You Don’t Need Game Art In The Beginning of Your Game Project

I have used 3D models in the very start of a game project. This is not necessary. I recommend the exact opposite: Don’t hire a 3D/pixel/gui/other artist to do graphics in the beginning of your game project.
– The graphics will most likely change anyway – that means the artist needs to re-do his work
– You don’t need final graphics to test your gameplay: you can use placeholders to test your game.

Instead of a human model have a 2 unit tall blue box moving in the screen. Instead of having an elven warrior fighting, have a 3 unit tall brown sphere moving on the screen. You need trees? Use green cones to represent them. Same goes with the graphical user interface: you don’t need animated button to “Start a New Game” – you can have a white box with a black borders and times new roman text: “START NEW GAME” in it.

Use of placeholders will benefit in the following ways:
– It enables you to test your gameplay faster (as you need to tweak the graphics later when the core game is done)
– It saves time, nerves, money or all of them as your artist doesn’t have to re-do his job over and over

Juuso Hietalahti


  1. I agree when you are at the beginning of any game project and wanting to get a feel for the controls of the game, and to test model movement, physics, etc., then basic shapes are all you need.
    If you are working with an established game engine and the goal is to rapidly prototype a couple of levels, then create basic models. These basic representations of the final models could be used as lods.

    Google 3D Warehouse can be a good place to look for basic models.

  2. I think the important thing is that you need art if you want OTHER people to show interest in your game. So you might not need art during the prototyping and early development, but when it comes to testing it’s sure gonna help. My game couldn’t even catch 5 seconds of attention from my girlfriend until the player became a proper character model…

  3. @Harry: Yes, I agree that there are times when it’s not enough. Per-polygon collisions and accurate physics would need more accurate graphics, but even then you can use untextured models which roughly look like the game character. You could use untextured walls etc. for polygon collisions. It is not necessary to bring all art at once.

    I would argue that graphics would take 90% of the time… the most time consuming task in the production is (as far as I have experience about this) definitely programming. Programming game engine, fixing bugs, coding those physics… it all takes a lot time – most of the time. I would estimate coding to take maybe 60-70% (whereas sounds 5-10%, art 20-30%) for an average project where I have been involved. (of course these figures vary… but roughly this is the case). In this sense using placeholders can be good.

    Now, what comes to the question about FPS: If you want to create bullet time, or code artificial intelligence, or gun physics you don’t need much art. You can bring your great ideas in the game without need to use art.

    Of course if your game is mainly focused on graphics (like some adventure game could be) it might be worth bringing in the content & art sooner. It depends.

  4. This is in general a good advice, but there are times that it is not enough. Specialy in modern games with per-polygon collision, accurate physics etc.

    Also today graphics tend to be the most expensive resource (in time and money) for games. If you are making an FPS for example.. what is the point of working with boxes? The gameplay is pretty much the same for most(all?!) games today. It is the graphics tha make the difference…

    The gameplay development is 10% of the time and graphics 90%. So you are basicaly optimizing the wrong part…

  5. I thought you were talking about concept art for a second. Whew.

    I agree with you. If you put the art first then your making a 3D art gallery. Create great gameplay FIRST, then great art.

  6. @Tony: Yes, I agree with this. I remember Lionhead presented their Black & White game couple of years before it was completed. They showed only “moving boxes” and it didn’t get much attention and was commented like “they have lots to do yet” (if I remember correctly). The next year they have built models, added lots of graphics etc…. and this time the press really took it well: “looking really good & interesting… zooming was nice” etc. Polished art will make your game look more complete.

    @Jyskal: Yes, that’s a very true. Very good point!

  7. While it’s true that you don’t need final-quality graphics to test your gameplay, you certainly do need final-quality graphics to *sell* your game idea. Publishers are looking for any reason to say no to you, and “placeholder graphics” will kill your chances faster than a crash mid-demo.

Comments are closed.