

    As for the "flow network" comment, I mean that the new prototype takes place on a mathematical graph (http://en.wikipedia.org/wiki/Graph_%28mathematics%29). The players control nodes (analogous to the suns in Auralux), which generate "liquid" (analogous to the units in Auralux), which gets pushed to other nodes through predefined pipes. It's essentially another minimalistic RTS, but I'm constraining the action within a more limited framework, and also preventing the player from holding back forces like you can in Auralux. This can make AI and controls a lot simpler, though I'll be losing the free-floating aesthetic that I enjoyed so much in Auralux.

    Evan said...

    That sounds really interesting. Can't wait!

    ZTrooper said...

    I look forward to your next endeavor. :)

    ZTrooper said...

    One thing I'd like to know, is whether or not upgrading the suns that can be will result in more units per tick of the game clock, or does it simply make the sun harder to capture? Also, was I correct in seeing that the longer a unit is around, the bigger/stronger it becomes?

    Daniel Fox Johnston said...

    Hi!
    So I know this is a little bit of an older thread, but I'd like to throw my two cents out there.

    Let me start by saying I am fascinated by the dialogue you two have had here, and I'm impressed that you both managed to be so respectful throughout the correspondence - the true signs of good blogging. All of Evans ideas are valid and interesting and would make this great game even better... But...

    I would like to give a huge shoutout to Mr. McNeill. I am a huge indie game fan and Auralux has become my favorite iOS app I have stumbled upon in a while. Why you ask with batted breath? Because of its simplicity.

    It takes a very delicate set of feet to walk the tight rope of complexity and accessibility within the mobile game platform (where Auralux (in my humble opinion) is most at home). Where your market ranges from 4 year olds who's parents thought it would be a good idea to buy them an ipod touch, to 65 year olds just learning how to use email for the first time, to nerds like me who eat apps like this up. Erring on the side of simplicity is the ticket! Although the AI's play is sporadic at times as Evan pointed out, it makes the game play all the more interesting.

    Now for a million dollar idea. As a lover of science fiction, I have recently run into an author by the name of Orson Scott Card, a giant in science fiction literature. And if your familiar with his books, most of the battle school games that are described in Card's book Ender's Game, pretty much play out exactly like Auralux. While playing Auralux, I found myself imagining myself as the big-wig operator calling the shots from for an interplanetary war with the dots being the battle ships. While the opponents ships being attributed to the "buggers".

    Let me throw this out there. If there is anyway Mr. McNeill could touch base with Orson Scott Card's legal team and find some way to attach the theory and execution of this game with a Ender's Game theme, you would have literally millions of buyers who would saw off their left arm to be able to play out the last, iconic battle between the humans and buggers. If this has not crossed your mind before, allow it to. Because the attaching of this great game to the context of an engrossing story and giving it thematic material could turn this indie game into a cross-literary work of art.

    Seriously, people would loose their cool if they knew something like that was out there.

    And with the nearing release of the movie, there is already HUGE hype stirring for something like this.

    Great game Mr. McNeill, and great points that could be addressed Mr. Evan. Thank you for sharing your thoughts with us all.

    Keep at it!

    -Fox

    Joshua said...

    I immediately thought of Emders Game as well.

    SirCasey said...

    Likewise thought of Ender's Game when I started playing and would totally be geeked out if it had the branding (not that Auralux is a bad name of course :). Bought all the expansions and played through most all the levels (with 6 or so left to beat) before I found out you -can- have an army but control no stars as long as it's big enough to capture another star. Anyway, just wanted to say here that I love the game too and recommend it highly, looking forward to your future endeavors!

    Post a Comment

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