How to make an AAA game in Unity (or fail badly)

That's fair enough, and that's not a bad idea for a lot of people, but this thread is for those who are interested in doing something much bigger. Whether or not it is a good idea is not the point of the thread.

 

Yes I am tackling that idea itself. It is too dangerous and risky.

And how much time you think to presumption to input? Maybe 2 more years? Yes even indie can make big thing if he continue over years.

But how can raise money? not for hire other, but feed himself and his family if any.

Release beta version and test in market and get money by sell it? Yes it is good, but when it will be formed at least beta version? and what if not much sales occured from that beta version? Then it failed at that point.

Then what is left? Make game fully completed and release? Yes but only if it is successful enough for compensate develop period. If not? Then I am saying input time of 1~3 (or even more) years is too much valuable to waste like that.

So I think this is only possible to who can control and afford worst result of that input money and years, at the same time who can raise money from other sources.

 

The best way to approach making an indie game imo is to make sure that at every step of the way, you have as little to lose as possible if for one reason or the other you can't (or realize you shouldn't bother) continue to make the game.

I think for starters, it's a great idea to build up an alternative income that doesn't depend on a 9-5 job, such as freelancing, selling on the asset store, selling Unreal marketplace, on turbosquid or whatever you like. It doesn't even have to be game-related, nor does it have to be a full income (assuming you're single of course, or that your partner can put up with you not holding your end of the bargain). That way, if you need a breather of a couple of months you don't need to go job hunting to feed yourself.

Secondly, I think the best way to start is to build a small, short demo that is fully functional in terms of gameplay and demonstrates exactly the graphics quality you want to achieve. This means you know exactly what is going to be required of you to finish the game. And it's at this point that you start marketing the game. Build a website, attract interest, and so on. If you don't get a good response, you're just going to waste your time, because either your game is crap, you're crap at marketing, or both, and any of these three will sink you.

Then, you map out the rest of the game content-wise, build a very efficient workflow for content production, and get on the grind for a year or two until it's done. During this time you continue to peddle the game, and if interest wanes and you can't pick it up, stop and re-evaluate what you're doing.

Another important thing is that after you make the demo, try to attract quality help from other people - very rarely does a collaboration project ever reach demo stage so if you get to this point you'll stand out. The other people will also know exactly what you want in the game and there won't be a confusion of direction or anything like that.

Now if the demo fails to gain interest, well, just pull the plug, it's a waste of time. Build another demo ad try again, or whatever. But the crucial thing is that you have something to show for your hard work, whether to an employer or on your portfolio or your family and friends, something is there that you've done, and people can play it. And that's what you call an achievement.

 

Yes it can be just achievement and portfolio even when it fails. But what is cost of that? Quite big I think.

No money,
Social disregard,
Being laughed at from the very closest people, parents, brothers, spouse, friends, who know you are failed get money from your game. Of course they will not say that at your front(Some rude one will do),
Left of friends, especially friend of opposite gender(lover),
Hellgate opened and hell will be reality in your earth life literally.

But of course there is light side.
What if someone succeed from that game? He will get everything. Specially fame. And enough money to establish his own company.
Even if this is like lottery, too attractive to deny include me.

 

If this is what you get when you fail at something, you need to get to know some different people.

Making a game of such a large scope is a bit like starting a business, it's a very high-risk venture and you have to be prepared (like most entrepreneurs, even those who succeed) to fail, fall over and get back up again. You have to be a survivor. If you know somebody who wants to drag you down when you stumble, either ditch them or ignore them, because if you give yourself the luxury of worrying about what they think you're finished.

Ask yourself, have all these people who would laugh at you accomplished anything worthwhile? Probably not. Do you want to be like them? Probably not. So ignore them.

To be honest, although I've made some pretty radical downward moves in my life (ditching my graduate career out of university for example to teach myself to program) I haven't gotten anything but good responses from family and friends. You just have to make sure you take care of what they think of you - stay humble, try to involve them here and there in your work, don't shut them out and just keep a good positive attitude. People don't often like to be nasty, so when they are it's usually for a reason.

 

Yeah but minecraft, the game with 8 times the size of earth, did good too, and no man sky, even with the hate is not too shaby at being bigger than the observable universe. People (alone) have done prettier and more complex than these two games, but they came late.

Scope isn't a problem for single dev if he plays right in automation, unique assets number is (that's discounting anything that can be automated, city and landscape being two of them), so far quality tailored animations is still a problem no matter what, especially when you need good acting, though game have been going away with sophisticate automated emote.

 

If you have failed then share it here -> http://forum.unity3d.com/threads/share-your-failures-here.433204/

 

Well I don't want this thread to end on a link to a thread for indie failures, so if you're interested in seeing success, here's the WIP thread I just started for my first real stab at making a game - a mobile space combat game to be completed in 3 weeks...

 

@Billy4184 - Want to get some feedback on your new title? FF #48 just opened. You can post here.
Gigi

 

I surely will, thanks for the heads up!

 

Good luck!

 

I think this is interesting to look and analyze relative to the goal of this thread. It's very specific though, it's not about the density of smaller details but the composition of terrain and POI.

 

Talk us through it!

 

This is sadly true. Unless you succeed no one will value your hardwork. If you fail they will try humiliate you.

 

In my case I'd always fail badly as I am not a gamer, can't play games unless I'm forced too...so at best I can make some software can be utilized in games but not quite a real game.

 

People just follow your cue. Whether you think you're a winner, or you think you're a loser, people will make sure you're right.

 

mmm Okay, but I think there is a video on gdc vault that speak exactly what I have seen there, I'll put t below. I'll try to stay concise (hindsight note FAIL):

- disclaimer, I haven't own or play the game yet, I followed a full walkthrough of the main storyline online and the game story don't travel through all regions, so there is a kind of separation of content from the linear path and open world (ie you can miss entire maps and questlines by blindly following story). I'm not commenting on the density of interaction, only on visual quality. We assume the ideal lone indie is having generic assets to help him too (like lost soul abide).

Relative to other open world:
- seen from above, the lay out looks way less organic and more fabricated, in fact flying reveal a lot of dirty visual trickery they used
- but on foot all of that look perfectly normal and natural, so much I question why they offered flight to expose the clunkiness of the layout (I think it was to update they knowhow because they felt they were lagging behind western style game).
- A lot of POI aren't destination POI, they are POV enhancement, ie they tend to be parallel to path and works like tourist tour, you turn the head and see an amazing vista, which is something I never thought about before!
- their open world is a "plain with dot", it's organized along road that form a networks of loop (very few dead end) so it allows continuous movement.
- road allow for stop point "open world" around them. Instead of relying on a aimless space, you have a very wide path structured between A to B landmark and POI (outpost), the width is use for typical open world exploration while not losing the illusion of direction of linear design, you never feel lost and it's more efficient than the POI (village) and surrounding region (cave, clearing where quest leads you) that can expend in any directions. Since stuff are organize around the road, you have point of action that just branch and return back to the road, it's a bit better than a simple hub because it conserve a sense of local progression and direction.
- the terrain visual aren't spectacular in execution, nothing I haven't seen done with unity with people abusing gena + gaia, in fact it's rather average, there isn't any particular density, tree and foliage aren't out of this world, we can see texture repetition, and there isn't particular blending with decoration like rock and the terrain. BUT the average looking are contrasting with clearly artistic interpretation of landmark, and they totally steal the show.
- A lot of use of level design "island" and "H" (except in the later game which people find tedious), ie an island is a blocker (generally movement) you have to move around and is a way that add dynamism an other wise linear path or empty place. Linear path also have clear gate transition when they don't have island to pace the traveling.
- they use terrace a lot to make natural wall visually appealing and not just a big blocker.
- path between to interesting point don't need to be interesting themselves if they are short enough, so there is a lot of relative repetition between stop, it make the next stop pop even more which build anticipation. There is still the banter and tourist sight to break the monotony though.
- tree around point of interest tend to be different than the average background tree, differentiating subtly important place
- they have clear scale contrast between decoration, almost to 8:1, big artistic interpretation of rock with fancy overhang are clearly on a different scale than the scattered rock below it.
- they are not afraid to duplicate unique asset like the big distinctive arch, but know how to use them in a composition that frame a place and keep the tourist sight technique.
- tree group are very often used as framing element to guide sight

Nothing here are out of anyone budget as they more composition principle used to the max, and they are sufficiently structured they be turned into procedural rules.
But this is the talk you should learn to take advantage of the techniques and principles they used:

Now a bit more about the production value and catching to AAA

Spoiler: Opinion: proceed carefully :p

Fundamentally they use the "halo effect", ie a big spectacular or distinctive moment or elements that color, with his quality, all elements surrounding it or followed it. It's visible in the dialogue system, the game is trailing relative to all other AAA games in that department, you have standing mannequin that barely emote and a camera directly on their face where they barely make eyes contact, but it's lead by generally a small cinematic with important character walking and have a cut to their face with big fleshy facial capture.

All main characters have top tier care on them, and these small attention bleed into the surrounding big time, many blandness on the environment are just pushing the character in the front if they are not damning error. The outpost where you are supposed to take the boat would feel much less impressive if there wasn't that scene with Ardyn with top tier facial animation and secondary motion. Let's say an indie can achieve 95% of the quality of the game in the general world, the remaining 5% are in these hi level moment and that only make the level of AAA impossible to reach, because of the halo effect.

Now it's still possible for an indie (WARNING OPINION) to reach that some of that level (the AAA feel, not the TOP TIER AAA level, not all AAA are equivalent) by playing his hand carefully, which is what lost souls abide did. Fleshy animation with expensive emotions are hard and costly, but you don't need them in all circumstance. The understate emotional animation, in Lost soul abide, was very on point, with the mood and feeling of mystery they were going for, much more expensive animation would have break immersion, because it would be unnecessary, out of situation, and be a waste of resource. So small moment of controlled highlight quality can help gave a bigger impression.

Old games use to do it all the time, cutting some effect to allow other to shine, planning around limitations or use them to build anticipation (modern game do it too, having cinematic model for close up and gameplay model for normal). Often wannabe indie just have a maximalist approach, they want the best and densest foliage everywhere for example, with all water simulated, regardless of impacts, and by doing so create homogeneous quality that lower impacts. And as seen, AAA more often than not do choose their battle wisely instead of cranking everything up to 11.

The main difference is that AAA can jump into wasting everything to find the right tuning, mostly by doing everything at the same time, ff15 is great school case, even recent promotional material didn't make it into the final game, entire plot point was dropped, entire character, main characters has more redesign there is character in other game. If want to achieve that level and be an indie, you can't afford waste, you have be sure right at the beginning and roll with what you have, and get it right the first time.

 

I watched it, and really only saw 3 things.
- horrendous vehicle design
- busted up LOD
- broken shadows

Those distracted from really paying attention to much else happening.

 

Well do you think indie can do better?

But since we are looking to learn, let's focus on thing that help us. Let's also make abstraction of "taste", we can always replace this by our own.

Though I make a case than we can do a lot of thing, ff15 stay a league above what's potentially possible, no way any indie can replicate that banter system, also JUST the implementation cost is rather high, I would need to double check with more footage to see if we can automate some of the thing I had see. We can still do the basis and reach AAA entry level, but that will stay hard in practice.

 

I must admit, I spent the first five minutes thinking about all the ways that that flying machine would never have left the ground (at least the front half).

 

The shadow never does. You can see it still on the road, even when the 'car' is over the terrain, it's pretty funny.

 

... and someone needs to tell these people about center of gravity, and center of lift and all that jazz ...

Also, they could probably do with Artomatix's Infinity tiles..

 

The empire fly in big empty box, what do you expect? magic do wonder!

 

Yeah but if you're going to put wings, at least make it believable. I mean, just because you can make a character with feet coming out of their wrists doesn't mean it's a good idea!

Anyway, it would be interesting to consider how a terrain generation tool such as Gaia might implement a foundation of level design into a terrain by creating it according to some of the rules you described - for example using less generic props around points of interest, or building more distinctive, eye-catching areas in such a way that they don't compete with eachother on the screen?

 

In my opinion, most procedural generation technique right now are mostly doing random distribution, with a few rules that design implicitly by magically tuning statistical property that break once in a while. Level design is about organisation of circulation, and few generator have explicit logic to reason about it.

I'm working on establishing procedural rules for that, I mean I need to validate technically the "paper tested" theory, I have implemented some of that, but I need to finish a prototype about enumerating position on quadsphere first (for terrain and lod at planetary scale).

The remaining thing to secure, for the POI system, is finding the formula for pixel coverage on screen, at a relative distance and fov of the camera. This will allow to actually compose, by finding relative scale relative to a specific POV, and have thing like arch or windows framing correctly more or less distant objects.

When I said that I mean I solve it for the infinite terrain case, which will generalize the idea. Here are the prototype result of the first step. It's very crude, the idea is to see if I can have infinite "circulation" on infinite terrain, you are only local to a cell and immediate neighbor. I'll evolve it to generate city like layout (using grid instead of point), use it for line of sight (with interplay with movement), road network, trade routes, etc ...
https://www.dropbox.com/s/m8p75tsu5edk3m5/Capture d'écran 2016-11-04 00.27.05.png?dl=1


Though gaia wouldn't need something as "elaborate", you will still need to define a logic to process some random network of POI, if possible by having theme grouping constraint above.

There is basically 4 differents interplay with POI and path:
1 - I can see and I can go (direct path, generally a pull)
2 - I cannot see but I can go (secret area)
3 - I can see, I cannot go (a simple POV)
4 - I cannot see and cannot go (wasting resource huh?)

The thing is that sight is always a straight line, so you can add complexity by shifting the interplay along the level inside the player's perception. For example, with the same POI, you can set up a 3, then go into 2 then 1, basically a path that curve around a visual blocker and reveal that it lead to the POI (U shapes or O shape path, or "tunnel").

Don't forget also the use of elevation, elevation can be great to structure long range and short range line of sight (like in assasin's creed, big POI are visible even in the crowded street with plenty visual blocker, but high point reveal more and show the actual mapping).

Because line of sight is straight, you have clear parameter to embellish composition by basically spawning stuff relative to that line, framing is just a matter of spawning stuff on the side, like we do with fence on procedural road, you can also explicitly put visual blocker (on the line) to arrange the visibility design among POI networks. Combining with path generation, you can put POI to pull the player into certain direction, until some point where you reveal another POI, leading the player by breadcrumb of POV, funneling it with visual blocker.

 

This is definitely an interesting point. Imagine if you created a terrain by first marking out places on the map where you want the action to converge, for example enemy bases or whatever, and then marking out the paths that you want the player to use to travel between them. Now what if the terrain built itself procedurally around that, for example:

• On an approach to an enemy base, the scene is backdropped by a very distinctive visual marker such as a mountain;
• It uses more interesting props to mark the path to the enemy base, for example using more colorful and distinctive trees, grass, or something unique such as a big boulder or a great fallen tree.
  
• The lay of the land follows the paths, for example a winding, descending canyon that opens into the enemy base.
• Or even the generator creates topology that directs the gameplay, such as an high overhang that you can snipe off, or a crack in the valley wall at the rear of the base where you can make a sneak attack, or some bushes that extend toward the base at a particular point making it easier to stay in cover as you approach.

It's an interesting idea anyway.

 

That's the spirit, here is more about this topic:

Full or partial automation of level design ease the cost of implementation greatly, thus reduce the cost and increase the possibility of what a single person can do. Even though I am a hardcore "proceduralist" there is still benefit for hand craft people by having an automatic prepass, especially if you have broad annotation before to resolve automatically and adjust. There is always that barrier of human experiences that takes R&D to implement in a procedural system that a partially automated system with hand adjustment is still preferable right now. Procedural system can't magically solve arbitrary human convention if you are doing life like human environment. But by having an automatic sight and circulation system you are automating a good chunk of implementation and can potentially double as checking system, but it also make easier to do a lot of stuff like trigger volume and co in a way that understand how player move through a level.

 

@neoshaman

How's it actually working out for your game? Anything to show yet?

 

I had to pull the plug and shift it to side project, I made sure to leave unknown but not any ambiguity, so it's just a matter of solving local problem, it's mostly because it's story dependent and the story is theme dependent, I'm struggling to have a good pace at solving the problem but at least it won't collapse at the slightest change.

So now I'm shifting to another project and make a "no man sky" light and validate all the procediral stuff I have been talking about lol. ON the 64bit I promised a run of 3 month into it after implementing some mesh code I'm doing, but since the bulk of the game is planetary terrain, I have roll that out in my mesh task, yep It's cheating . Also I have start rough prototypes on blitz3D to validate many formula so that's doubly cheating And right now I'm dealing with hashing position on a cube sphere.

After that disastrous run on my former project, I don't want to really do "creative" task anymore, I crave technical validation, so I'm not sure I will bother make the no man sky demake "fun" , I just want to make game. Though R&D on black character continue on it, the main character will be a black woman by default

So yeah I'm upping stuff in direction of procedural generation and be more action less talk I also never made a game entirely alone. So that's trial by fire too.

 

For the type's of games we're talking about procedural generation is not the answer to everything (or much really), it speeds up specific workflows to a certain extent especially creating things like large worlds but for anything else time invested vs. relative gains is a false pretense.

It's like being able to afford a massive mansion, but having no money left to add furniture so you're left with a massive empty house. Even AAA with all their money and prowess can't get it right, I've recently played FF15 and the main story is great and it's a return to form. But the openworld element I can take or leave, it's full of what's essentially fetch quests and dull as dishwater, it has it's odd moments but meh.!

Not only that, you can tell they've had to compromise too far in terms of aesthetics I mean I could write a whole section on how they've skimped on art and tech, there's a severe lack of general interesting gameplay outside the main campaign's etc.. When they could of gone down the path of a more linear FF7 / FF8 approach which was what everyone loved about them in the first place.

I don't know why it's "in" to sacrifice mass quality for size (no pun intended). All I've seen so far of massive worlds and procedural generation is mainly half baked games. Games like TLOU don't suffer from this.

Do what YOU can do to the best of YOUR ability, even if it's aimed right at the top tier of games. I think most people would want 20 hours of interesting rather than a hundred hours of boring.

 

Well FF15 is handcraft, looks like the solution isn't procedural generation or hand craft, but simply better design Procedural system still has be well designed, they represent the designer's bias and intention after all.

Procedural generation is just like game mechanics after all, a bunch of rules design to create an effect, not all mechanics (aka system of rules) are fun, procedural will be the same, it's about finding the right tuning. People find the no man's sky survival update "fun", it's the same game but now the number have been tweaked! So the same system can elicit very different emotion base on just tuning of a few numbers.

Also there is a number of people who hate the main story of ff15, for being jrpg stupid, but like the game and are eager to come back for the open world. It's always hard to separate design from tastes. I'm not sure that's the problem of procedural generation. Also a lot of limitation the game is having is because of the target platform's power, overambition relative to the platform and a lack of direction for 10 years straight.

Anyway, we will see if I can show some potential once I get over the creativity crisis. I don't expect procedural generation to replace creativity in my new game, that would be stupid, but I want to focus on implementation and technique to wind out, so I don't expect the game to be that interesting, because I will be focus on technique rather than fun.

I'm aware of a lot of procedural designer get mired in the tech demo and forget the game and the fun in the excitement of technique, I know it gave procedural generation bad reputation because of it. However those technical validation are supposed to get the base done to address your concern, so wait and see if it works!

If I can get the system to work, then I will tune them to fun and gameplay.

Tl;Dr
When I say I don't want to do creative task, it has nothing to do with procedural generation (it's an unfortunate collision with the discussion in this thread, it could have been anything, like making a platformer), it has to do with me having burn out my will to think about things and wanted to focus on implementing straight stuff into a concrete form. I may have express myself incorrectly.

 

No way they would of made all that terrain by hand, seen as they could just buy a $200.00 piece of software to do it all for them.. It all comes down to a matter of common sense, it's great as a supplemental tool to aid in some areas of rapid iteration. It's not made to be a focal point and / or to create more than a developer can chew..

Also missing the point slightly, having a sense of scale / depth and potentially mucho artwork (with proc gen) is cool but it's nothing more than a 3D scene, it isn't a game in the slightest. I've experimented with way too many types of software and implementations to "ease" the burden of large open world games, when it comes down to it you just have to roll up your sleeves and work through it step by step. Unless we get the almighty make game button, big FPS / RPG's / MMORPG's / MMO's etc. will always be a massive slog and a massive time sink.

In short, there's no "easy" way around it..

Just as a side note, I don't agree it's anything to do with the hardware. Witcher 3 managed to look great even on console and it's a game of mammoth scale in every way, also they are one of the VERY rare few who made the openworld feel alive, an achievement within itself.

Best of luck with your project, keep us updated..

 

It's one of these instance where we can say literally yes it's a real life sculpt
http://www.dualshockers.com/2014/11...-huge-has-many-towns-and-intriguing-features/

They 3d scan it and continue in the game, and that's not the only thing they designed by hand, hair too! It's a workflow similar to what you see in CG animation too.

Witcher play on the strength of the platform and had focus design heavily supported by automation tools. FF dev was a mess and focus on teh wrong part, they have released demo and people had the time to pick apart technically what they did (see NX gamer or Digital Foundry's youtube channel). A lot of problem appear when you fly too, much less when you walk, you can't fly in the witcher I think, which goes back to my theory of the halo effect, aka perception is colored by the worst and high point, hide your weakness and show your highlight and the average baseline will tip accordingly.

On procedural generation, I include mission design too! And I have some lead into mechanics generation I need more work on. Also I'm not fitting a vision to clash with the tools, I design with the tools in mind, just like we design with 3d in mind or 2d, or side scrolling, etc ...

But yeah I'll keep you updated so you can call me out when I fail miserably, no bet, no risk, no reward!

 

Sorry, couldn't resist posting this:

Maybe you should make your own thread, I'd love to see your progress. I have to ask though:

Do you really want to make a game? To me the "creative" and "fun" aspects are a pretty big deal and if you already are starting with a mindset of "might not bother to make it fun", why even start making a game? Most people really care about the fun part, and yet many of them fail at it. If this isn't your priority and you mainly care about the tech, why not make an Asset for the Assetstore instead of a game? Chances are it's more profitable and enjoyable to make for you anyway. And maybe you'll lay the groundwork for someone who does care primarily about the fun?
Maybe I misinterpreted your comment though, it's just a thought.

I have worldmachine and I generally like it, but I also have to say that the more I use it, the more I see its shortcomings. Maybe my workflow is just bad though, I don't know. When I use a procedural tool I prefer to just use the procedural tool and think of it as a stack in an nondestructive workflow. If I sculpted 100% manually I'd want to go for stuff that can't be done with worldmachine (overhangs etc.), if I go with a procedural tool I'd like to never touch it with manual sculpting tools, so that I can always go back and just generate the whole terrain with another seed.
Have you played/seen Squad? On the maps of their levels I'm 99% sure I see the typical erosion patterns of worldmachine, and I also see very obvious manual brushstrokes with a soft round brush. It's not noticable while playing the game and most of those areas are outside of the walkable zone though.

I don't like the layout tools that my version of WM has, not sure if they made progress on those recently. Imho there is plenty of room for improvement. I did try workflows where I just feed it a very crude heightmap painted in photoshop, that serves as a functional level layout, but the end results always looked kind of... primitve and weird.

I haven't tried it yet, but I'm warming up to the idea of generating terrain from premade "stamps" that are blended together. The procedural terrain generation of Rust looks like they do that. That way you can compensate for some of the issues of procedural generation while still maintaining some of the advantages. Maybe...

This is where I'm currently at:





P.S.: second one is closer to ingame perspective, and I've scaled the cloud shadows up a lot for the screenshot. They are about 1/5 the size normally, but the pattern repeating accross the whole landscape looked super artificial at the zoomed out perspective.

 

@Martin_H ... sometimes I feel that we overthink it a bit. I can't tell you how many hours of my life I've burned playing on maps of this quality or similar!

 

Actually I think it is quite likely that more thought went into the map you posted if you directly compare the two. Roads, forests, streets and rails, buildings, rivers. And I bet much of it has tactical effects too. Mine might look more complex on first sight but it really isn't, it's just more random. And that is part of the problem I'm having: how does one make optimal use of such a landscape from a gamdesign perspective? At the moment I just stick giant concrete foundations in it that have sloped edges that are traversable from all sides. Feels somewhat like a copout though. And also like I'm not fully working with the terrain, and a bit against it.

 

What game is that? Looks like Civil War Generals 2 but not as perrtty

 

The texture looks nice, but I think it could easily benefit from height-based texturing (so that the lower ground has a different look) as well as masking in a different heightmap for the lower ground, like something more soft and soily/dusty where it's collected at the bottom.

The way I see it, somewhat unfortunately, world machine and substance designer share the fact that to make something good you generally have to add a lot of nodes and a lot of different influences, and it's an art in that you can't always tell immediately how something you add is going to impact the final result, especially when you have a long procession of 'processing groups'. I haven't spent much time in world machine, but in the case of substance designer, I've seen some really nice textures along with a screenshot of the graph and it's often absolutely chockers.

So for a lot of people (including me I must admit, but it does make sense now) I think these tools are a bit of a disappointment in that they aren't easy to pick up and utilize. They're non-destructive and suit some people's way of doing things - for example I can quite easily make something from a procession of noises and blend nodes, but if you asked me to hand-paint it I wouldn't enjoy it and probably couldn't do a half-decent job - but they take plenty of practice and have a steep learning curve once you get out past the basic stuff. I've been using Designer here and there for a year and it still seems like a mysterious beast, but slowly I learn more and more stuff.

Really most of my progress in Designer is due to Allegorithmic's tutorials, World Machine (I think?) is seriously lacking tuts and it's kind of hard to bridge the gap between what you see on the nodes and what you want to see in the result when there isn't a hands-on example available. Also substance share was a great idea, and World Machine could do with that sort of thing badly.

 

Just something I picked at random from The Googles. Originally I was looking for Battlefront.com Combat Mission maps (fun but the graphics were primitive) though I've spent plenty of time in hex games as well!

 

It's a matter of focus, I need to learn to walk before learning to run, So now the focus is on building and validating the tools. It's not that the game will be not fun, I don't know yet, But I will focus on things I have less a mastery of, ie how to put all these idea together and make them works, not just as paper tested idea and little prototype. I'm currently planning to make a thread about it once I have cross a major threshold, a functioning planetary terrain with basic spawning and generation, it's the prerequisite for the game to even exist in the first place! I mean the game can't be fun if it cannot even works that's what I intended to say. Usually we develop both in tandem, but here I want first to build the muscle, then use the learning from it to make something that is design with the opportunity and constraint in mind, toward fun.

 

If the end goal is to make games, I'm not sure this is the best idea. I've always said that I want to make a few mobile games first - well now I'm finding out that making games fun is another one of those skills that you just have to plug away on for a certain period of time. It doesn't always come as naturally as one might expect.

I agree with your idea of focus, but I definitely do not believe in leaving other things out of your day-to-day for long periods of time. Like any muscle, whatever skills you don't use wither away and then you have to start working on them again. To be honest, I think even if I'd spent some time each week tweaking flappy bird clones, I would be making faster progress on this space game - I've lost that sense of fun factor and now I'm having to flap around trying to get it back.

While I'm a die-hard graphics fan, I'm not fooling myself that what makes a great game isn't mainly the fun factor - and judging by the majority of indie games it's not always an easy thing to come by. So if I were you I wouldn't let it rust.

 

You mean those countless industry professionals, who have been telling us for years to make lots of small games, have been giving good advice all along? Who would have thought ...

 

I send all teh wrong signal right?

The thing is I'm supposed to be trained creative, being creative on demand you need to go a bit further than inspiration, you need to think about the why and the how of what you do, it must be internally consistent but also tthematically with a stated intention, that was my old project. Well that's what I threw up in space for this project!

here is an example:

If I were to make the project creatively, I would have called it Mela9. It would be base on the idea that long inter planetary space travel failed once, because of the lethal background radiation of space (since were not protected by the magnetosphere of our cradle), that slowly killed us. With the boom of biohacking, we successfully transfer the dna of this, https://en.wikipedia.org/wiki/Radiotrophic_fungus , to human. It allow to literrally feed from radiation (called then the glo) instead of food (at least it replace the mitochondria energy production), it's like photosynthesis. Part of thefictional universe would be defined by the cultural fallout of this dscovery ... It's something that would take time to develop seriously, and a lot of mental energy.

Meanwhile the current project is called nomad space, it's a pun on no man'sky, the entire background is that you are a character in a ship going from planet to planet and meet random humanoid alien civilization ... how creative is it? it's basically tropes, obvious permutation and stealing! I won't care about plot hole, coherence, etc ...

The approach is similar about fun, I won't say it won't be there, but I will have greater concern. Let's look at quest design, there is a lot of stuff to make a quest fun, but my main concern here is to know how to actually implement a quest, how do I track a quest technically, how do we manage potentially infinite log of half finished quest? How do I distribute quest on an infinite planetary space? What would an interface look for planet scale planet localization of quest? how do prevent collision of quest line? It's enough to burn myself out before doing anything interesting with the quest, because making something interesting rely on those thing to be solved before hand. I need to balance the effort.

It's a bit like drawing, before doing amazing result and composition, you need to do boring study and learning, this is a learning project. The anatomy sketch you made to study won't be as impressive or impactful that the awesome character you will build out of them, but that character can't be made without nailing down anatomy in the first place, it's my anatomy project!

The last example is this simple napkin sketch, not that it's a great work but it's really illustrative!
The character on the left and the right are the same character, they have been made roughly at the same time. It was a simple character design challenge, I took a bunch of random popular character and merged their visuals signature even though they are essentially incompatible. I start with the left character, I had no idea what I was doing and made things on the go, it's awkward, stiff, janky and you would thought I had barely grasp of drawing. By the end I had a clear idea of the character and how everything fit together and made the right one effortlessly, it took actually less time and have way more details, the idea is that it's a reluctant hero, the savior died and failed his mission, and then entrust his sidekick to hold the mantle of the quest, but she is just the jester she never aspired to greatness, it something you see in the composition even though I didn't actively thought about composing the scene.

My point is that you need to go through a preparatory work to lay foundation to better works, If I had expected for the left one to be great right away I would never gotten anywhere. I rolled with the flaws and once I find out the purpose, it had come out much better in another on.


https://dl.dropboxusercontent.com/u/24530447/divers/n7217074791816690255029.jpg

 

Well I've never disagreed with that advice, and I've always said that I wanted to make a few mobile games first, because it's good practice. In many ways, I don't think the abstract fun factor mechanics of large games differ a whole lot from mobile ones except in the details.

Like neoshaman though, I wanted to build up my 'big game capability' first before anything else, now I think that's probably not the greatest idea. Everything in balance, and all that.

In the end though, chasing the AAA dream is what floats my boat so anything else is still just a build-up to that.

 

The painfully-limited input capabilities of mobile leads me to disagree with this. I have a pretty good phone (Nexus 6P) and I don't even bother with mobile games any more. (Though to be honest, partially it's because I'm tired of wading through endless ad-bloated crapfests. But only partially.)

 

Well by 'abstract fun factor mechanics' I mean stuff like flow, feedback, rewards etc - stuff beyond the scope of any given genre or input system. Games like Halo and Destiny (unfortunately) are becoming very close to mobile games in this respect - the only thing that differs is what specific high-level game mechanics (e.g. fps mechanics, level-design considerations etc) these fun factor mechanics are mounted onto.

But I don't want to try to make it black-and-white, all game mechanics are derivatives of the simplified ones that are found on mobile. For this reason I think practising getting the fun factor on smaller games helps a lot even if you're trying to make some Huge Openworld RPG. In a larger game, they're more subtle, further abstracted into indirect stuff like story and drama, often linked together to form larger chains of reward systems, but they're still there and identifiable.

And it's an instinct you have to develop I think, like art, in the ability to see your work unfolding through the eyes of someone who's not familiar with it. So yeah, I think I should have practised more..

 

Thank you for the explanation, now the whole thing makes sense to me. I never really thought of carrying the "art as craft" analogy that far, but it makes sense.

Sorry, I didn't mean to imply anything. I just had to laugh reading that sentence in the context of the thread title, the "meme" of beginners planning too big and pros telling us to start small, and imagining how industry professionals must feel reading something like that. I just wanted to share my amusement over the irony of this being posted here after over 2 dozen pages of talking about how to make "AAA" games ^^.

If anything, I'm the one sitting in the glasshouse and throwing with way too complex ideas that I never finish x].

 

Well while it might sound like something has changed, it really hasn't - I still have the same point of view. From the beginning of this thread I stressed that I expected it to take a long time to reach where I wanted to get, that I thought the best way to approach game development is by doing a few mobile games, and focusing on short demos for stuff of higher scope until money/collaboration comes in your direction, that I didn't exactly expect to be able to rival anything of current generation but rather something five or six years in the past, etc.

The problem with the thread is that it wasn't focused enough. I just wanted to talk about stuff - chew through some 'out there' ideas with people who might have similar aspirations at pushing boundaries. At the beginning I decided to make a bit of drama by inserting 'AAA' in there, which was just a cynical joke at what I considered to be a pretty meaningless term that I kept hearing bandied arround alongside 'next-gen', usually in an attempt to make something sound better than it looked. Well, there was a bit of mayhem, although when the proverbial fans slowed down a bit we had some interesting discussion I think. But overall not an incredible lot to take away unfortunately.

At the moment I'm just chugging along my usual course. I want to make a few mobile games, hopefully establish myself financially with my games as well as online work. I'm working here and there on a number of ideas for procedural stuff such as a sound effect generator and a tool for generating and building modular assets. I want to make a procedural animation controller and possibly make an attempt at procedural facial animation as well in the near future. High-end rpg games are still my aim. But all of that is on the backburner right now to just getting some games and products out there to establish myself. I'm also working on my site and I hope to take that a long way as well, with a blog to maybe bring together a lot of stuff I've found out trying a bunch of different things across game development. It's all pretty much what I expected, I knew it would take time to build myself up and that's exactly what I've found.

I hope that in some ways at least this thread has recovered from the antipathy toward huge and ambitious projects that I think sometimes does the rounds of the General forum, and that it can just be what I wanted it to be, a place to 'go crazy' with ideas and maybe learn a thing or two from eachother. There's been some fantastic stuff posted here but unfortunately in a lot of cases it hasn't been discussed as much as it could have I think. In any case, I hope everyone's just making progress and getting closer to where they want to get.

 

I have always said that most people don't listen. More specifically they don't listen to unsolicited advice. Tell them a thousand times not to do something and they will do it anyway. Making mistakes is how you learn. I think people need to make those same mistakes before they do things like try smaller projects over larger more complex ones.

My point being there's is no point in trying to convince someone not to do something they obviously want to do. They won't listen. Why try? Let them be. They will figure it out eventually.

 

In a more practical way I think there is 2 main milestone to validate the idea:

1. Making an "horizontal slice", ie, the way I see it, is a full game with simplified assets but all system to validate methodology and workflow of big projects.
2. Making a vertical slice, to master the workflow of high quality asset.

I'm going to do both, and my uncreative project is a step within 1.

The catch is that you are doing a magic trick, it's all about diversion and controlling the attention, you must design WITH your limitation instead of against it.

It's not suited to particular "vision", the more specific your idea, the less lesson learn will translated, the less people will recognize the marker of quality, the less safe the workflow will be. For example there is a bunch of way to have hi quality character through multiple tools (make human, manuel bastoni, fuze, etc ...) but they will never gave you the character you want, it's best to design WITHIN their limits (and highlight their best) rather than retro fit a vision that don't play on their strength.

You must aim for impact aka "recognizable signifier", and use the few Hi quality assets strategically to benefit from their bleeding halo effect.

They gave similar advice in this video, but coming from animation media:

A similar lesson that this slide might be ideals, explanation in the video.