Is this doable? (Not a "can I do this in Unity?" thread, I promise)

Is it possible for a one man team to create an open world RPG like Elder Scrolls in 3,000 hours (about 1 year's worth of work)?

Granted, you would need to cut a LOT of corners.

Namely: cut out all character customization, use extremely simple facial animations, use lots of store bought assets, do almost none of the art myself, and cut out dialog based on stats and turn most NPC's into "sign posts".

But, keeping a character controller with detailed RPG elements, crafting, stealth, spells/skills, day/night cycle, a quest tree/web, NPC schedules, and of course all the other standard RPG stuff.

Also, I'll go ahead slap together a "to-do" list and estimated time based on people's input.

 

RESERVED

Spoiler

Core Game
Main Menu
Saving and Loading
Settings

Character Controller
Movement
Stealth
Attacking
Object Interaction
HUD

RPG Engine
Core Stats and Level
Combat Skills
Crafting Skills
Reputation (maybe)
Quests Log
Lore Logs

Map
Enemy Spawning
Asynchronous Loading
Time of Day Cycle
Weather

Items
Inventory

NPC
Schedule
Behavior Tree
Dialog Tree
Bartering

Enemies
Patrol Path
Behavior Tree

Design
Over World Map Design
City Design
Dungeon Design
Quests
Dialog
Opening Sequence
Ending Sequence
Publishing

 

All the things you list here are doable in a weekend, they mean nothing on their own. There are tutorials out there showing how to create a fully fledged inventory system in a couple of hours yet I've spent probably a couple of hundred hours on my own inventory system and I'm still far from happy with it. Unity has an official 1-hour tutorial showing how to create an ability system when this guy can spend close to a hundred hours on just the VFX for a set of abilities, with no coded functionality whatsoever, and he's extremely proficient at it, I would probably need 2-5-10 times as long to do what he does.

In my opinion the only thing that requires time in game development is polish, and an idea is worthless on its own, the only thing that matters is the implementation.

I've probably spent a thousand hours on my RPG project and it's still extremely bare bones.

 

Aside from such an obvious answer like 'it depends on your skills', I believe the answer could vary according to how you define the word 'open world'.

Even if we don't consider adding a modding support, Skyrim platform provides quite many useful 'procedures', 'packages', and 'conditions' to make such open world interactions possible, which would take tremendous time for a single developer to duplicate.

Probably, there might be a few premade events or actions your behavior tree asset provides, but I doubt the number would be comparable to what Skyrim provides.

Aside from that, there are a lot of other things, like furniture system, idle markers, relationships, scenes, magic effects, impact sets, and many others in Skyrim platform that might be needed depending on how you define your scope.

So, if you really want to use Skyrim as a reference when you create your own game, then I'd suggest you to take some time to play with the Creation Kit and read its documentation first, as it might help you to define your scope more clearly.

 

I'd say yes...

But with this restriction I'd say probably not, unless "one man team" means "one dude willing to spend 10 million dollars".

Make it 5 years, drop any ideas about having any kind of realism, stick to super low-poly graphics (minecraft/another world style), then there will be a chance.

That's the way I see it.

 

Here's the problem with:

It only covers code. Not content. And guess what? Content actually takes longer to do than code. Code... for as hard it is with this subject, is also the least time consuming part for an MMO or even RPG. The time consuming part is the content.

So have a think about that. Even designing the story and what people actually say will take much more iteration than just integrating dialogue code. Then multiply that by the amount of story and number of characters and branching and so on. It's utter madness to imagine it will take a year.

It'll take a year to get all the dialogue done let alone the rest, for one person.

 

You can probably create a crappy open world RPG that no one wants to play. But a good one? Not likely. It took me several years to create a non-open world RPG that is less than 10 hours long. But then again, I sucked and had to learn a lot so I'm probably not a good example.

 

This is true. I need to work on the list for sure.

And I agree that content would be the big stuff.

Coding would be about 1/16th. ~185 hours

Rigging Prefabs another 1/16th. ~185 hours

Making the entire world, including maps, cities, and dungeons another 1/8th ~375 hours

Then populating it with spawn points and items would be another 1/8th. ~375 hours

Then the quests would be 1/4, dialog 1/4, and cut scenes 1/4. ~750 hours

Or at least my best guess.

Sound about right?

Maybe not, I also suck and have a lot to learn too.

 

I would suggest taking what is that you love about the game you want to make and trim all the fat until it's something that you can manage You don't need a lot of dialogue if there are not a lot of characters. And you don't need a big world to make a fun world. And so on.

 

Sounds wrong.
10..25% of time coding. 50..70% of time making art assets. 40% placing those and stitching them together. And another 400% polishing.

Coding alone can easily take 2..4 months.

 

I plan to do something like that, barebone, strip down copy of huge games, to learn to manage the code as it is my main weakness. Can't say anything about time lol.

 

What takes more time, at least on my experience is the huge amount of art for only one man to create, if you are willing to cut this from your to-dos and buy them, I think it's pretty much doable!

I was working on this game called Creacorp, it's a game inspired on Pokémon, not going into details but, it's 3d, everything was fine in the beginning, I was adding the features pretty fast, until I started to make a lists about what I was going to need to model, my milestone (which was a playable demo) had 5 maps in it, it was not open world but connected maps, when I was writing down this list it started to grow and grow and grow and I started to get scared, all those houses, and each house has lots of things in it also all the outdoor decoration, man-made and nature-made, characters, usable items, the list was huge...

But you also need to take one thing into consideration: iteration, you will not add lots of features in your game and let it there as it is and think: "This is good, works and it's funny, I don't need to tweak any mechanic or values in it". Do you already have a GDD? Or are you going to make exactly like The Elder Scrolls?

 

tbh it all depends....when as example using the help of a lot of assets from the asset store, it might actually be possible to remove a lot of the coding aspects. sure you have to learn those programs but it might still be faster than coding everything from hand...and graphics wise etc. one could try to go with simplest possible, like I created a way for me to create assets relatively fast by making heavy use of the lasso tool in photoshop and not caring that some edges might look jagged or weird, since, when keeping it consistent it might give it certain flair...>.> *coughs* but I do have a background in art....
quests....could be problematic if wanting to write one coherent and interesting story, but if you would make it more open world and less story based, one could easily auto generate quests like "kill X amount of animals" or "bring x amount of logs" etc.
I think a combination of automization and simplifying everything would make it likely possible with one year timeframe.
but....let me make my open world survival game in 3 months...then I come back and tell you how it went lol, but it's really stripped down game mechanic wise and trying to focus on my strengths...

 

It's utterly impossible to the point of being laughable.

Making something "like Elder Scrolls" requires realistically around 100 men working for about 5 years. For the record: in making of The Witcher 3 1500 people participated directly. Yes: ONE THOUSAND FIVE HUNDRED employees/contractors.

Also 3000 hours is not one year of work but close to 1.5 years, unless you are working 12 hours per day.

 

They are actually less than 100 (50ish) people for skyrim, and had only 8 people dedicated to level design. So it's not laughable as the engine is already given.

 

I think they could do it because they reused the same engine from Oblivion, which is much higher level (in abstraction) than Unity.

 

No, you cannot make skyrim solo in 3k hours.

Yes, you can make a very high quality open world rpg in 3k hours. As in, commercial quality, higher end experience.

No, you cannot come close unless you are already immensely skilled, well versed in absolutely every aspect of game dev, and never make a moderate scale mistake. You basically need to get every aspect right on the first pass.

Only realistic devs who might be capable of something like this would have already worked on (to the point of release) multiple full feature, commercial grade open world RPGs. You must be a specialist with close to a decade of previous experience in the specific field of making open world RPGs to come close.

"I did level design for Skyrim" <-- not nearly enough
"I worked as core generalist on every elder scrolls title in the last decade, and also worked as core team on witcher" <-- now we're talking (I don't think that roles like "core generalist" exist.. so maybe this person is impossible).

If you are that person, then yes, I believe one man can achieve this goal in that timeframe.

Similarly, can a programmer write a mid sized system with 100,000 lines of code that never has a bug or compilation failure? I would answer yes, a dedicated specialist in a very specific field with enough experience can achieve that.

 

To be frank, he already dismiss quality for scope, and I'm pretty sure his quest redux mean no implementation of radiant AI anyway. So given quality reduction of that level, it's possible, use pgc

 

Dude, do you read what you write? "8 people dedicated to level design" FULL-TIME FOR YEARS. Do that solo and everything else in a year. XD

 

I was talking in the context of the quality reduction mentionned in the title.
I think when people say can I make skyrim, they don't actually mean all system and depth of the game, they basically see the scope (as in many village and town and plenty wilderness) which is something possible to do fast, the faster the lower teh quality with copy pasting as the lowest point. But scope with those condition is not a problem at all. I mean the OP spend time enumerating everything he would cut, that's the frame I answer to.

I'll stay away from now of this strand of the discussion. It's useless to continue without actually doing it.

 

People don't buy scope. People buy quality.

 

Can you do it in 3000 hours? Maybe...

Will it be worth playing to the average consumer if it was done in 3000 hours? No

 

You can't. I can't. But I think there are people who can.

The thing is, those people will be deeply experienced and well connected. So they won't need to. They'll be able to raise funds, assemble a team, etc.

That's kind of the problem, anyone who is really that good will have established themselves as a known quantity and won't need to execute such a bare bones project. At minimum they'll be able to put together a small team of like 3-4 (the ideal team size for ambitious indie).

It's not possible because you're a magical snowflake, it requires you to be a magical snowflake who has also actually worked on multiple projects through to release and learned from other magical snowflakes on the top of the game.

 

I don't think selling it was a factor, just possibility.

 

He never said which Elder Scrolls game. Here is the team responsible for Arena and Daggerfall. Daggerfall took 2.5 years.

 

I want my money back.
Those things that came out at night in that starting city were mean to me.
And those damned shop keeper threw me out at night or attacked me if I stayed.

 

If you want to do something, take a shot at it. If you ask others, they will always say what they are capable of, not what you are capable of.

Remember everything is undoable until someone goes ahead and proves that it can be done.

 

Boy, no matter how explicit you are, people just hear what they want to hear and argue with whatever strawman they make...

I would make deep cuts to the systems to trim the dev load and I would use almost exclusively all store bought art assets (and maybe some hired work).

The way I see it, there's lots of people paying good money for walking simulators. Throw an RPG in it and you've got a good shot at a good product.

I'm kind of thinking that doing photo-realism is just as fast and cheap. On one hand, low poly means that you don't need gross details. But on the other, with photo-realism you're just taking pictures of real life.

Either is a good avenue though.

I have an RPG engine I've been polishing for years and a lot of great quests in mind.

I don't know. Have you seen Dwarf Fortress?

Also, if you read my first post you'll see that I was talking about cutting back the scope dramatically.

 

Don't start a genre you don't understand ROFL, walking simulator is all about environmental storytelling, it's the polar opposite of what you want to do.

The money they save from having gameplay system they invest in highly intricate story and level design. And the genre have moved pretty sophisticate now, Passage was the proof of concept, Kuleshov effect style (google it), dear easter was the proof of execution, amnesia the proof of atmosphere, Gone Home the proof of structure, Stanley parable being the overall genius work and what remain of Edith Finch is the evolution. They are basically adventure game without S***ty non sequitur that puzzle used to be, instead they rely on story as a puzzle, hence why they opt most of the time for obtuse storytelling and mystery to draw player in.

They rely on DENSITY of details, don't make a noob mistake that people walk for the sake of walking, they walk for the sake of discovery and inspecting the nook and crany of the world, to let the environment soak in them. Don't expect to sell or have an audience for a cut down version of anything most of the time.

 

It's totally possible for one person in a year to do what you asked if they cut a lot of corners, had mostly repeated content, used existing assets from the store, didn't make any mistakes along the way and didn't care about fixing any bugs.

However that person is incredibly rare, and anyone who needs to ask "could I do that?" almost certainly couldn't.

 

There are definitely some good quality assets on the store these days. I remember mentioning the Polygon packs from Synty Studios a while back but I'll mention them again. My inclination just from glancing over these five is that the dungeon pack and the knight pack would work well together.

https://www.assetstore.unity3d.com/en/#!/content/80585 - Adventure Pack - $19.99
https://www.assetstore.unity3d.com/en/#!/content/83694 - Knight Pack - $19.99
https://www.assetstore.unity3d.com/en/#!/content/89551 - Samurai Pack - $14.99
https://www.assetstore.unity3d.com/en/#!/content/85664 - Viking Pack - $19.99

https://www.assetstore.unity3d.com/en/#!/content/75980 - Simple Dungeons - $16.99
https://www.assetstore.unity3d.com/en/#!/content/76478 - Simple Fantasy Interiors - $16.99

Compared to those packs photo-realism is much more expensive in my opinion. Manufactura K4 has some of the better packs but they're consistently two to four times more expensive than Synty's packs.

https://www.assetstore.unity3d.com/en/#!/search/page=1/sortby=popularity/query=publisher:585

Sometimes you find really inexpensive packs though. Like this one that has over 700 assets. It's on sale again too.

https://www.assetstore.unity3d.com/en/#!/content/7101 - $19.99

Every open world game by Bethesda has hundreds of minor problems. I'm not suggesting you simply ignore bugs but at the same time their community has been willing to live with them. Much of this is likely due to how other companies seem to completely ignore the niche the games occupy.

 

The thing is, if you cut scope enough, you can make any game in any given time frame.

Trouble is, scope is what defined the Elder Scrolls games. If you cut the scope down, you really can't call the game 'like Elder Scrolls'.

If you don't want people attacking your straw man, don't put one up in the first place. There is no way your cut down scope will be anywhere close to 'like Elder Scrolls'.

And let's face it, you come up with a new idea like this every month or so.

 

Oh no, I get that. I didn't mean any disrespect to the genre, either. Although I can see now that it sounded like I did.

If I were to ever do this one thing that I would DEFINITELY spring for is some solid set pieces. All of the big cities would look like a Jame Gurney painting.



I know that you can't just throw in random crap, dust your hands, and call it a day.

No, I would most like split the game into five regions with unique biomes. And I would avoid long stretches of single textures like the plague. And I would make sure to keep in mind verticality.

What I would do is most likely use something like World Machine to get the broad strokes in place, then place designate "set pieces" peppered throughout the map(s) and make sure that all the main cities have a uniqueness about them.

 

Oh, I meant that art assets came up and I had said that I would use almost exclusively all store bought assets or work for hire.

True.

 

You might be able to save a considerable amount of time with a visual programming asset like GameFlow. I'm only bringing it up because it was pointed out in another thread that it doesn't have to compile when any changes are made which may save a considerable amount of time.

https://forum.unity3d.com/threads/interested-in-visual-scripting-try-gameflow-for-a-change.471455/

Outside of content linked to the main story I would make the bulk of the dungeons via procedural generation. Daedalus would be an excellent asset because it can be used during design time as well as runtime whereas most generators only support one or the other. It could very easily save time when building more static dungeons.

https://www.assetstore.unity3d.com/en/#!/content/14492

Once again a mixture of procedural generation and static placement.

Just discovered that @AdamGoodrich has an asset for creating cutscenes. His assets are practically a requirement for creating an open world game very quickly.

https://www.assetstore.unity3d.com/en/#!/content/65397

 

Now this is going somewhere lol, even though we are still theoretical.

 

Why thank you

Gaia 2 progressing really well and a new terrain shader will be launched in the next few weeks. This is a standard 4 texture unity terrain in the new shader

 

Alright, NOW we're getting to the real questions.

To be sure there are LOTS of terrain editor extensions out there and yours is way up there on my list.

But to be honest, as a customer, I always feel exhausted and over run trying to piece together the right assets to use together.

I could see using World Machine OR Gaia. The trouble I'm having is I'm not entirely certain how well either will work with some other assets I'll need, such as RTP, easy roads, a terrain stitcher, or Landscape Auto Material.

Then on top of that I need to think about time of day, water, skyboxes, and weather.

To be frank it's is all very, very dizzying.

And at times I also wonder if it would be best to just do the entire map in Mudbox in tiles.

So, why Gaia?

And what addition assets pair well with it?

I for sure will want alpha blending,
triplaner mapping,
holes,
stitching,
roads/streams,
water system/renderer,
decent fog,
decent grass,
volumetric clouds,
weather,
time of day,
a good terrain renderer that lowers tiling,
tessellation would be nice,
the ability to paint height and textures at the same time,
options to do texture by height and slope,
objects (such as grass or rocks) automatically painted onto textures (like LAM),
a good work flow for working on multiple terrains (I would like to be able to paint on two terrains at once),
and this is pie-in-the sky but I would LOVE the ability to make broad strokes in 2d like you do in World Machine.

Sorry, that's quite a list, I know, but what would you recommend I get with Gaia in order to achieve everything I want to do here?

Also, sorry if you answer some of this on the store page but currently it seems to be down.

EDIT: Okay, I just found your list of recommended assets on your site.

So Aquas or Suimono for water,
Volumetric Fog & Mist,
Horizon Based Ambient Occlusion,
and Distingo.

AND I see a lot of the features I'm looking for are supported by Gaia.

That still leave me hanging about alpha blending and triplaner textures.

RTP handles these. Does Gaia play nice with RTP?

 

Yes.

World Machine is very popular with some developers but the work flow involves bringing your terrain in and out of WM and the engine you're developing it for. If that weren't enough you have to consider that WM uses a noise-based procedural approach. Back when I last tried evaluating it it felt like I had to struggle too much to get anything I liked.

Contrast this to Gaia where the actual terrain creation takes place within the editor and is built around assembling stamp-like templates and blending them together to get the result you want.

That said it doesn't have to be an "OR". You could very easily use World Machine to create stamps for Gaia or export the heightmap from Gaia and bring it into WM to add features to it like roads.

 

This has all been very useful. Thanks!

 

Coding something "like Elder Scrolls" in.. 185 hours. LOL!

You said nothing about NPCs. You need dozens of them. They have to be modeled, textured, rigged, animated and scripted. Don't forget particle FX. You say photorealism is cheap? You are not serious...

 

Man, I thought this was talking in hypothetical.

This sounds like a bad idea to me @Not_Sure

You seem like a good dude, don't do this to yourself... lol

I remember reading about some dude who built a low end Skyrim style open world, it seemed somewhat competently built although clearly mega budget.

What's the end game, do you want to sell this or just do it for giggles?

If you want to sell it, then I don't recommend this path, getting a sale worthy title is ... that's going to be a tough road. If you just want to do it for giggles, then cool, but don't expect much.

 

Well, I'm mainly looking at a planning phase for my next project after Bravely Bear is wrapped up.

I think that it's more of a matter of what needs to be reeled in and trimmed down.

So if this were an MVP, what would that look like?

Enemies, an over world map, some dungeons, an inventory, some spells/skills, a leveling system, and maybe a shop.

Surely, that's possible.

 

The problem is, this will be a bad game. Getting it to be "good" is another story... and that's where the nightmare begins.

 

If you're going to go down this path, I would strongly suggest not looking to make it "sale worthy" and to stay in the "for fun" category. You can totally make a "for fun" open world game in a year. You will not be able to make a "sale worthy" open world game in this time period.

I would expect this to take closer to 3-4 years depending on your quality targets. Keep in mind also, that the market will be more demanding in 3-4 years and you will have more indie open world competition, some of which will be quite skilled.

 

Making any half-decent open world RPG is a team effort. These are one of the largest and most demanding games to build. There is no way "cutting corners" will make it viable solo. In a year solo it's possible to make a crappy prototype which will be in "forum-playable-alpha" state. Going from that to something which is worthy of Steam early access will easily add a year... or two.

 

Well, I'd be a fool not to look at the over whelming amount of people trying to talk some sense into me.

What if I brought it way, way, WAY down to the basics.

And by basics, I mean Doom.



Doom is widely considered a masterpiece and yet it only has a handful of enemies and a handful of weapons.

The reason it worked so well is because the enemies had a unique set of attacks that required you to react by prioritizing targets while observing the environment and added the weapons as another matrix in this "dance".

I figure I could make for an interesting mix of enemies with ~10 base models and a couple variants.

Archer enemies would act as "hit scan" enemies.
Mages would work like imps.
And melee would act like pinkies.

So the way I see it:
-Skeletons
-Imps
-Bloater Corpse (boomer)
-Demons
-Overlord Demons
-Some flying enemy

That barely any models at all and each one could easily have three different attack types.

 

I think you're entering more realistic territory here. It's worth noting though, that "dungeon crawler" is very different from "open world adventure".

Given a lack of experience working in RPGs, I would still expect this to take over a year to do good work, but you're in much more realistic space here.

I think even the way you're talking about this idea is much more promising. Here you are talking about gameplay elements, enemy attacks, etc. Previously you were listing different kinds of assets, a sure sign that something is off.

 

I didn't read the entire thread but I think the correct answer is Spiderweb Software.

 

I'm interested in see you try and see where are the bottleneck personally. If you go for it, please document every step