Tuesday, July 30, 2013

Fake Graphs About Real Things

I have a tendency to be pretty verbose in my blog posts. My typical topic length has inched ever higher over time, and I think it's starting to impact my comfort level in starting new topics. Each new post, it feels harder to write about something because of the growing number of things I've discussed, and the growing number of words I seem to use each time. And as readers, I'm sure the sight of a wall of text is none too enticing.

This week, I had an idea for something different: fake graphs about real things. Basically, there are some concepts in game development that seem like they are easily described in graphs. They're not always possible to quantify, but the trends are pretty easy to qualify. Like describing volume as a function of distance from a speaker, or speed as a function of time spent accelerating, we can describe certain things in our industry in relation to other values.

So, before I accidentally type another paragraph, let's start "graphing!"

Fig. 1: Amount of time spent actually working on game over length of project.
The above one's pretty easy. Particularly if you're a solo indie, this is what you should plan for. In fact, it's a bit incomplete:
Fig 1a: Everything else you'll be doing.
Oh, and one more thing:
Fig 1b: It ain't over 'til it's over.
Here's another one for those running a game-as-a-service or Minecraft-like beta funding model:
Fig 2: Anxiety as a function of build release schedule.
Those with heart conditions shouldn't wait too long between releases! Particularly if:
Fig 2a: Ohgodohgodohgodohgod!
This one, while sobering, is probably no surprise to most:
Fig 3: Game sales over time.
"That's not so bad," you might think. But we may be missing some points of reference:
Fig 3a: WTB rent-free treehouse with free electricity and internet...
There are a few things you can do to help, though:
Fig 3b: Or this might just be a NASA pulsar graph
Each of these peaks will likely have a diminishing effect on the subsequent peaks, as market saturates and relevance decreases over time.

You could also try adjusting the price of your game:
Fig 4: Sales as a function of price.
$0 is a popular price! Conversely, not many people are into mortgaging their homes to buy a game. Though, as the app store has shown us, there are 8 people who will buy your game no matter what (1 of which, accidentally).

How does this look in terms of actual money? Probably something like this:
Fig 4a: Number may not actually be magic.
This assumes some minimum price, of course. Most places aren't selling games for $0.000001. If that were the case, the vertical slopes would be more gradual.

Let's talk about another topic. Phil Fish has recently made the news (again), and you might be wondering what you'd do in his shoes. Being criticized can be unpleasant, but not all criticism is created equal:
Fig 5: Amount of attention paid to criticism vs. how much I respect the critic.
Put another way, if someone is the type of person who slings insults at complete strangers over the internet, it's hard to take them seriously. After all, with such a low barrier to lashing out, they probably do it to everyone. I might even feel left out if they didn't.

Similarly:
Fig 5a: Attention paid as a function of bias.
Overly negative (and positive!) commentary isn't very useful, so I tend to pay the most attention to balanced critique and critics. Overly biased feedback is about as helpful as a ruler with all 8s.

Speaking of usefulness:
Fig 6: Likelihood of reading feedback as a function of its length.
If I open up an email or forum thread, this is pretty much how things are going to play out. Don't get me wrong, the following is also true:
Fig 6a: Wanting and doing are different things, though.
But if it's going to take me 30 minutes to read, 15 minutes to digest, another 10 to re-skim, and then an hour to reply...well, I might not. Of course, I tend to dwell in the far right hand side of this graph when writing, so there's some irony/hypocrisy to be had here.

And while we're on the topic of weaknesses, here's a big one for me:
Fig 7: Desire to postpone decision as a function of its importance.
Decisions can be hard. Big decisions even more so. While it can be healthy to take time to consider options when faced with a decision, it can be unhealthy to postpone for too long:
Fig 7a: "Just a little more research..."
The important feature of this graph is the peak and subsequent negative slope. For every decision, there is some amount of consideration that is optimal, and beyond that, an ever-increasing opportunity cost. The real-world version of this graph may be bumpy, but the general form is the same.

There are more, of course, but this is already quite a bit. Let me know if you found this format to be helpful, entertaining, or even erroneous. And enjoy the break from my walls o' text while you can!

Tuesday, July 9, 2013

The Hardest Part of Making Games

I was playing Rock Band the other day. We dug the plastic instruments out of storage, and fired up the 360 to dust off our faux musician skills. And at one point, we played Nine Inch Nails's "The Hand That Feeds."

As the little, colored notes dropped off the screen, I found myself imagining how cool it must be to make music like that as a career. I love musical instruments, and to be surrounded by synths, sequencers, 1/4" audio cables, humming amps, electric guitars, and mixers would make my day.

I've dabbled in composition from time-to-time, and I've even had some relatively pleasing results. Sure, they're amateur at best, but creating them was fun. I actually long for the time when I'm simultaneously un-busy enough and motivated to get back into that.

That last thought brought me back from my daydream, even as the colored notes continued. "I'm too busy and too tired," I thought, "because I already have a cool day job like that." It was as sobering as it was uplifting.

In reality, Trent Reznor probably doesn't sit in his studio, admiring his collection. He doesn't sit with a cup of tea, trying out all the synth presets on his Korg M1. He doesn't feel the blissful absence of pressure as he surveys his laboratory.

Mr. Reznor has a business to attend to. He has music to compose, a brand to promote, investments to consider, and relationships to manage. What we picture him doing each day is blissfully ignorant of pressure.

What's So Hard About Making Games?

Like with music, I think there are many on the outside looking in at game developers. Each has their vision of what goes on in a developer's day, and some are more accurate than others. Some may picture developers coding away in the late night hours, illuminated by a cold monitor. Others might picture long conference tables, half-empty pizza boxes, and colored dry-erase diagrams.

Some people picture...other things.
So when it comes to game development in practice vs. imagination, what actually goes on? What are the types of work being done? Are any of them pure enjoyment? Tedium?

And what is the hardest part?

As with music, games are complex products, the creation of which involves a wide variety of tasks. How hard or enjoyable each of these tasks is will depend on the person(s), their abilities, and their preferences. It will also depend on the type of game being made, the tools being used, and the project constraints.

Since NEO Scavenger and Blue Bottle Games are primarily a solo project, I wear a pretty good number of hats. I do most of the dev work, content creation, business administration, and I've even collaborated with a few talented individuals.

What follows, then, is a review of the types of things I've had to do as part of developing NEO Scavenger and running Blue Bottle Games since it started. Along with each, I'll try to call out some of the more enjoyable, challenging, and tedious aspects of each.

If you're considering getting into game development, maybe this will help illustrate some of the rewards and challenges you'll face. And if you're already there, maybe it'll comfort you to know you're not alone :)
  • Game Ideas - It won't surprise you to learn that this is one of the easier, more enjoyable aspects of game development. I love coming up with ideas. I can't not do it, in fact. When I get up from bed to go to the bathroom at 3am, I have new ideas by the time I crawl back into bed. I have journals stacked in a drawer with ideas, a massive folder on my hard drive, another on Google Drive. When I bought my first Palm III, it was so I could jot down ideas as I rode the subway to work. I gave up a lucrative job at BioWare so I could make my ideas into games. I've been known to turn down perfectly good ideas in favor of coming up with ideas of my own. Have I mentioned I like coming up with video game ideas?

    While usually not tedious, there are a few times when it can be frustrating to come up with an idea. Usually, this is because one needs an idea for a specific problem, but none are coming. Ideas pour when they rain, but they're not always what you need when you need them. They're also notorious for derailing otherwise productive work with distraction.
  • Research - This one's a bit like ideas, above. I have to restrain myself from researching things if given the chance. Before I start drawing something, I research it. Writing about something? Research. Start building a new system? Research what other games do.

    Research can be tedious, of course. Plowing through an academic paper is rarely as instantly gratifying as Wikipedia. And many times, there are conflicting sources that need to be resolved. Vocabulary can also cause research effort to mushroom, as one is confronted with a wall of unfamiliar terminology.

    And then there's the question of research's value. Some of it is justifiable, as research can reveal useful strategies or opportunities. Most of the time, though, it's a guilty pleasure that gets in the way of real work. The challenge is often keeping that in check.
Not the hardest part of making games.
  • Pixel Art - Drawing game graphics is moderately hard work, but I enjoy it. It was harder and less enjoyable when I started, as the results were more frustrating. However, with practice, I'm getting better at producing the level of quality that I want, and I enjoy the process more. (The process grows faster, too.) The least enjoyable part of this job is when a piece requires other "infrastructure" art to help it.

    For example, drawing a new rifle is lots of fun. Drawing that same rifle again and again, once for held, worn, and stored versions, and again for the combat UI, then all of the above again for the strapped version, again for the scoped version, and again for the strapped, scoped version? Tedious. I'd rather draw a new rifle. However, for it to work in the game, it needs all of its requisite art done, not just the fun part.
  • VFX - This one can be a lot of fun, but also very frustrating and complex. The cool thing about it is that subtle changes in code and art can make some really spectacular and interesting things happen. It's almost like a mini game development project within a larger project. (E.g. programming the flying, blinking lights in the Detroit map within NEO Scavenger, below).

    The hard part comes when the VFX requires complex algorithms, wrangling multiple coordinate spaces, or bypassing limitations in the platform (e.g. alpha blending that doesn't kill performance).


  • Audio Design, Recording - Audio work is something I am terribly amateur at, but still enjoy immensely. As hinted at in my musings about Trent Reznor, I'd love to sit down with some audio equipment and just noodle for days. And I've even afforded myself a few such days during the course of NEO Scavenger, field recording, cutting and looping effects, and even composing a title screen piece (mercifully replaced by Josh Culler's great soundtrack work).

    It's not fun all the time, though. Sometimes, it means listening carefully to hours of recorded audio for useful bits, cleaning them up, slice and package them, and labeling them with tags for reference later. And in the worst cases, it can be doing this process over and over until it sounds "right."

    Frankly, producing good audio is a full-time job of its own, and I haven't given it the love it deserves. It's beyond my capacity right now, as features and bugs eat the lions share of my development time. I still want to, it's just been back-burnered while fish are frying.
  • Programming - This is where a lot of time is spent. If I had to guess, this probably takes the largest share of my time. For just about everything I create in NEO Scavenger, some coding needs to be done to make it appear. Even coding sometimes requires more coding to work (i.e. bugs).

    As I approach the end of the project, that seems to be changing, though. Most of the early art/writing/design was for new systems that needed to be built. Now, I'm mostly filling in content variety in established categories, so the coding is shifting more towards maintenance and fixes.

    The difficulty and enjoyment in coding varies pretty wildly. Like with VFX above, coding can sometimes mean big changes from little code. This is also often the case when adding a design feature. At such times, coding is my favorite task.

    At other times, coding is a tedium that is hard to endure, or impossible to sort through. Rewriting the GUI for NEO Scavenger is one process that was agonizingly boring and involved. And some of the more complex problems can involve abstract and difficult thinking, especially when the system one is building is dependent on one or more systems that also haven't been built (or defined) yet.
  • Writing - Writing may actually trump artwork, in terms of time spent on tasks. NEO Scavenger is pretty heavy on text, so a lot of work goes into planning and creating that text.

    As with many tasks above, there is fun writing, and there is boring writing. The fun writing comes when my brain is in the right mode, and I have trouble keeping up with it. The story just seems to flow, and I always know what to write next (at least roughly).

    The tedious part of writing comes when one needs to handle all the peripheral infrastructure. In NEO Scavenger, players can take multiple paths in encounters, and that means those paths have to be written. The tedium is helped a bit by the fact that I try to make each path interesting. However, nothing helps that sinking feeling when I realize, "ugh, they might do that, and that means a whole new string or tree of encounters I need to conjure up."

    The hard part of writing is rather like what I described in programming, above. When so little has been defined, it might seem like one has lots of freedom. However, that freedom can be paralyzing, as one is constantly second-guessing whether the idea or character is plausible, interesting, or whether it might sabotage the story somehow down the line.
"...and the player will be able to choose ANY action they want!"

  • Game Design - Design is a bit hard to separate from the above, since it involves almost all of them. On a good day, it means recognizing clever ways of making two systems interact, producing a complexity greater than the sum of its parts. At times like these, one can almost enjoy their own game as if they're a new player, watching systems interact in surprising ways.

    On other days, there is a design problem that appears to have no good solution. I've spent days on some problems: thinking; starting down a path; recognizing a new issue; erasing; repeating. It can be very disheartening and frustrating. Sometimes, the problems with a new system don't emerge until all the work is done. Then, it becomes a trade-off between spending extensive time fixing, or extensive time rebuilding.

    And tedium can rear its head here, too. Some systems require reams of data entry and proofreading. Other times, navigating a web of interdependencies to ensure the systems don't undermine each other.
  • Build and Platform Management - This refers to the process of getting the game to run on the platform of choice, whether it be a browser, Windows, Mac, or Linux. It also refers to the process of getting that game into the hands of the person who needs it, be they oneself, a collaborator, or a customer.

    I'll admit, there's little here that I enjoy. I'm thankful for some of the experiences I've had in making NEO Scavenger cross-platform, but I wouldn't call the work enjoyable. Mostly, it's a mixture of tedious and challenging.

    The challenge is usually wrestling with the idiosyncrasies of each platform, be they resolution, user input devices, UI conventions, performance, or unexpected bugs. Flash helps with much of this, but as it turns out, causes almost as many problems.

    The tedium is usually the actual build process for me. Creating each of the platform builds requires a lot of PC-hopping, file uploading, changelog-writing, and form-filling.
They LOOK friendly enough...
  • PR - Interacting with press and public is a job in itself, and can take as much time as you offer. There will always be more conversations out there about your game than you can handle, so you have to set some limits to leave time for game development.

    Without a doubt, reading press and player feedback is my favorite part of PR. Praise can be highly motivating, as can constructive criticism. Even knowing that people are simply playing the game can raise one's confidence.

    I also enjoy learning about the art of promotion, and creating plans for publicity (e.g. special events, giveaways, strategic link placement, game festivals, etc.).

    I'm terrible at executing those plans, though. I hate asking people for things, so cold-calling press outlets is hard for me to do. I also have a hard time pulling the trigger on a PR event, knowing that "just one more" fix or feature would improve the first impression, perpetually delaying my plans.

    When it comes to customers, some interactions are easier than others. I love helping people, so if my response can offer something useful to the player, I feel good about it. Conversely, if I don't have anything helpful to offer the player, responding stresses me out.

    There are critics out there, but they're actually not as stressful as you might imagine. In cases where the criticism is valid and constructive, it's actually welcomed and helpful. And in cases where it's more insult than critique, it's easy to disregard.
  • Team Management - NEO Scavenger may be a solo endeavor, but I have had help along the way. And most of my experience prior to Blue Bottle Games was with large teams.

    At its best, teamwork is awesome. Being surrounded by talented, creative, and motivated people can make you do amazing things, and create stellar products. Some of my happiest moments (in any career) were when surrounded by a tight team of bright minds with a shared vision.

    Working with Josh on NEO Scavenger's music, and Cameron on NEO Scavenger's story and design, was equally inspiring. They could see things I missed, and offer ideas I never would have, helping to produce a fuller, more professional game experience.

    If there's any tedium I experienced, it was due to one major mistake I made: making myself the bottleneck. There's nothing more mitigating to a high-performance machine than forcing it to drive in rush-hour traffic, and that's what I did with Josh and Cameron. I never built the infrastructure for collaborators to contribute directly, and I wanted to own as much of the process as possible. As a result, I slowed down their progress. One could argue that I failed at one of the greater challenges of team management: delegation.

    Of course, there can also be bad teams. I've been on my share of those as well, and they can cause no end of tedium and frustration (for both the members and managers). Diverging goals and opinions, authority issues, poor communication, and lack of clarity can ruin even the best collections of talent. Bad teams are jobs in themselves, and making games becomes a side project compared to keeping things from melt-down.
  • Business Analysis - As with PR, this is an area I enjoy exploring quite a bit. And that came as a surprise to me.

    Learning about economics, business strategies, and entrepreneurship can be really addictive. I enjoy looking at the game I want to make, the market it's trying to reach, and hashing out features, pricing, sales channels, and partnerships to make it viable.

    Running the numbers can be tedious, of course. As can digesting info to keep one's picture of the market up-to-date. However, there's a certain Skinner-like reward to reading news about the industry, and suddenly recognizing an opportunity your product is poised to take advantage of. It's one of those rare moments in life when your "money-making idea" can actually do what it says on the tin.

    The hard part, here, is often just finding useful information. Nobody is in the business of writing "A Market Analysis Tailored for Dan Fedor." There's a lot of filling in blanks with educated guesses, extrapolating data, and in some cases, blind faith. Will your next game idea be a hit? Nobody knows. Even the AAA analysts are clueless here.
Let's not forget DIY desktop support.
  • Testing and Debugging - When folks say things like, "oh, you just play games all day?" this is probably of what they're thinking. And they're wrong.

    Just about everyone on a game dev team tests and debugs their work. Or at least, anyone worth their salt does. Fire up the game, navigate to the area where your recent work shows up, and start testing it.

    This can actually be a pretty fascinating area of work. It rewards good experimental method, note-taking, statistical analysis, and deduction. Sometimes, it can feel like forensics or detective work. And yes, once in a while, you get a chance to play through a section to see how it "feels."

    Unfortunately, this type of work can also require a lot of patience and persistence. It's not good enough that one tests entering combat and exiting again. One has to test cases where the player initiates combat with AI, AI with player, player with AI already in combat with AI, AI with player vs. AI. What if the player dies instead of leaving? The AI? What if one AI leaves and another stays? What if the player leaves, and more than one AI stays?

    There can be a combinatorial multitude of scenarios to test, and doing this job right means testing them all. Worse yet, once a change is made to the system, this entire suite of tests needs to be run again. And changes to the system are always happening.

    Some testing can be quite frustrating, too. A notorious example in NEO Scavenger are the null-pointer bugs that only show up later in a game session. They're frustrating because the cause is unclear, and one must play the game for long periods before the bug might appear. And often, the symptom of the bug varies from game-to-game, making it hard to tell if two bugs are related. If the bug cannot be caused reliably, and requires long periods of play to trigger, imagine trying to fix it. And worse, verifying that the fix works.
  • Project Management - In an ideal world, each developer could just release the game they want, when they want to. In reality, the pressures of time, budget, and manpower impinge on that ideal. Project management is the process of quantifying those needs and constraints, and finding the compromise that balances those forces best.

    A lot of folks picture project management as Gantt chartsscrum, status meetings, and spreadsheets. And a lot of project management involves those things.

    However, all of us do some amount of project management subconsciously when we assess the value of a new feature against our capabilities, schedule, and budget. Whether it's a customer asking for widescreen format support, or a teammate asking for alpha channels in their sprites, we're assessing cost vs. value somehow in our decision-making process.

    There is little that I enjoy about this process, honestly. Gathering data can be a lot of work, it often involves telling people "no," and the results are based on a lot of guesswork.

    That said, the process can reveal useful questions and concerns that are missed in more casual assessments. It can also serve as a common reference point for discussions among others. Even as a solo developer, it can be very helpful to have a to-do list that I can sort by effort, value, finished/unfinished, or other criteria.

    So, for me, it's one of those things I'm glad to have done, but dislike doing.
  • Taxes and Accounting - If you're running your own business, the CRA/IRS wants to know what you earn. And likely, you'll want to know as well.

    There is nothing enjoyable nor rewarding about doing taxes. It is 100% tedium and challenge, so I'll leave it at that.

    Accounting can be grueling, but the rewards are there if done right. It's important to know how one's business is performing, in order to make sound decisions. And it can sometimes be an ego boost if sales are good. But it's a lot of careful, detailed work, and it can involve some challenges of data-management and research.
More fun than taxes.
Some Additional Hard Parts About Making AAA Games

There are quite a few other areas that I could discuss, particularly from the point of view of a larger team. NEO Scavenger is a pretty small undertaking compared to AAA efforts, and I don't (yet) have to worry about such things as:
  • Animation - Either traditional or 3D rigging, keyframing, motion capture, and the like.
  • 3D Modeling - 3D models include characters, environment art, props, etc.
  • Materials - The development of textures, shaders, and other surface material properties for models.
  • Lighting - 3D worlds, and even some 2D worlds, require sources of light to be placed strategically.
  • Cinematics - Positioning cameras, moving them around, as well as managing the content being shown, settings for playback, and related work.
  • Localization - Translating the game for multiple markets.
  • Physics - Assigning volumes and physical properties to items in the game world.
  • Licensing - Negotiating permission to use licensed properties.
  • Other - And everything else I've forgotten about.
What is the Hardest Part About Making Games?

So what is the hardest part about making games? You may have noticed a pattern in the list above, having to do with the aspects which I find tedious or difficult. "God is in the detail," as they say, and this is as true in game development as any other field.

Dabbling in an art or science is fun, but due diligence is work. The hardest part of any task above is when I have to get my hands dirty and follow through with all the details. There's a reason organizations usually have a limited number of leaders with the broad-strokes vision, and an army of workers to do the detail work.

However, even the tasks listed above are manageable on their own. Many are measured in hours, days, or weeks, and not much worse than school assignments in that respect.

The hardest, in my opinion, is finishing a game. Spending hours each day, every day, consecutively for over two years. Waking up on day 520 of NEO Scavenger, and finding the motivation to do one more. That's the hard part. I like making games, but even I get tired of making the same game after all that time.

Don't mind if I do.
The hardest part of making a game is finishing it. For all the things I can do on my project, that's the one I still haven't.