Terminology

The only interface design you presented that I have always quite enjoyed is TF2. I hate how "immersion" gets thrown around in the latest years. Dead Space is a 50/50 for me. The health and stasis bar are great because their overall design is solid and makes sense while at the same time fitting the game universe, but the rest is just holographic versions of things you see in every vanilla interface design. And the map sucks.

Perhaps looking into the V/H Field of View that is needed across platforms would be a good idea. A low/narrow setting is fine on consoles but terrible on the PC (all down to distance away from the screen). Getting the simple matter of the viewpoint correct is usually a good idea. FC2 and Deadspace both had issues with this, while TF2 allows you to change it at will which means players aren't inconvenienced if they suffer from the effects of a low FOV



And DICE need to work on their front end UI's much more than their in-game ones. I dont think any of their games have had a menu that could be considered good by anyone but a sloth. Pretty? Sure, but the usability (or complete lack of) leaves a lot to be desired.

I'm surprised that the Metroid Prime series wasn't mentioned. It has quite an elegant solution, because it's perfectly natural for Samus to have an in-world HUD. The HUD is there for the character's benefit, which helps immersion and makes the game more believable since the player and character have access to the exact same information.

Maybe my definition of immersion is different, but I would argue that WoW's interface improves the game's immersion, not impede it. Unlike the traditional clairvoyant approach to UI design, Blizzard gave people a framework, then allowed them to alter it to their particular playstyle. Control over the game world is the key to immersion, not direct visual "property."



The impression I got from this article was that adding UI elements detract from immersion. It's like saying that driving a car would be improved if we took out the dashboard. Access to key gauges -- speed, fuel, temperature -- would enhance immersion, since you can tell how the car is performing as you are driving. If our only concern was the driving experience, I would say that it would get very boring quickly.

I read an interview with the TF2 designers not long after the game came out; one thing Robin Walker said that really stuck with me was how they realized that dropping realism and embracing a "cartoony" style allowed them to do many interface/communication tricks that they couldn't have justified otherwise. And in fact they seemed to consider it a major priority to put the information "in the world" as much as possible, with "non-diegetic" elements being used only as a secondary and less reliable channel of communicaton. Walker's key example was the masks that Spies wear to communicate what class they're disguised as.



Found the interview, here's a link to Part 2 of it, with the part that I was quoting: http://www.rockpapershotgun.com/2007/10/03/rps-team-fortress-2-interview-part-2/

Finished the article. Really enjoyed it, especially the last page. The "gun on a box" image is a powerful one for reminding us of the sometimes-dramatic difference between what a game looks like and what the game mechanics actually are.



I think the "prosthetics" analogy is even more powerful, and important to remember in not only UI design but all game design. But I think the choice between whether to make a 'realistic-looking prosthesis' or a 'highly-usable prosthesis' actually might be a balance that differs from project to project. For instance, in Dead Space it was crucial to make the experience immersive - it was a survival horror game, and much of the point of it was for you to fully occupy your character and be frightened and stressed when your character is in danger. In WoW on the other hand, immersion is an extremely low priority (after all, any attempt to maintain it would be constantly broken by Chuck Norris jokes showing up in your chat window - the "Golden Circle" is extremely porous in an MMO), and so they made the correct decision in making the "prosthetic arm" be one that did nothing to enhance immersion, but was not only highly functional, but easily replaceable.

This is a very interesting article however I question the conclusion that the developers should strive to make the player feel that he or she IS the character.



In his book, "Shared Fantasy" Gary Alan Fine argued that players will readily switch between frames of reference. At one moment identifying as the character, the next identifying as being separate. In either cases a player can be fully immersed in their experience. It's possible to observe the same phenomenon by watching anyone playing a video game that includes an avatar. For another example look lets at movies. A viewer of a movie can be fully immersed, while at no point feeling that they ARE the character on the screen. My point is immersion through avatar embodiment is not the only, or even the best, method produce to an immersive experience.



I fully agree with your conclusion of functionality preservation. At the most basic level a game interface should strive for simplicity of controls and interaction. The use of an interface as a technique to immerse players should always take a backseat to usability issues, because these issues may prove more detrimental immersion than a non-diegetic HUD.

I wonder to what extent the type of UI preferred for a particular game depends on the kind of gamer who is the intended audience for that game?



One of the distinctions I've come to draw among types of gamers is between those who enjoy "playing in" a gameworld and those who primarily enjoy "living in" a gameworld.



The former kind of gamer is there to win. There are rules to be followed which result in scores achieved. For these gamers (and thus the games they prefer to play), the look of the interface is less important than its utility. It doesn't need to fit the gameworld in any coherent sense; it just needs to support winning. If the UI is so "immersive" that it gets in the way or makes it hard to grasp necessary information quickly, then it's a fail for the "play in" gamers. It doesn't have to be ugly -- it just has to communicate key information as rapidly as possible.



The "live in" gamers might be OK with playing a rules-based game to win. (They'd better be, since virtually every game is designed under the assumption that this is what everyone wants.) But the more a gameworld is designed to be logically and emotionally coherent in all its elements, the more satisfying the play experience will be. These gamers will prefer a UI that makes rational or emotional sense in combination with everything else in the gameworld, and that doesn't constantly remind them they're playing a game. Utility-driven UI elements that have no representation within the fiction of the gameworld will make the game less fun for these gamers.



So form and function both matter in UI design. But I suspect that each matters to a different degree to different gamers.

@ Bobby Breen: The viewpoint is a good example; especially we often develop with a TV on our desk a few feet from our faces even when we develop console games. Something to keep in mind.



@ David Gates: Agreed, Metroid would have been a good example in the article, can’t include them all… Metroid is a great example of where the fiction actually makes a traditional HUD diegetic just by context. It’s also a good example of empowering the player in the game world, actually giving information that is needed to assume the part of a super soldier.



@jaime kuroiwa: Your definition of immersion is what I would call involvement or commitment – or even appreciation. The above things can happen if the interface is performing its job of interfacing the player into the game world. Obviously the game has to be fun on top of that. My definition of immersion is the “suspension of disbelief” where you really forget yourself and feel as if you are in the game (something I actually don’t believe in). I think it would be more constructive to pursue involvement, commitment and appreciation rather than “immersion”.



@Shay Pierce: Great feedback! I think you are spot on; I want developers to think about this, the end design has to be custom tailored to fit the game you’re making. However, I strongly believe that your first priority must always be to enable the organism to be aware in the game world, something that Dead Space did by having fairly traditional UI (although holographic) and Far Cry 2 by superimposing extra information.



@Ian Livingstone: I agree with you and I mean that the developer must make sure the interface give the player “organism” what it needs to act the part assigned by the game. Let me give you an example: Say you play an elite soldier in an FPS. The FPS perspective inherently denies you any clue of what’s going on behind you, together with all the other disabilities of being in a game world (lack of tactile sensation, awareness of pain, etc). An elite soldier is likely very adept at picking up dangers approaching from behind and are in general pretty good at sensing potential dangers in the near surrounding. So if you play said game and get shot from behind, point blank by a machine-gun wielding terrorist the UI has failed to make you feel the part of an elite soldier.



@ Bart Stewart: You bring up a big issue but it is my belief that you can hardly do harm if you try to make the players feel comfortable in the game world. You always have to adopt your UI on a per project basis and two great UIs can look and behave vastly different.

Very thorough analysis, would be interesting to see something like this oriented toward minimalism and experiences other than feeling "badass". Like the use of greek terms. Perhaps that's telling, the term "diageses" means "to tell" or "to narrate" while "mimesis" means "to show". Perhaps there's another school of interface design we're missing.

Very interesting. I have to agree with many points in this article. More and more we are inspired, or even pressed to believe on unecessary urgencies of environmental (diegetical) visual interface requests. Most of the times, all that is really necessary, is to offer comfort and reliability to the player.



Priming for usability is key to any interface design in my opinion, be it a software, or even a cup of coffee (as a physical interface). Aesthetics could come second, even if strictly necessary for the definition of a product. It must though, whenever possible, add "immersion" to the user's experience.



Suspension of disbelief is also something I do not tend to believe totally in. I'm surprised to see I'm not a lone sailor in this ship.

@stone bytes



I completely agree that it is highly unlikely that 'real' military forces would ever use high-visibility, glow-in-the-dark neon lights on their combat gear, especially in a stealth situation, but look at the design of Sam Fisher in Splinter Cell - he has three lights on the front of his goggles and one light in the middle of his back. Yes, in a real situation, he's instantly visible from any angle in the dark, but in the game how else would you know what way he's facing when sneaking through an almost pitch-black corridor?! I don't think that was a frivolous oversight by the character designers, but rather a deliberate reality trade-off for the sake of enhanced playability.

@Patrick Dugan



I think we have crossed into the age of "mimesis" - showing is much better than telling. One example of this is stamina and health. Both of these would have once had a UI element to tell the player how much was left. This now seems pretty dated - your stamina lasts until your on-screen avatar is tired, your health lasts until you die, the severity of your predicament is shown to you through graphical overlays, changes in colour tone, changes in sound; the mood and tone of the presentation changes to reflect the players physical condition.



I wonder how much further this could go and whether reducing UI increases immersion. Consider a first-person shooter without a UI element for ammo count. Of course, totally withholding such information from the player would be both unfair and unrealistic, but if subtle clues are provided they could potentially provoke a deeper connection between the player and their avatar. Possible "fuzzy" UI could include guns that click with each shot as the clip is running low or different reload animations such as reaching into a back-pack for the next clip when those on the belt have run out. Player connection with NPCs could also be enhanced by removing this UI - NPCs could provide more explicit information such as chastising you for hoarding a ton of ammo or telling you to be careful when low.



Most importantly, though, the choices should be right for the experience. None of the above ideas would work in a fast-paced mechanic-heavy shooter like TF2 where the situation must be made clear at all times, but for survival games which promote the experience equally as much as the mechanics (e.g. Stalker, Far Cry 2, or even Resident Evil) I believe a degree of obfuscation would be interesting to explore.

As a newer game designer interested in current design trends and exploring today's challenges in the field, I was drawn to your article analyzing the benefits and drawbacks of diegetic user interfaces in specific games. I appreciate the range of genres which are covered by your examples, which goes a long way to highlight how one method may shine in a particular genre and be much less useful in another. I also agree that although diegetic interfaces can create a more direct connection between the player and the respective "organism," the multitude of player commands and abilities in some cases (such as many "massively multiplayer" online games) outnumbers what you refer to as "real estate" within the game world. This idea of the organism is compelling, and although the concept itself is always at the center of decisions regarding a game's user interface, it seems extremely useful to encapsulate it with this metaphor. Furthermore, the comparison of the organism to a human prosthesis is similar to the "mental model" described by Tracy Fullerton and Chris Swain as an extension of the player's perception of their role within a game in their book, Game Design Workshop. According to Fullerton and Swain, this mental model "can either help players to understand your game, or it can cause them to misunderstand it." In relation to the article, I interpret this claim to resemble the patient's understanding of a new prosthesis with preserved functionality versus one which sacrifices such functionality for "authenticity."



The article mentions during the "Far Cry 2" example that the combination of both diegetic and non-diegetic interface elements makes the game feel "conflicted." Do you think that this property holds true as a general rule? For example, would you come to the same conclusion for a game such as "Grand Theft Auto 4" which contains diegetic elements such as the cell phone and non-diegetic elements such as the world map? I certainly agree with your statement which says that a compromise between the two methods is greatest "if a diegetic interface is the goal from the beginning." I believe the same to be true for a non-diegetic interface. However, was this not the goal in "Far Cry 2?" I think one of the main issues of its interface may have been in the non-diegetic elements being "forced in" rather than the diegetic ones as a result of the games "lack of real estate." Perhaps it is also useful to explore this question on a grander scale for upcoming technological advances that are expected to be seen in games such as 3D viewing and Microsoft's Natal. I'm curious about your thoughts on how these technologies will affect both the usefulness and importance of diegesis in game user interfaces.

Good article, good discussion. ;)

@Michael Grattan



I do not think that the conflicted state of Far Cry 2 holds true as a general rule. I encourage interface designers to use any method as long as it benefits the "organism".



The issue with Far Cry 2 is that they have clearly wanted to make the interface diegetic but failed to make the game playable without the complimentary HUD. FC2 is an example of the "wasted" effort in trying to enable the player by diegetic means only - despite their best efforts.



Thank you for your interest in my article. I wish you the best in your future endeavors!



best regards

Marcus Andrews

Great article. Luckily, there is no annswer to "what is the best UI design for games". Also happy not to read "10 thumb rules" or something like that. What UI design works best all depends on the game you are creating. The only thing that is extremely important is that the UI designer has to make the game usable. Wonderfull features are uninteresting when they aren't be used by the player. It is not that much different from anything else (user electronics). This sounds obvious, but I think there are many games that could do so much better.

thanks for this...hopefully people shouldn't get confused with User Interface, HUD and Graphics User Interface as they have their own doings and meaning.