Search Unity

  1. Megacity Metro Demo now available. Download now.
    Dismiss Notice
  2. Unity support for visionOS is now available. Learn more in our blog post.
    Dismiss Notice

How Close Can Indies Get To AAA Games?

Discussion in 'General Discussion' started by Assembler-Maze, May 1, 2017.

  1. Deleted User

    Deleted User

    Guest

    These threads keep popping up and we're all still in the same boat, if you can create a 500 hour game with the consistancy across the board for art / graphics / sound / UI / story / interaction / content like W3 and then have the marketing budget left to make sure your multi-million dollar game doesn't flop around like a fish out of water you're not an indie...

    I don't think many understand exactly what's involved in this entire procedure, also I've noticed the asset store mentioned a couple of times.. I've recently been through the cycle of many asset store items, not shouting out names but I'm yet to find something that's useable in a high end indie game never mind anything to do with AAA..

    Example, a certain asset came out of the box with 6ms CPU overhead and 3ms rendering overhead.. That's more than I'd expect for a voxel cone traced realtime GI solution, never mind this little asset (which is highly rated)..

    None of them seem to be truly integral solutions out of their own defined parameters like you'd find in any AAA engine, if I use Unity I have to write all these things from scratch and that alone will highly impact the amount of time.

    Just try and make a mediocre (indie) openworld RPG with an hours worth of (interesting) content, then come back and tell me how viable it really is. Trying to finish any RPG is a massive undertaking of it's own, I'm not saying for a second indie games in their own way can't be better than X.. I'm saying is it doesn't matter..
     
  2. zombiegorilla

    zombiegorilla

    Moderator

    Joined:
    May 8, 2012
    Posts:
    9,042
    Hearthstone was significantly more than 5mil, probably 10-15ish before marketing/ua. The 10-15 developers is also misleading, that refers to the core developers on the project, that doesn't include outsourced hours/staff and internal shared resources. I know a few artists on hearthstone that weren't part of the "team". Large studio numbers are not something to use as a measuring stick, as they are dependent on studio structure and resources. And you can't judge cost simply by head count. in a multi-studio setup p&l costs are billed from other areas. (qu, it, techops, bs, bi, backend, analytics, hr, etc...) though that staff usually isn't considered part of the "team".

    A AAA title is a heavily invested, large scale, large budget title, usually multiplatform and planned from the beginning for a large marketing / ua spend / dev budget. It's a blockbuster film. It's about production and investment. It says nothing about quality or even success. Usually they are high quality and successful, but that is just a result of the large budget and large/skilled staff, but neither is guaranteed. There isn't a specific number that defines a AAA, but generally it is very high. 10m is not uncommon (dev budget) for a mid range, large studio, mobile title. Indie is a nearly meaningless term, but AAA is pretty clear.
     
    frosted, Kiwasi, theANMATOR2b and 3 others like this.
  3. neoshaman

    neoshaman

    Joined:
    Feb 11, 2011
    Posts:
    6,493
    At least I like that those discussion are shifting more and more toward break down number ...

    Let's say I have a project I have been working for 10 years, if I take an entry pay rates that would 10x12x2000 = 240 000€, if the game sell at 20€ I need 12 000 units to break ROI, if I pay myself highest rates (or goes to a team of 4) that's 48 000. That's nowhere near anything the budget and break even of big AAA.

    When I argue about can indie even make a mediocre AAA, there is also the fact that risk is at a much lower level with potentially bigger return if it goes right. Can you find a market as big as 50 000? that's the real question.
     
  4. zombiegorilla

    zombiegorilla

    Moderator

    Joined:
    May 8, 2012
    Posts:
    9,042
    Or just spend 1 year working on it with 5 people, and only pay each member $40 a year, and then charge $2,500 per copy, and if you sell 1000 copies, and you'll each make $499,960.00! If you are going to use random, unrealistic numbers, you might as well make them fun.
     
  5. Billy4184

    Billy4184

    Joined:
    Jul 7, 2014
    Posts:
    6,010
    1. Why would you do that ... just why?
    2. By the time you released it, it would be CCC or DDD or worse compared to everything else.

    The real discussion is:
    • How can I best spend 1-2 years and $50,000 to achieve a result as close as possible to an AAA game?
    And the answer is:

    Build a self-replicating clone army of Unity game-development bots to do it for you.
     
  6. neoshaman

    neoshaman

    Joined:
    Feb 11, 2011
    Posts:
    6,493
    :D
    My bait was perfect ;) expected actual number to answer though, my number are inflated regarding my own situation though, but I took the samourai way, no social life make your life very very cheap!

    No but I did work 10 years on a project(I'm not alone), I end up quitting my job to accelerate lol but didn't work as plain, because most of the cost is not execution but making all details work together, which is something that is part that people never talk about, whatever the size of the project. I'm not doing AAA though :p
     
  7. zombiegorilla

    zombiegorilla

    Moderator

    Joined:
    May 8, 2012
    Posts:
    9,042
    Obviously its not going to be AAA, that isn't what AAA is. The reality is that working on a game for solo or small team for 10 years is pretty much guaranteed to never actually launch. If you can't commit and complete in that amount of time, you aren't likely the type of person that has the ability complete something like that at all. That doesn't even take into consideration the massive amount of tech changes that happen in that span. Tools, hardware and market is radically changing in a decade. You are going to be spending more and more time just updating your project to work on current hardware/os/tools. Moreover, it is unlikely you are going to make $20 a copy, especially when virtually every marketplace is going to take a cut, and selling 12k copies is completely unrealistic. And on top of all that, a decade of software and hardware costs. Really there is no point in just making up numbers and planning like that. 1) make a game. 2) market research and sales projections based on existing titles/marketplace.

    It's been said many times before, the most important feature is shipping. Without that, sales/budget planning is just daydreaming, not game development.
     
  8. neoshaman

    neoshaman

    Joined:
    Feb 11, 2011
    Posts:
    6,493
    Well I never said I was doing AAA (yet), I mention it because of the reaction to "10 years". And didn't link my project to going to market, it will eventually go, but that's not the goal. At least not for this one I mentioning :p

    I put the bait because I was forgetting something and wanted a refresher (that was the store cut! and I only use free software anyway, the software cost is roll into the pay because indie ... lol). Part of the joke is that people were relying already on team size and budget to make argument (works for industry, questionable for some indie initiative), originally I was having a statement like "wow I didn't knew I had so much money!" but played it straight instead *cough* *cough* knowing it would have flipped some of those arguments, and now people are arguing that making budget and planning is unreasonable, that's quite a swing of the pendulums.

    This is off topic but it raise the question of why are you making game at large (fame, validation, money, hobby, duty? obviously I don't chase money), it intersect the question when we talk about how specific is what you want to achieved. I mention in another thread, about how there is many wastes weighting on AAA production's cost, due to figuring out how to fit disparate unknown details, and how most failed gracefully, explaining why they rely so much on sequel or cloning trend (less specific details to figure out, clear vision, ultimately cheaper).

    In a 5 years of AAA dev, not all 5 years is about "executing" the game, half is about figuring out. Mass effect Andromeda did turn around in only 18 months by cloning the formula of older mass effect, after failing another visions for 3 years, it's not as broken as the internet want you to believe. Planning DO help make decision and know when to cut and when to commit, market or not, planning make execution swift and clear, it allow you to spot reuse and recycling and where to best spend effort. And if you are cloning a trend It helps you, that's where I diverge with my project.

    I have nothing else to compare for my game, not because of the gameplay, but because of the settings (which impact gameplay big time in my case). Also by being NPR by default it put me outside of the AAA tropes (nobody would rate a painting style at AAA, no matter what effort is put into it), even if I had big scope (whatever that mean). In fact I think AAA would be easier (relatively, I mean less problem to solve, more things I can rely on and not invent) and fit right into my training :(

    (off topic) Speaking of my game, it's very specific in way I didn't anticipate. If you want to make a game, any game will do, good job, you can make a game in less than 10 mn. If you add one degree of specificity, for example, make a good game, it can jump from 3 days to 3 month, add polish to it and that's 3 years up to 10 (see the fangame metroid 2 remake).

    I wanted to make a caraibean game (it limit a lot of what I can do, the goal is to avoid the awful european vision on us), it has many implicit I didn't consider, it's a lot of small decision to make, there is no great example to clone, and specifically it's a game with black women, I spend YEARS just looking at hair to "get it", there is no good precedent. At the beginning I wanted to make a kind of zelda light, I trashed that for a an animated VN style with short set piece mini games (clear separation of system, most content follow a known animated workflow, linear and story driven). Right now I invest time into evaluating a digital puppetry (tool) system to reuse for future project and premiere with this one. It just a project that turned out to have many complicated nested project, I need to complete them because what matter is following the vision.

    I don't want to make games, I mean making any random games wouldn't be satisfying, like @ShadowK want to do a specific rendering and don't want to settle for less, I pursue this project because that's what I care about, not market, money, fame, validation, respect, why else quitting my jobs and cutting my expense to minimal?

    It's really not about planning and daydreaming, sometimes you just have to suck it up and do a marathon instead of a sprint, there was a lot of piece to make fit together. I mean I worked on it everyday since then, without a break (almost I did try to prototype other games like a dream pgc space game I keep coming to), after leaving my jobs.

    I mean making an AAA rpg would be much more comfortable, I know what to expect at all level and I know the market, it's expectation and everything ... no matter how many time it takes.

    At the end it's about what do you want to achieve, if you just want to make game, you have a lot of flexibility, that flexibility decrease each time you get more precise and more original.
     
  9. Billy4184

    Billy4184

    Joined:
    Jul 7, 2014
    Posts:
    6,010
    Don't forget to look up from the map every now and then ;)
     
  10. Kiwasi

    Kiwasi

    Joined:
    Dec 5, 2013
    Posts:
    16,860
    Still doesn't cut it. Even if we do for a moment run with the wild assumption that a clone army is the way to dramatically increase developer output.

    If you could build such a clone army on your own, then a hundred people can use the same techniques to build a clone army that is bigger, faster and more efficient.

    Same goes for any techniques you care to name that will magically 'make AAA accessible to indies'. Big games with big budgets can use the same technology. So all you've done is upped the bar for AAA games, without getting any closer.
     
  11. Billy4184

    Billy4184

    Joined:
    Jul 7, 2014
    Posts:
    6,010
    Doesn't work like that. Even if a 'clone army' of proc gen algorithms could make something 90% of the quality of Naughty Dog's art or something, it would not be good enough for a company like that. And with procedural tools it's not even close to being a linear relationship between difficulty and quality, so it's not like they can simply apply more people/dollars to the problem.

    Nor is it always (or even often) possible to simply substitute hand-crafted content for only the parts of a proc gen tool that aren't good enough. Just to throw random figures around, the cost of exposing 20% of the functionality of an automated tool in an expressive way is probably 1000% the cost of creating the automated tool on it's own. Once you take a leap outside of a mathematical formula, all hell breaks loose.

    The fundamental advantage that indies have is the fact that it takes years from the point that a technology becomes available - or obviously useful in some way - and the point that it achieves utility for the AAA industry (very high quality, high flexibility, compatible with hand-crafted art etc). No Mans Sky is an example of this - the tech would not be anywhere near good enough for Destiny but it's quite good for an indie team. If Hello games had created a finite game with proc-gen tools they probably would have been able to create something with an AAA level of content and 70-80% of the quality of AAA art, with only their small team, which would have been a huge success and example of the advantage of procedural generation.

    To expand a touch on this point, the way I see it, there's a huge quality difference between what's achievable in a 'generalizable' proc gen algorithm and one created for a specific purpose. For a tool made for a specific purpose, not only do you not have to deal with all the possible cases (the 99% of them you will never use), but you can specifically avoid complex problems with 'dirty hacks' as long as they work for your situation.

    Which brings me to my next point: so proc gen tools need to be made 'in-house' for a specific purpose, and this means high capital cost (you need to spend a lot of time researching, building and testing the tools), high risk (e.g. you're a year or two in, and you find that customers are hot for corner case X which you never factored into the design) and very reduced flexibility.

    So to sum up, AAA studios want very high quality, a very high level of control and low or no risk, which is the antithesis of what proc gen is all about right now. Indies on the other hand don't have anything to lose.
     
  12. frosted

    frosted

    Joined:
    Jan 17, 2014
    Posts:
    4,044
    All of it requires tweeks, but there is high quality content there if you dig.

    Almost all of my 3d assets come from 3 specific vendors. Two of them worked on Witcher 3, one was core to the character design of Geralt (main character), the other was core environmental design on Witcher.

    In terms of the code from asset store, I use TextMesh, FinalIK, PuppetMaster, Tenkoku, RTP, JSON.NET and a destruction kit.

    I've had to heavily modify each of these to fit my needs (including TextMeshPro), I had to very heavily modify FinalIK/PuppetMaster and Tenkoku to the point where I won't merge updates because I've dramatically deviated from the core.

    That said, that these assets were there as a jumping off point is easily tens if not hundreds of thousands of dollars worth of development cost that was saved. Asset store stuff (and other royalty free sources) are absolutely revolutionary and will dramatically effect the cost of development moving forward. Ignore them at your own cost.
     
    Dustin-Horne, Ryiah and TeagansDad like this.
  13. CarterG81

    CarterG81

    Joined:
    Jul 25, 2013
    Posts:
    1,773
    AAA indeed have so many employers, they hand-craft every area, every level.

    I would love to see some Research on this.

    Have some experienced procgen, inexperienced procgen, experienced level designer, inexperienced level designer, and procgen combined with limited level designers.

    Then have groups rate each level based on some scientific criteria & their opinion. Shuffle the 4-20 levels so each participant receives them in different order, same procgen (but multiple procgen levels), etc.

    Wouldn't be surprised if the only level determined by most users was the inexperienced level designer's. Or if the favorites

    Hypothesis: Players dont actually care or notice.

    Would be worthwhile to see the results, IMO, and dirt cheap for AAA.
     
  14. Deleted User

    Deleted User

    Guest

    I can't think of a way to respond to this without sounding like a complete jerk, so I'll just say we'll agree to disagree on some points.. Although sure there some great assets out there and let's face it in terms of high end game production buying nearly all top rated assets is a drop in the ocean.

    What's more important is my time, if it takes me a week to backwards engineer something like a TOD system I could of made a better system from scratch in the same amount of time.. It's about picking your battles really.
     
  15. Billy4184

    Billy4184

    Joined:
    Jul 7, 2014
    Posts:
    6,010
    At the very least, I think there would be a huge benefit to using them as learning material for making your own stuff. When you have an example that 'works' it makes it much easier to approach making a complex tool yourself. There's not much I'm using off the asset store in my own game (partly due to the fact that there's not a lot of stuff in the genre) but when I see something that is relevant to what I'm doing, and it's not buried under 10 feet of pretty UI, I'll buy it to have a look through it. That goes for art as well.

    It's also something I'm trying to keep in mind for my own assets - I'd rather make something very lightweight and straightforward to understand compared to an all-inclusive solution for everything. While I'm trying to make it as easy as possible for non-coders to just use my stuff for whatever they need, I'm also keen to make it something that I would have been happy to buy as a coder simply to understand the core of the system.

    I'm still not sure if it's the coder in me or some kind of phobia against cluttering, but I'm completely put off by heavy-duty, expansive tools that are loaded with features and menus, something makes me want to keep them well away from my projects. I know some of them are very well maintained and very good quality but I just don't want to use them.
     
  16. zombiegorilla

    zombiegorilla

    Moderator

    Joined:
    May 8, 2012
    Posts:
    9,042
    That's never stopped you before. ;)
     
    Ryiah and Deleted User like this.
  17. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    Then you lose all the money you spent on the project and never get it back.

    10 year long project requires different business model - patreon, for example.
    If you try to make a 10 years old high fidelity project, you're screwed, because by the time you're finished, graphical tech will be upgraded multiple times, meaning that you'll end up chasing the visuals instead of working on your project.

    If you can't finish it in 2 years and have no funding stream, it may be reasonable to mark the project as dead.

    Otherwise, expect to never recoup the expenses.

    I prefer to have a normal discussion and not mind-games.
     
    Kiwasi and zombiegorilla like this.
  18. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    The "But" part is the one that kills projects.
    Heavily relying on 3rd party tech means they may refuse to work together.
    Heavily relying on art assets means they may not fit together.

    If the asset requires too much tweaking (heavily tweaking textmesh/ik/etc), it might be reasonable enough to roll out your own solution. If you aren't planning to turn your own in-house solution into a product for an asset store, you can cut quite a lot of corners.
     
    zombiegorilla and Deleted User like this.
  19. neoshaman

    neoshaman

    Joined:
    Feb 11, 2011
    Posts:
    6,493
    My business model was to start living a super light lifestyle. And 10 years is not full years, I used to work for a large amount of that time :p In my specific case, it's not tech dependent, I did try to do stuff as early as possible but the planning exposed was the most expensive items was the world building, if it change or if it's not well define, it impact everything else, trash and redo, you had to structure and solidify the world building, if your asset don't match that you have done free but useless work. half of those years was me wasting time in aimless prototype because I didn't spend time researching and making a plan, Now I have JUST finished research and having a plan, stability will help me ramp up production. IMHO a patreon (or equivalent) are only good when you have a solid plan and can make definite progression, not stumbling around trying to define your vision.

    And that bring us back to the discussion, What I wanted to highlight with my example, it's that "world building" or "finding the vision/fun" is a hidden cost that is ignored in the discussion, game can be made rather quickly at decent "quality" (production value) if you could remove that burden. Half of AAA dev cost is spend on this aspect too, production ramp up in the second phase when you are sure about what you are trying to do (ie have a plan). It's not just the asset but asset depend on it (it define the asset).
     
  20. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    IMO (I might be overlooking a lot of details, because I haven't read everything about your situation), this sounds a lot like a classic "shattered dreams" project in the making.

    Basically...
    "After ten years of research I've come up with a perfect way to learn how to ride a bike without falling even once!". <--- this kind of feeling.

    Basically... people get an idea for some sort of an opus magnum (especially for their first work), research it, look for a way to making it, throw years of their life into it... and then it falls to pieces because they were lacking something.

    The way I see it, to make a project happen, rather than preparing for that one big thing and doing the research (for ten years?) you must fail. Many times. Repeatedly. Because that way you accumulate practical experience, which will be necessary to maybe make your that one game.

    So, to have your magnum opus, after ten years, you need to have a dozen of several of failed projects. A hundred would be good. Commercial failures, situations where you overestimated the scope, where things didn't work.

    If you don't have those... your project will die. Or will follow road of the Poncho.

    I recall one guy I briefly knew that had an idea about starting some sort business around certain app. Here's the problem: He spent at LEAST 3 years pondering the prototypes, idea, etc... and as far as I can tell never finished it. He failed to launch his idea, failed to make it gain enough traction, and as far as I know, never finished it. There are a lot of people like that. "Hey, I work in sales department, but I had that dream of starting an equivalent of star wars franchise, using that car racing game" (never heard of that succeeding). There are millions of those guys. And that's horrifying, honestly.

    As I understand it, when you get an idea... there's a very short window to start working on it. Few months, pretty much. Past that point, the idea is pretty much dead, and someone else most likely will make it. If you started working on it... there's a brief window to turn it into an actual project, especially if you expect to have some profit for it. This window is ... for a game is around 2 years top. Otherwise you'll a different funding model, similar to patreon.

    When I hear someone say... "Hey, I spent few years assembling a design document, and now I'm finally ready to start my own game".... I think those statements are, honestly, horrifying. . Because it means "I spent few years doing nothing useful to make my project happen". Or there was that one instance where @BoredMormon mentioned a project that spent inordinate amount of time discussing details of lizardman architecture styles rather than making prototypes.
     
    Kiwasi and Ryiah like this.
  21. frosted

    frosted

    Joined:
    Jan 17, 2014
    Posts:
    4,044
    Honestly, other than newbish projects I've never heard or seen an example of a project being "killed" by this stuff.

    I won't overly advocate the asset store, since it's got a bad rep (often, but not always, deserved). But to ignore the wealth of tools and assets out there for dirt cheap is foolish.

    If was going to roll my own IK or SDF font renderer, that would be all the more reason I would buy those assets. Being able to work from an existing, functional (and reasonably complete) product with full source is miles beyond starting from scratch, again, you're looking at tens of thousands of dollars in saved time.

    Ignoring the wealth of material out there (from the asset store, or anywhere else) in 2017 is just foolish, and given the subject here (how much can be done with how little) not talking in realistic terms about 3rd party material is a giant omission.
     
    TeagansDad likes this.
  22. frosted

    frosted

    Joined:
    Jan 17, 2014
    Posts:
    4,044
    This on the other hand, I agree with a thousand times over.
     
  23. neoshaman

    neoshaman

    Joined:
    Feb 11, 2011
    Posts:
    6,493
    This project has nothing to do with magnus opus, there was cuts and they have been DEEP. Also It's born from a series of failure, except I file them as being the same project, early draft was a pirate project, didn't worked, now it's maroon, that's stable.

    I'm not looking for traction, I'm looking for the project being done, but still met the value it originate from.

    And more importantly, my mistake was to not have a gestation phase. Most creative project do, even before pre prod, someone has been thinking about it for some time. It's not unheard to have creator say they based their project from childhood idea they wanted to see into fruition. But generally people don't put gestation phase into the duration of the project.

    Also I don't believe it will be anything special either. Don't wait for it.

    But I can also show the failure along the way, like trying to make a sonic game with mario galaxy physics (hence why I rant about unity physics and performance) bad idea because the math involve was tricky (they had to have a dedicated programmer to solve the most hairy aspect of the physics and they don't even have a free cam like I do), I try to make a racing game without planning, code was fine, but no motivation made the gameplay rather crap and graphics an afterthought (that's not my goal) make me left it with some good lesson about unity, I also made multiple games with people and in a AAA company (and see waste first hand), I tried to make a platformer in blitz3D but jumped to unity free with the sonic clone. I agree with fail fast, but it doesn't mean leaveing project behind. The project is the goal, not making game.

    Now about that magnus opus, I have this pgc dimension/galaxy hopping space game I keep making prototype for since i'm 17, 20 years ago :D THIS will kill me.
     
    GarBenjamin and frosted like this.
  24. CarterG81

    CarterG81

    Joined:
    Jul 25, 2013
    Posts:
    1,773
    This is a pet peeve of mine, and it is indeed very common for people to do. Vague swatches of time, summed up without any context.

    "I've been working on this game for 10 years!" (includes 9.9 years of "I wanna make a game. Hmm...What game?")
    "It took me 3 months just to [blank]!" (includes 2.5 months of Vacation)
    "I worked on [feature] for 2 full months." (Misleading: Feature did not take 320 hours to implement. They did it off/on while doing other stuff.)

    You don't just see this with amateur developers either. You see this with real, legitimate releases that make it to market. I won't name names, but I've heard very successful developers say those EXACT things. When I called bullshit...well, let's just say both devs & fans really dislike you pointing out they're exaggerating because the truth cuts deep.

    So I'm talking Gamasutra articles, game journalists peddling quotes without investigating the claim, bigtime indie celebrities, and minor indie success stories - all participating in this "Let's just make up some arbitrary number & make the claim that is how long it took." Rarely do developers want to talk actual numbers. Not just because they don't think to keep track of such important information, but also because even if they did- they would not like it if they did. (I know some indie gamedevs who are notorious for slow updates, and I can promise you that if they kept track of time working & what they did with their time, then they would hate themselves for being so damn lazy & depressed.)

    So developers don't know anything about how long it actually took (or want to hide it) and so they much up random numbers, exaggerate those numbers, and try their best to make themselves look as hard working as possible. Hence the "This game was the result of over a decade of work!" when the game hadn't started development until a year ago. And no, most of those people don't have complex & detailed GDD's. Those who claim they do? I bet the GDD is vague & full of fluff.

    TLDR: Happens all the time & I want people to be more specific.
    "I've had this idea for a decade, really began fleshing it out 5 years ago, and began actual production 1 year ago."
    is much better than
    "This game is a result of over a decade of work!"
    But the latter is what you'll see time & time again.
     
    Last edited: Jun 12, 2017
    Deleted User likes this.
  25. CarterG81

    CarterG81

    Joined:
    Jul 25, 2013
    Posts:
    1,773
    Is this really what people do? And could you elaborate on "because they were lacking something"?
     
  26. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    Err, no. If I were going to roll my own IK or font renderer, I definitely wouldn't be buying anything from asset store. Definitely.
    The issue here is that asset store license is not good for source code. If you make a derivative, you're forbidden to redistribute it. That makes it significantly less useful.

    It is not a platform to "build on top" of. It is the platform to grab an asset, and plug it in without ever changing. That's how it is supposed to work. If you need to spend few weeks tweaking something on roll out your own solution, then it is not working the way it is supposed to.

    Now, if the project is actually sitting somewhere on github under Mit/bsd/zlib/cc0 licenses, with asset store merely being a mirror, it is good for grabbing. Examples of that are UMA, and I believe DoTween had similar license. Post Effect Stack is another good one. SabreCSG. Everything else should be usable without tweaking, and should not be used for "building on top of".
     
    Deleted User likes this.
  27. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    So no one tried to recreate a single AAA game scene yet, just for fun and giggles?
     
  28. Deleted User

    Deleted User

    Guest

    I sum it up as three years of messing about and now production starts, although without that messing about I would fail out the gate..

    @Arowx

    Bit of a pointless endeavour, just download one of Epic's many high end demo's and reverse engineer.. Give you more time to spend on your own stuff.
     
    Last edited by a moderator: Jun 12, 2017
  29. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    One scene is less than one percent of the total work involved in making the whole thing. What's the point?

    In a single scene you'll be dealing with a level building only. Meaning you'll be missing the fun parts where you tackle character modeling, programming, animation, voice acting, sound effects and music.

    A better idea would be to attempt (if you feel really confident, that is) to recreate a level. WIth characters, combat, movement, and a couple cutscenes. All of them fully voiced, of course, with high fidelity and facial animations.

    For example, Kaer Morhen segment from Witcher 3, or Ciri's segment.
    ciri.jpg
    Good luck
     
    GarBenjamin, Ryiah and neoshaman like this.
  30. frosted

    frosted

    Joined:
    Jan 17, 2014
    Posts:
    4,044
    Either we're talking about apples and oranges or this is a big problem. There are explicit points in licence that state end user may modify assets for use in commercial projects. If I am misreading those, that's a problem where I would need to email various vendors to get rights for legal sale.

    You more or less must have modification rights to royalty free game assets, "modification" in some of these systems is nebulous at best - changing a value on a material could technically be "modification" (as well as deeper lower level issues like format conversion or reprocessing). So without the some degree of modification rights, royalty free game asset sales can't really happen, at least not on the scale of the asset store...

    If you're talking about reselling the asset as a stand alone - that's obviously a different issue entirely and really kind of out of scope in terms of discussion, no?
     
  31. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    Yes, and there are also explicit points prohibiting you from redistributing those assets outside of games.

    Collaboration with anyone goes out of the window, pretty much.

    This one will make working with freelancers very fun:
    Basically, for coding assets you need ease of collaboration. Asset store makes working with remote workers more difficult, and once you're done with your mega project and want to share your code with the world, you can't do it, because you're locked into "are not allowed to redistribute" by eula. This is definitely not the license for software to "build on top of".

    The way I see it, asset store has limited usefulness for modeling assets and art assets, and best avoided if you 're looking for coding solutions you plan to extend. For those scenario you'll need non-viral opensource software instead.

    --------

    Also:

    Are you SURE about that? At least some of the examples you listed are not a difficult problem, and it is possible that amount of time required to read the documentation will be comparable making a knockoff solution.

    There are problems where it is better to buy an asset, but font rendering and IK are not those problems.

    BIG stuff worth purchasing:
    1. Any sort of serious level editor with CSG.
    2. Full replacement of standard lighting pipeline.

    Those are big enough problems, especially when they were tested on large array of hardware. But then again, you won't be tweaking something like that through code.
     
    Last edited: Jun 12, 2017
    Deleted User likes this.
  32. Arowx

    Arowx

    Joined:
    Nov 12, 2009
    Posts:
    8,194
    Well I did get this far, wonder how much more is needed to make the full level?

     
    CarterG81 and Deleted User like this.
  33. frosted

    frosted

    Joined:
    Jan 17, 2014
    Posts:
    4,044
    Yes, if you want to distribute the entire source of your project to a freelancer off location then you will need to buy them licenses to your 3rd party assets...

    Admittedly, I don't know how this works - but if you're going to send the entire source of a mid+ sized project to random freelancers off site then making sure they have $2-300 in licenses is the least of your worries.
     
  34. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    You're missing characters, their voices, combat mechanics, working netcode, music and the rest of the map.

    So, you haven't started working yet.


    Nobody said anything about sending entire source, you know? Nobody also said that freelancers are random.

    The point is, I think the asset store is best avoided when it comes to scripts and code solutions. In most cases. I think I also recall @zombiegorilla mentioning an in-house rule about avoiding asset store, due to licensing issues.
     
    Martin_H likes this.
  35. frosted

    frosted

    Joined:
    Jan 17, 2014
    Posts:
    4,044
    Hahah, come on dude... this is just getting silly. No, you can't do a comparable job to textmeshpro or Patel's animation stuff in the same time it took you to read the documentation...

    Is that stuff rocket science? No. But it's time consuming, detailed work. Doing a good job can easily eat up months of time. Easily. And considering programmer hourly rates, $10,000 is what, 100 hours at best?

    That's like 12 working days.

    Will using these kinds of assets save you huge amounts of time and/or money? Yes. Obviously they will. The idea that these things are not cost effective is bordering on absurd.

    There are valid arguments against third party stuff, but you're not making those arguments.

    ______
    EDIT:

    Look, if you don't want to look at third party code for whatever reason, great. I won't fault you. But if we're talking about using time or money to make games you're wasting it by not at least using this material as reference.

    The idea that you wouldn't want to download something like TextMeshPro if you're thinking about rolling your own SDF font renderer is close to an objectively wrong decision (when dealing with efficiency). It's always a good idea to use other work (especially well respected work) as a reference when planning anything non trivial.

    As for large enterprises, yes, these organizations have complex issues. But frankly, wasting a quarter million dollars is also not huge deal in most of those environments. It's a very different calculus.
     
    Last edited: Jun 12, 2017
  36. Deleted User

    Deleted User

    Guest

    I thought you said you'd heavily modified them? Some times it's easier to roll your own solution if it doesn't fit right, sometimes it ain't and most of this is heavily project dependant.. We're talking about high end / large production games where what you find acceptable they wouldn't..

    I'm doing the last iteration of a large openworld game and some of it just ain't good enough, most likely it never will be. I'm not saying all of them aren't but playing the asset store lotto can be a little frustrating.

    Right now, this second.. With these large "epic" ;) types of games it's utterly perplexing why someone wouldn't choose a toolset / engine dedicated to these types of projects (with a proven track record). If you are able to heavily modify sub-systems / tools like you specified then every single engine I've come across would be nothing but a breeze.

    It doesn't half make me chuckle when someone says, oh I want to make an MMORPG but this engine was too hard to use. Talk about a contradiction..

    I understand Frosted our situations are somewhat different, we're either too deep into something and / or when we started said options simply weren't available (in a decent state at least). It's one of the reasons I've gone back to the drawing board (for the last time).. I can give 500 reasons why but I don't believe at this point it matters..

    I was honestly truly impressed with the latest iterations of Unity as a core engine though, since 4.2 it really has come a long way.
     
  37. neoshaman

    neoshaman

    Joined:
    Feb 11, 2011
    Posts:
    6,493
    I had rewatched a video for another thread, and apparently it's impossible to have streaming asset without stuttering (unless you have code source access) with unity, dixit Inside's dev who had source code access, as of unity 5.6 in their 2016 talk ... rip open world :(
    Is it fixed?
     
    CarterG81 likes this.
  38. frosted

    frosted

    Joined:
    Jan 17, 2014
    Posts:
    4,044
    I did heavily modify them. And including that, there's still a massive time savings and often a massive quality gain. As for what different people find acceptable, you might be right. Or... you might not be right.

    PAMELA used a number of asset store items in a presentation that many (including you and Unity Technologies) found quite impressive. That includes code. So... yeah... one of (if not the most) graphically advanced games released with Unity included asset store code in presentation...

    Hearthstone still includes playmaker dlls in release! That's a friggin Blizzard game!

    Again, if you don't want to look at asset store stuff, great, but many of the biggest and best released Unity games include work from the Asset Store. That's not opinion, it's fact.

    Pretending like it's all worthless or bad is silly and wrong. The reason those things are used is because it saves time and money. Period.
     
    Last edited: Jun 13, 2017
  39. frosted

    frosted

    Joined:
    Jan 17, 2014
    Posts:
    4,044
    I would actually go further and state that of the released and commercially successful Unity games in existence - far more include work purchased from the asset store than don't. Only including commercial success. I'd bet those margins are huge, probably around 90%+ would include at least a few assets.

    Yeah, making the most use of those resources that you can (within reason) is absolutely a smart move. If we're talking seriously about efficient use of money, then yes, you should be looking at those resources. Anything else is just sitting around stroking each others ... ahem... egos. :p

    I shouldn't even have to be making this argument. This should be self evident.
     
  40. neoshaman

    neoshaman

    Joined:
    Feb 11, 2011
    Posts:
    6,493
    Speaking of wich, wasn't Republique using a lot of the asset store, I remember it being branded AAA quality mobile game by unity.
     
  41. Deleted User

    Deleted User

    Guest

    You're missing the point by a country mile, some assets would probably be fine in 95% of games but we're talking fringe cases here where a re-write might be a necessity.

    In terms of graphics I use Scion a lot, it looks great and it's lightweight so no issues there.. Other assets could be used in a smaller game but are completley useless in others.. Some don't save time and they definatley do not save money vs. traditional workflows and self implementation which a lot of it ain't that bad if you know how.

    What's Hearthstone got to do with it? It's a 2D card game, quite a fair amount of difference between that and the Elder scrolls series, also Playmaker is just a preference over anything else.. If they found it essential to their workflow cool, doesn't really mean much. The words "AAA" developer means nothing, it can include anything from a match three to a MMORPG epic.. It's all about the project (as I keep saying).

    Huh, I'm doing well not to be blunt as a hammer here.. :)..! Also whatever works for you.!
     
    Last edited by a moderator: Jun 13, 2017
    zombiegorilla likes this.
  42. frosted

    frosted

    Joined:
    Jan 17, 2014
    Posts:
    4,044
    Necessity can be subjective. I'm also going to be heavily modifying the tone mapping and eye adaption from the post stack to better fit my needs as well.

    Does that mean I should rewrite the entire post stack? No way man. That'd be a waste of time (and money).
    Does that mean that the post stack is crap? No. It's obviously pretty solid.

    Again, artists like M4K worked heavily on Witcher 3, the guys work is good. Same with the guy who worked on Geralt and also sells stuff on the asset store. Those guys are immensely skilled, and their work is very, very good. You don't need to use it, but if you do, you'll probably see a gain in efficiency (and likely a gain in quality if used carefully).

    Replace some shaders, tweek some textures, rework some code to fit your needs. These things are invaluable and a huge value add to working with Unity in the first place.

    To be clear, I am not advocating using everything off the asset store and building a heinous Frankenstein... but to ignore the tremendous value that having those cheap assets provides is folly.
     
  43. frosted

    frosted

    Joined:
    Jan 17, 2014
    Posts:
    4,044
    Material Example
    http://i.imgur.com/MculuWI.gif
    (cant embed larger animated gif)

    Here's an example of my use of IK/Puppetmaster.

    This is more complex than most combat systems, accounting for uneven ground and footing in melee combat with turn based control. I'm using terrain slope, attack direction/velocity, foot placement, facing etc for calculation.This is not a generalized "when he dies go to ragdoll" use case.

    Because it's turn based, there aren't even guaranteed physical collisions. All of it needs to be simulated and manipulated (attack range is stat driven not mesh collider driven).

    Could I have written the entire thing from scratch? Yeah.
    Would it have taken four or five times as long? Probably.
    Would the results be as good? No.
     
    Last edited: Jun 13, 2017
  44. Kiwasi

    Kiwasi

    Joined:
    Dec 5, 2013
    Posts:
    16,860
    Let's not forget that was about the seventh or eighth race they had built lore for. All without a core game loop or prototype.

    These days I don't go near collaboration projects for that reason. Lots of ideas. Lots of plans. Very little actual progress.
     
    CarterG81 likes this.
  45. neoshaman

    neoshaman

    Joined:
    Feb 11, 2011
    Posts:
    6,493
    You can't really escape it if you are doing something with real world culture though, that's why most people do fantasy game, have a problem? invent stuff, in my case have a problem or a great idea? Oh I must find some contrive reason to make it fit or cut it. Sometime the intention is the culture. But spending time on non existing lizardmen is kind of stupid anyway. It's a matter of focus, if you want to make ANY game you will be satisfied with everything. SURELY that's the last time I do this kind of project though ...
     
  46. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    Months? On IK? Not happening. Absolutely.

    Up to 6..12 months, depending on your region.

    Time/cost saving starts at huge features. Most of the coding assets at the asset store appear to be aimed at small problems. Those can be helpful.for people that can't code at all, but if you're at least normal/intermediate level, you can live fine without most of the stuff, without losing much or anything at all.

    The number of really time-saving packages on asset store is really tiny.
     
    Last edited: Jun 13, 2017
    GarBenjamin likes this.
  47. frosted

    frosted

    Joined:
    Jan 17, 2014
    Posts:
    4,044
    Firstly, yes it can. The poorer or more half assed your system for dealing with stuff like integrating physics and animation the more cost incurred. The costs of a half ass system will be spread across the implementation time elsewhere. If you're dealing with a wide variety of problems, the total time involved can absolutely be months.

    Time and quality will vary widely from person to person and domain to domain. You're entirely ignoring ramp up and learning costs, these costs are frankly the majority of all costs in code and development.

    The time required to simply type in code, assuming you have a perfect vision of what the result will be is nearly nothing. Almost all the real cost in development comes from learning, clarifying and disambiguating.

    A domain expert can certainly put together a top quality work in a fairly short time, but legit domain experts are fairly rare and hard to come across.

    I assume you have built IK systems yourself, since you sound fairly certain of the time costs. Do you want to compare your work against FinalIK, quality for quality, output for output?

    I'm willing to believe you may have done a better job on your IK solution, but if so, it certainly took you longer than 5 or 10 days.
     
    SquadraCorse and GarBenjamin like this.
  48. frosted

    frosted

    Joined:
    Jan 17, 2014
    Posts:
    4,044
    Assume that when I am talking about $ I am using standard costs in terms of US region. A competent freelance programmer usually runs at around $100-$200 USD hourly.

    There are 2,080 work hours in a 12 month period (40 hours a week). If you can hire a freelance programmer for 12 months for $10,000 you are paying him less than $5 an hour.

    If you can hire top quality programmers for $5 an hour, then absolutely do it.

    Top end talent for $5/hour rate is the best way to match AAA production value at very low cost. That's the truth.

    ___________________

    I really think that it's possible to match traditionally $5-10 million dollar work on a very low budget, the two legs of this are: royalty free assets, and focusing on extremely low cost talent (globalization). There really are a lot of very skilled individuals who will work for a fraction of the cost.

    I tend to stay away from this kind of discussion on these boards since this subject can get kind of controversial and political.
     
    Last edited: Jun 13, 2017
  49. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    Yes.

    I've worked basic multi-link long chain IK (think about dragon's neck kind of IK limb with 30 + links in it or so) using iterative method and also tried to do the same using Jacobian Matrices as described in one of the ik-related papers floating around. So I have fairly solid idea about process involved.

    Antropomorphic IK would be significantly easier to implement, because due to nature of human limbs, you don't need any advanced operations, and IK may be solved in a few lines of code, using trigonometric methods.

    A normal IK rig setup already provides primary and secondary targets for your IK chains, and reusing those in unity would be trivial. If the secondary targets are not available, they could be recalculated from provided animation.

    Look, from your video, I do not see several months of complexity.

    In unity it is possible to make a small game in two days and that's including art.
    2..3 months timeframe would be sufficient to make a basic fps game engine from scratch without using unity, possibly with your own basic physics engine.

    There just isn't enough depth in the problem to sink so much time on it.

    The exception is ... if the the IK system is not actually an IK system, but an euphoria clone, capable of procedurally driving a muscle-driven character avatar. This is a complex problem, that is big enough to start a company off the idea. But an IK? I'm not seeing enough complexity.

    Now, one preparing an asset store package could sink signficant amount of time into preparing documentation and examples and polishing UI buttons, but when you're on your own codebase, you don't need any of that. You could just stick with the algorithm.
     
    Deleted User likes this.
  50. neginfinity

    neginfinity

    Joined:
    Jan 27, 2013
    Posts:
    13,554
    Well, the difference here is that I agree with globalization (look up programmer salary in regions outside of USA), but I'm not so sure about royalty free assets.

    The reason why I don't think too highly about the assets is because trying to find something I can use has been generally quite problematic.
     
    Deleted User likes this.