Congratulations on your new job as a VR designer at Jedland.

"It's just a bloody bench" the client said. If only it were that simple.

Feb 04, 2025

**It might not be your dream job, but for god’s sake don’t let** me **hear you saying that** | Image : Author

There’s a central conflict that occurs in all immersive design. Let me tell you all about it. It’s a very straightforward and simple concept to grasp, and it’s one of the first things I tell new VR creatives about, because it’s absolutely fundamental to so many design decisions you have to make when making immersive experiences. And I like to introduce it early as it’s a bit of a bastard, as you’ll see.

That conflict is the constant tug-of-war between simulation vs. convenience. You might want to think of it as replication vs re-imagining. Or maybe you’ll recognise it as expected utility vs virtual practicality. And it’s a conflict that pops us regularly in everything an immersive designer does. Don’t worry, I’ll explain.

Whether you’re designing spatial VR apps or experiences, mixed or augmented reality experiences, or even if you’re designing real-life immersive experiences such as immersive theater, or a work that’s going to be shown in a museum or art gallery, this is something you’re going to run up against time and time again, so it’s important to know about it, to recognise it, and to help you make the right decision for your project when this conflict inevitable rears it’s head during pre-production & production.

To talk about this, we’re going to do a little role-play. Don’t worry, it’s not the fifty shades kind, instead we’re going to use the example that you’re designing a theme park in virtual reality. So the brief you’ve got is that me, your client, wants to basically recreate Disneyland, or Universal Studios, or Sea world, Europa Park, Alton Towers, something like that. At a certain point, the client (who is I) gets really excited.

“I’m going to call it ‘Jedland” they proudly announce, and then lean back with the kind of self-satisfaction that only comes from someone who’s just revealed the very best part of their Trillion dollar idea. And you agree with the client (for I am he) and tell him with the most genuine honesty you can muster that “Mister, it sounds like an amazing destination already”.

“Right???” he/I/we says, delighted.

Sympatico. This designer gets it. The client (he me) gives the designer (you you) the gig.

After an evening out drinking to celebrate getting the gig and wondering whether you can trust this devilishly handsome and disarmingly charming client, you sit down the next day with a banging hangover, trying to think about how you’re going to start designing Jedland.

Because you’re a smart designer, you know that this reality vs convenience conflict is going to be there from day one, and it’s probably going to piss off your new boss because it’s not a straightforward thing to think around. He’s a really cool dude, and you know he writes a great Substack, but even so you’re worried that he might get pissy back at you in return. And he (me) don’t like getting pissy, because when he was a kid he once got called Polly Pissy Pants, but that’s all backstory for my character you don’t really need to know.

But it’s inevitable that you’re going to have to have to discuss the same issue many, many times during ideation, pre-production and probably a fair way into the full production cycle. That’s because your design for Jedland needs to choose between recreating a real theme park experience, and creating a convenient and comfortable user experience in VR.

And there are lots of disconnected and conflicting elements that will play into that choice. Many creative decisions will need to be biased one way or the other – towards recreating realism, or towards facilitating something that might be more convenient than reality can offer. It makes a lot of sense to keep this conflict in mind as you move forward and accept it as a constant source of challenges that you will run up against and have to make creative decisions over.

Here's the bullet point break-down of the problem you’ll face.

The design of a Real Life theme park flows from a sympathy towards the comfort and convenience of real life users
It’s built to make the visit as enjoyable as possible within the constraints imposed by the real world.
However, some of the things you can do in a real park won’t be possible in a virtual theme park.
The design of a synthetic, virtual reality should also flow from the same sympathy towards the comfort and convenience of virtual users but now those requirements may be very different.
A virtual park doesn’t need to adhere to as many real world constraints (but does of course need to adhere to VR’s own unique constraints). This means there are new opportunities possible in the virtual realm that can make users’ experience more convenient, easier and more enjoyable.

So as a designer, you’ll constantly bump up against these two dilemmas –

If it’s something you could do in a real theme park, like exploring or riding a coaster, should it be included in a virtual theme park?
If it’s something you couldn’t do in a real Theme Park, like having a 200ft robot version of your mascot doing an extensive improv dance number as part of the parade, do you want to enable it in your virtual theme park?

That doesn’t seem like a particularly challenging pair of butt-pains, but believe me it will offer up its fair share of head-scratching critical decisions time and again while designing this experience. The creative decision-making stakeholders (in Jedland’s case let’s make that me as this century’s suavest and most debonair virtual theme park owner, and you as my immersive designer) will need to consider these two sides of the conflict over and over, and decide which should take priority in each case. And as strange and esoteric this might seem initially, you’ll soon realise that these can be foundational choices with wide reaching consequences.

Firstly, let’s imagine an innocuous element of a Real Life park and think about the conflicts in bringing this to VR.

Let’s think about the potential impact of just a single innocuous element – Benches.

“Benches? Of course. God yes, what a stupid question. Why wouldn’t we?” I ask. Sorry, but in this scenario I’m just that kind of client. To me, it seems barely worth asking whether you want benches in Virtual Land. Theme parks have benches. Duuh. (I mean, I’m role-playing here as someone who’s a trillionaire, so an unsympathetic, self-absorbed arsehole who lacks empathy and politeness towards my employees seems on-point, please excuse me if I’m enjoying burying myself in the role).

So yes, on the face of it, that’s a non-decision, it makes perfect sense to include them. The logic is that you’d expect to see Benches in a theme park. They’re literally part of the furniture. Simple. Obvious in fact. We’ll put benches in and users can sit on them. I can’t believe we’re having this discussion.

“Jed, I respect you so much as a boss already”, you say admiringly, “and as an awesome human being. And that Hagrid thing you’ve got going on with the beard and long hair – that’s really working for you. I wish I could pull that off. You’re a wizard.”

“I appreciate your candor, kid” and we flash you a real Tom Cruise grin. Your knees might buckle if they bend that way.

“But boss, I do have a question…”

“Hit me with it” I say.

“The question is … erm… why?”

The client narrows his attractive, steely eyes at you, strokes his firm jaw, and you wonder if they’re wondering if they’ve made a terrible choice in you as their designer.

Forget what the boss thinks. The designer is asking the right question. Good job, you. Because you realised that your audience will see a virtual bench and expect it to work like a real bench. That’s reasonable. But you know it won’t work like a real bench and it won’t serve the same purpose or satisfy the same needs that the user is expecting. And that means your users will then, inevitably, find themselves pulled out of the fantasy of being in a theme park, reminded that this isn’t real because this bench doesn’t work like real life.

The aim in VR Design must always be to satisfy the user’s expectations; that way users will accept the surreality of what they’re dealing without thinking and remain immersed.

That’s my number one rule of VR. At least for this essay.

IRL, benches are pretty essential for physical, real life visitors to be able to rest and take a personal moment away from the slate of entertainments on offer, whether to put sunblock on the kids, take a call, rest in the shade or figure out how to put all that merchandise into one bag to make it easier to carry.

But none of those physical functions that a real bench provides will be required in Jedland. Because most of those essential functions of a real-life bench don’t provide the same utility in an immersive environment.

Sitting on a virtual bench only affects your avatar, and the avatar doesn’t need any of those things. Your avatar doesn’t get tired, your avatar doesn’t need to sit down or have a rest. They don’t need to step out of the sun for a while. If you want to look at your map, you don’t need to go sit on a bench to do it so that you’re not ignorantly blocking up the thoroughfare. If you want to sort out your merchandise, you’d probably want to do that via some kind of an inventory system -- and in VR like in videogames, an inventory system doesn’t really need any physical space, and it’s something we can open out and get the thing we want at any point simply with the press of a button. We don’t need to be emptying out nappy bags, toys, snacks and merchandise to dig out that sunblock. A virtual bench simply can’t have the same degree of utility to the virtual visitor as a real bench, and it won’t be faced with the same use-cases. So that significantly changes it’s intrinsic value as part of the user’s experience.

This means before you can be expected to implement benches in Jedland, there’s a number of creative decisions that need to be considered. You have lots of choices in how to handle this, and the right answer will depend entirely on the context and situation.

**A bastard of a bench in Jedland. So much trouble.** | Illustration: Author

So you turn to me, and ask which of the following I would like to implement. To save those with sensitive reading skills, I won’t write out the full, very sweary dialogue that would occur in our exchange, but you can probably imagine how our hot (but still looking cool) boss would react to the impossible dilemma you’re putting him/me in when you take a look at the following breakdown of options, questions and considerations:

(1) “Should we just make our benches props only, set dressing that can be seen, but not sat on?”

That’s cheap and easy to implement, but not very immersive for players. They’ll look the part and seem inviting, but they can’t serve the functions as in real life, and might disappoint your players who try to interact with them and find they’re fake benches, just scenery. It might even seem lazy or low on ambition. Users will be like your boss – not as cleverly witty or well dressed, perhaps, but sharing that same inability to see the more complex problem space that you can see with your amazing designer-vision™ abilities.

(2) “Maybe we can avoid the problem - make the benches props only, and then sit Non-Player-Characters (NPCs) on many of them to make it look busy and deter users from trying to sit on them.”

“Genius!” I declare. Then you hit me where it hurts. The characters and animations will need to be created and tested, and that’ll all take money and time to make, and will then eat into the processing budget to display. A measurable and budgetable unit of work, for sure, and it’s achievable. But it’s important to think through the implications of every major and minor design decision in production terms. None of this exists yet, it all needs building and putting together, and the combined mixture needs to run comfortably on the target hardware and come in per your budget.

So what you put into that mix can be a critical decision. You have a limited capacity for creating things, there’s a limited ability to effectively simulate some things, and you need to decide how they’re all going to work. And some of these solutions might add up to a lot of work.

“Dammit. But I see your point.” (Your read of this is correct. I’m playing him subtly nuanced. He’s complex but not a total monster). “What else can we do?”

(3) “So okay,” you say, “we could let user avatars animate to visibly sit on the benches”

“This”, I proclaim. “Wow, you really buried the lede here. Yes, just do this”. I’m visibly rolling my eyes now because I can’t believe we’re having a meeting about this, when I have an important 18 hole appointment for working on my short game.

But you, my great designer, recognize my awesomeness and how I just really command a room, and so you are patient with me, and point out again that it’s not that simple. From a designers point of view, this just takes us down into another nested, multi-choice mystery of possible solutions.

While they’re sitting there, their animations need to work in a sitting pose too, affecting how your anim systems need to be designed.
We’ll need a visual language that tells the user they can sit on a bench, and where they’ll end up. If your user isn’t using teleporting but using other locomotion methods, it’ll still need to work the same.
If the bench doesn’t offer any benefits to users, can you be confident many benches will ever be used? We want to avoid the waste of building systems that few visitors will get value from.
Any nearby geometry and texturing will be under greater extended scrutiny in VR from users sitting there with nothing else to look at, so these spaces may need to take a higher priority from an art standpoint. We could avoid that by perhaps considering keeping benches away from detailed objects as an option – it’s a compromise that keeps the benches in but avoids extra art concerns. But who wants to sit on a bench away from all the interesting stuff?
If the benches do their jobs and users like them, it will affect the flow of users moving around the park. When you’re thinking about processing loads, you want to design plenty of opportunities to take user avatars into smaller instances – such as when they go into an attraction or ride -- or take them out of sight of other users through environment layout, with walkways twisting around and through architecture that occludes them from view as often as possible. The biggest fear for the processing load is going to be a massive amount of real-time characters in view at the same time. Putting in really cool and popular benches could mean a bigger portion of your users are sitting around in full view, each sporting their unique avatar costumes, but not actually taking part in the activities. They become processor squatters, and the more users sitting around like this, the less foot traffic you’ll be able to support. The more users you want in, the more you’ll want to keep everyone moving – just like a real life theme park.
And then at this point, maybe you start thinking that stylizing your aesthetic might get around this problem (and maybe others like it that you’re facing). Sure, but that would mean going back to the drawing board in lots of ways. But if we’re not in full production yet, it might be the best choice -- so this is why you, a considerate and prepared designer, are asking this question now, at it’s cheapest and easiest point.

“Well, as your boss, I know you admire me and I appreciate your intentions, but DUDE, you are making my brain explode. I’m pretty sure this sort of shit is why I hired you as a designer, right? You sort it, I’ve still got time for 9 holes if I drive fast.”

But as we stand up to leave, you’re smart enough to dangle a truffle at me.

(4) “Wait! Before you go – how would you feel about maybe losing the benches altogether?”

We pause at the door, seeming to think for a moment (it’s genuinely hard to be sure) and return, somewhat reluctantly, to our seat.

“OK. That’s seeming increasingly like a smart solution. And a cheap one. Go on”.

Of course you could just lose the benches. Our users clearly don’t need them. But would this mean Jedland will look, well, odd? Unrealistic? Does a theme park space without benches dotted around feel like a real theme park? After all, we started out believing they’re an expected part of the furniture. Now we’re considering just avoiding the issue by taking the benches out back and feeding them into the wood chipper. That does sound appealing at this point. Maybe we can accommodate the loss after all, if we can make up for the shortfall in theme-parkey atmosphere that having those benches would have given us. It might mean we need to work harder elsewhere on the art design front to make up for it, make it look and feel ‘theme-parky’ enough once again – or, you never know, it might just work great and still feel like a theme park without them. Let’s start with some concept art of that scene, but accept that the spatial nature of VR means it’s impossible to know for sure without building the environment and trying it out.

Then the boss has a wonderful, terrible idea.

(5) “I got it! Maybe rather than losing them, we just reduce the number of benches so they’re there but not repeatedly suggesting themselves to players?”

This makes sense with their reduced utility in this virtual park, but fewer benches doesn’t stop users encountering them altogether. The decision how to handle these hasn’t been bypassed, just made less important. So we still need to decide whether players can interact with that bench — and if so, what affordances will we allow?

Can they sit on it? If so, can they also jump onto it and stand on the chair? Can they then transition from that to sitting? If a user is allowed to only sit on it that’s great, saves some work, but then the next considerations are what do we allow them to do while sitting?

Maybe we think they will want to look at their map. In which case there are more considerations to take into account - for example, you’d need to make sure all the seating areas allow enough space for the users to bring their maps up, we don’t want tables and chairs clipping through our Map UI and looking janky. And of course you’ll want to know this up front before you built the environments. Therefore we may decide what we really need is to design and build a comprehensive global seating system to handle all kinds of seating arrangements that scopes in every ride and chair in the park, and define the parameters for size and functionality of different types so that the art guys can get on and build it. It’s a meaty design task to take on, but you’ll have 99% confidence in your current sitting arrangements and a good feeling about any possible sitting needs your avatar guests may face in the future. Plus you’ll then have the option to make yourself feel all boss-like and bad-ass when you say “Check the CGSS doc, it’s all detailed in there guys” to your tech designer and your code guys any time they ask you a question. And if you’re lucky, you’ll get to hear them muttering behind your back that documents suck, the boss is a bastard, and that they hate you too. I imagine it’s just that kind of company.

(6) Next question; if we’re having a sitting state for player avatars, then we also need to decide where else should the user be allowed to sit?

Is it only on chairs and seats? What about on the floor or on low walls? Can I perch on the arm of the bench in that kind of ‘I’m trying to look too cool to commit to sitting down properly, but really I just look like I have old and broken knees’ way? Oh wait, whatabout - can I sit wherever I want on the grass? That sounds idyllic and cool, and Disney probably wouldn’t allow it in a real park so that’s a +1 for VR… but it’s another branch of animation poses, and anyway do we really want users to log in and be met with a sprawl of users all sitting about untidily on the lawns when they enter the game? First Impressions count. And hopefully you’re aiming for this to be successful and pull in lots of users, so your eventual goal is for plenty of people to come, who will be logging in and meeting up with family, friends, their game group etc. Lots of users potentially waiting next to or near the spawn point for their buddies to arrive might become the trend, and this might be the busiest meeting places where frame rate threatens to tank as people arrive in the world - not a great experience! - and no matter what you do you can’t solve for it because forming new meeting habits is something only your users can do.

So maybe just allow sitting in some areas of grass, we could provide dedicated ‘sitting fields’ I guess? They can sit on the grass then - but only where we say so, and that way we can control the lag and design the area around that need. Oh but wait, as a user that would just seem stupid, right? Maybe even oppressive, controlling? You know what gamers can be like. Technically they could sit anywhere because all the work is done, and they’d realise that - we’re just stopping them. First it’d be a Reddit post complaining we’re controlling their freedoms, two days later the common acceptance across the subreddit is that we’re herding them into meadows, and before you know it the lead story on Polygon is about how Jedland is making grass lovers into a segregated minority, it’s a slow news day and suddenly the kerfuffle appears halfway down the front page on the BBC News site and you’re summoned to the bosses office when he gets back from golfing. Ridiculous? Yes. But I figure that only makes it more likely to happen in today’s world.

So if you want to avoid potential rework later down the line, those are just some initial thoughts about the scope of usage across all chairs and benches that must be considered before you can go away start even building a single bench in the world.

(7) And then the actual most important question gets asked : are any of those even the right solution anyway? All we’ve really talked about is the best way to translate benches into something in VR that might replicate some of the real world functions and utilities of a bench, but will fall short of most, if not all. We’re trying to fool the user into thinking this is real life here, and we can use various tricks to maybe emulate the the idea of benches just enough to get away with it. If we’re lucky, what the user gets won’t be noticeably different enough from their expectations, and they’ll accept the compromised implementation without questioning and then move on without it breaking their immersion.

Let’s remember the point of benches IRL is something they don’t need to do in VR. Zero utility there. Maybe there’s some other utility that would be beneficial in a VR world. Can we reimagine the utility of benches in a Jedland?

Maybe users sit on them, but we repurpose the utility of benches for our Virtual theme park so they have a new use? Maybe peeps sitting on a bench as the way to open their comfort settings or read park news. Perhaps everyone sitting on a bench triggers a party-up, more immersive than a menu invite for sure. Or could sitting on a bench pick your seats for when you all get on the ride? Hell, maybe the bench flies off and becomes the ride - VR, right? Or perhaps it’s the only place users are able to access their social menus, thereby neatly stopping hordes of newly arrived users blocking the way with their menus open to check their in-game messages? Virtual spaces like VR Chat and Horizons have arrival spaces so that users have freedom to tend to their menu chores - emails, messages, news - in a private space when they first log in, before they enter the public world, because they know how entering a world of confused users all staring into their menus, rather than enjoying the showcase virtual world, might look to customers - and how stressful it can be to then have other users buzzing around trying to ‘help’.

Granting new uses for real world elements can open up plenty of creative opportunities to surprise and delight users in VR, and the lack of real world constraints mean sitting on a bench could trigger literally anything you can imagine as the designer of the experience.

But if it’s an important element of content you want to expose to users then it might not be the most intuitive way for users to discover and access that content or figure out how to even get that function to work. For all the innovation and cool factor it exudes when they first discover it, in the longer term they might prefer the convenience of opening their map, say, directly from their UI controls, and not need to be at the location of the bench to perform it. In the long term, convenience will win, and then we face the problem that the benches will be barren again and be viewed as pointless, and around we go.

You’ll run up against so many issues when you consider the IRL aspects too. Here’s a few examples: if a user wants to physically sit down IRL to trigger their avatar to sit in Jedland, it will all be lined up and feel amazing when it works, but you need a clever system to enable that without it getting misaligned and the user missing the IRL chair? Or what if they’ve pressed the button to crouch in VR while standing IRL, and then they physically sit down IRL, will their avatar go through the floor? And what if the user does sit down on a real chair, but their avatar is not near a virtual bench to sit on - will they be seen sitting in midair in VR? What happens if the standing user walks 4 feet over to the left IRL while the avatar is supposedly in a static seated position? Or what happens if the user is seated on a real chair, but they shuffle it a few feet over to get away from the edges of their playspace?

Et cetera, et cetera. Possibilities and choices. A lot of potential implications for workload, design, processing overheads, user intuition (and thus the all-important quality of the user’s experience inside the space) and many other elements. But all those are decisions you have to make early on as a designer, to create some foundational rules. Because if those rules aren’t in place, things can get wacky really quickly. And you’re expecting plenty of other weird situations you haven’t thought of to crop up around your seating system as you work through pre-prod and production too, of course. Every decision might be important. Tiny stones, but big ripples.

You’re frantically explaining this but when you look up, I’ve already Bugatti’d out of there 15 minutes ago and left you a note.

“You were waffling and life is too short. Do your job, solve this yourself. I have to go deal with more important handicaps. I trust you to make a decision, but obvs just don’t fuck it up.”

Great. Thanks boss.

These all seem like potentially minor decisons in isolation, but collectively they’re all ingredients that define various parts of the user experience, and touch and affect other parts at the same time. In implementation terms they’re all systems that need to be designed, built, tested and maintained. And you’ll notice the design part comes first there. It’s not at all unusual to find awkward, immersion breaking design decisions in VR apps, and I genuinely believe that a lot of these will have happened because important decisions got made too late, and design constraints that should have been identified before kick off are instead being recognised and imposed late in the final period, with the clock already running out.

The difference between a really good, robust Immersive experience, and a so-so one, can often be seen in how they handle the little things. Something like the exemplary Half life: Alyx has got really-well considered answers to all these little things (born of long development time and a complete development reset once they’d realised the approaches they should have taken) and they add up to a very solid, very realistic-feeling experience as a whole, which in my admittedly subjective but professional opinion (let’s not forget who’s your boss here) it manages to sustain for long periods without the player ever breaking out of their immersion and remembering the real world - your mileage may vary but on the design side I think it’s a master class in considering what the user will be expecting to do at any given time.

Other projects will be unlikely to have that luxury of development time and a do-over, of course, but that doesn’t mean you have to build these things and find them lacking in order to learn from them. So many of these questions and choices - the vast majority I would say - will be addressed inside the designer’s head. If they look pained and deep in thought, they may be thinking about something the way we just thought about benches. Trying to consider all the angles and fit the pieces together tightly is what the designer is all about, after all. Trying to understand the player’s thought process and predict how they’ll react when faced with any virtual object which has a real-world analogue is part of the job.

He walks in and breaks your train of thought.

“Right, good first session. Well done. I’m not sure we arrived at a resolution for benches but that’s okay, we’ve got to push on now to talk about whether we need a monorail to get people around the park or not. Monorails work the same in VR, right?”

He/I gaze out at the crappy, drizzly Jedland skyline as the sun heads lazily over to flop in an ungainly heap behind the horizon, done for the day. As the last rays of sun fade, light seems to catch in the man’s devastatingly disarming eyes. For a moment, that’s all you think it is. But as the daylight fades, the sparkle in your bosses eye does not. He turns back and stares at you with a look that speaks of the kind of self-satisfaction that only comes from someone who’s just revealed the latest very best part of their Trillion dollar idea. He checks over his shoulder before he leans in.

“Rocket Powered Golf Carts” - words he gifts to you in a loudly confidential whisper, orating each word carefully so you can better admire their shining brilliance, the words hanging in the air between you because they honestly don’t know what to do with themselves.

“Mister,” you say, “that sounds like an amazing idea already. I do have a few questions though…”

Sympatico. Right? This designer gets it.

Until next time,

Jed

Discussion about this post

Ready for more?