Sunday, 24 August 2014

Stop thinking about Oculus + FPS!

If I had a nickel for every time I've heard someone say they want to play some first-person-shooter or another on an Oculus, I would have a moderately large pile of useless coins. This is frustrating, because the Oculus really shines with different types of games, like racing simulators:

As more people become aware of this fact, the Oculus experience is going get even better.

Wednesday, 20 August 2014

Australia vs United States Income Tax

There's one question that everybody in the world is asking. It is the number one trending topic on Twitter and radio shock jocks will not shut up about it: How does income tax in the United States compare to Australia? Well rejoice all: Here are the answers!

Over the coming bazillion graphs, we will take a look at what elements make up Australian and US income taxes, and find our which country has the more greedy Government. But first, we need to set out some assumptions. Taxation is such a broad topic that without an extremely limited model, it becomes nigh on impossible to make reasonable comparisons. This is why the services of international tax professionals are expensive!

  1. We are going to compare residency in any state of Australia to that of the state of California (CA). CA is the most populous US state, and the one I currently reside in. That makes running this comparison easier!
  2. We are going to assume that our taxpayer is an employee of a company. Not a partner, not a sole-trader, not anything fancy - A regular old employee of a bog-standard tax-avoidance free company.
  3. Our taxpayer has private health insurance. In California, this means ignoring the tax effects of the Affordable Care Act, and in Australia the Medicare Levy Surcharge. Under both systems, those without private health insurance face additional taxes.
  4. Superannuation (Australia) and 401(k)s (United States) shall be banished from this realm. For Australia, this might at first seem odd, given the mandatory nature of superannuation contributions. However, compulsory superannuation contributions are not a tax. Instead, they are a mandatory savings measure under which the payer retains ownership of their contribution, and contributions, not benefits, are defined. Therefore, they should be considered separately from taxes, which have defined benefits. N.B. that the US Social Security taxes under the Federal Insurance Contributions Act are not comparable to Australian superannuation.
  5. Our taxpayer has no deductions. What? Again, a model must be limited in order to be useful. Delving into defining a set of deductions for a hypothetical taxpayer will make this article less generally applicable, and accelerate the greying of my hair.
  6. Australian dollars and United States dollars shall be referred to as AUD and USD respectively.
  7. It is the tax year 2014. This is important! If you are reading this article in a year other than 2014, all calculations, tables, and figures are likely redundant. Governments like to tinker with these things, you see. This assumption also means that we are comparing one Australian tax year (ending 30 June) to the United States tax year (ending 31 December) - The difference in year end has no effect on this model.
Without further ado, lets dive in to possibly the most entertaining article you have ever read!

United States

The United States is supposed to be a low-tax capitalist paradise, looking down on all the dirty socialist Europeans, and Australians. In reality however, Uncle Sam does in fact collect some tax. Income taxes are generally levied by both Federal and State governments, and we will start with the former. The US Federal tax agency is called the Internal Revenue Service (IRS), and levies income tax through seven individual brackets:

Fig1: US Federal Marginal Income Tax Brackets
When applied to to gross income, these brackets produce a fairly flat effective tax rate that levies tax on the first dollar earned and works its way to just above 20% for an individual earning USD100,000.

Fig2: US Federal Income Tax on USD Income
The IRS uses a mechanism called the 'Standard Deduction,' ostensibly to acknowledge deductions the IRS deems too immaterial to warrant itemisation. In reality, the Standard Deduction simply shifts the Federal brackets upwards by USD6,200 for those with few deductions. The result is a lower and more progressive effective tax rate. Note that the Standard Deduction does not 'stack' with other deductions - Itemised deductions must exceed the Standard Deduction before lowering taxable income.

Fig3: US Federal Income Tax on USD Taxable Income (Standard Deduction)
There are two more components to US Federal income tax. The first is a tax on salaries and wages called the Medicare Tax, which pays for a portion of the US Medicare program. This program provides health insurance to those over 65 and younger people with disabilities. It starts at 1.45%, and rises to 2.35% at USD200,000.

In the graph below, note that we're not referring to gross vs taxable income anymore. Medicare Tax is levied on salary paid by an employer (There is also a payroll component, outside the scope of this article!) and therefore taxable vs gross income is not relevant to it. The Federal income tax component continues to assume taxable income using the Standard Deduction.

Fig4: Selected US Federal Income Taxes
The next component is US Social Security tax. This tax pays for a portion of the expenditures of the US Social Security Administration (SSA), a behemoth welfare system that accounts for 37% of US federal government expenditures. The most widely recognised function of the SSA is to provide defined-benefit pensions to retirees.

Like the Medicare tax, Social Security tax is levied directly on salaries and wages. The individual component of the tax (Like Medicare tax, there is also an employer component) is 6.2% of salary and wages up to USD113,700. And so the graph grows:

Fig5: Selected US Federal Income Taxes
The purpose of the is article is to compare a Californian resident and an Australian one. Now that we've completed the federal components of this hypothetical taxpayer's outlays, it is time to add the state ones. The Californian state tax agency is called the Franchise Tax Board (FTB), and its major levy is California Income Tax, which is broken into nine brackets:

Fig6: California Income Tax Brackets
Like US federal income tax, Californian income tax has a Standard Deduction that operates in a similar manner. In 2014, this deduction is set at USD3,906. All graphs and calculations from this point forward assume the application of the Californian Standard Deduction.

An interesting characteristic of US federal tax system is that state income taxes are deductible - For higher income individuals, Californian taxes will eventually overtake the federal Standard Deduction. All graphs and calculations from this point forward account for this, and assume that an individual paying more than USD6,200 in state taxes will choose to deduct those taxes, rather than take the federal Standard Deduction.

Fig7: US Federal & CA Income Taxes (Selected)

But wait, there's more! The CA FTB also levies a tax called California State Disability Insurance (SDI) which is drawn directly from salary and wages, and funds partial wage replacement for disabled workers. SDI is levied at 1% from the first dollar of income through to an upper limit of USD101,636.

Fig8: US Federal & CA Income Taxes
That's it - we now have a full income tax bill for a Californian resident. Here is what the effective tax rate looks like:

Fig9: US Federal & CA Income Taxes, with effective tax rate


Now to drop down into the correct hemisphere, and examine the Australian side of this comparison. Until this point, all figures have been in USD. From here on, they will be in AUD. Keep in mind that this means that the figures above and below this point are not directly comparable. While the AUD has traded close to USD in the past few years, assuming a one-to-one relationship will generate a significant error at even lower income levels. Later on, we will look at comparable figures by using average exchange rates.

Australian states do not levy any income tax. This means we only need to consider federal income tax, administered by the Australian Tax Office (ATO). While the states do levy payroll taxes on high income earners, they are the responsibility of the employer and therefore beyond the scope of our concern. Federal income tax is broken into five brackets:

Fig10: Australian Income Tax Brackets
Which produce the following income tax curve:

Fig11: Australian Income Tax, with effective tax rate

To this, we must add a tax called the Medicare Levy. This tax pays for a portion of the Australian public health care system, which provides comprehensive health care to all Australian residents. The levy starts at AUD20,543, and phases in to a flat rate of 2% of all income at AUD24,167.

Fig12: Australian Income Taxes
One interesting point to note is that unlike US Social Security, US Medicare, and CA SDI taxes, which are levied on salaries and wages, the Medicare Levy tax is levied on taxable income. This will have either positive or negative effects, depending on whether a taxpayer draws the majority of their income from salary and wages, or other sources, and depending on the magnitude of deductions that may be applied to reduce taxable income.

Total effective Australian income tax looks like this:

Fig13: Australian Income Taxes with effective rate
With no state income taxes, and no employee-payable payroll taxes, we can stop here. It is now time to make a comparison between the Californian an Australian amounts!


Because Australia and California use different currencies, we can't just overlay their respective effective tax rate graphs on each other and be done. Instead, it is necessary to convert each one into a single currency. Let's start by presenting both Californian and Australian effective tax rates in USD.

Fig14: Comparative Tax Liability in USD

And for those playing down under, the same comparison in AUD:

Fig15: Comparative Tax Liability in AUD

Both graphs are quite similar, but the slight difference between them is very important. The difference can be large or small, and it depends on exchange rates. Here is a hypothetical AUD comparison, where the price of a single AUD is USD0.50:

Fig15: Hypothetical Comparative Tax Liability in AUD

This is a good moment to recall that this entire exercise is hypothetical. What figure 15 says is that in a hypothetical world, if AUD1 costs USD0.5, someone earning more than AUD70,000 would be better off as a CA resident. Conversely, an employee earning any income up to at least AUD100,000 is better of as an Australian resident at current 12 month average exchange rates.

In reality, very few taxpayers will be in the position of earning an income in a currency other than that of their state of residence. This exercise was never supposed to deal with reality, it is after all an examination of tax! And to close, why don't we depart reality all together, and imagine ourselves earning an income of AUD500,000 per year. What does the tax graph look like then?

Fig16: Comparative Tax Liability in AUD
At present exchange rates, Californian residency becomes preferable at about AUD240,000. So there you have it. If you are in the position to choose your remuneration currency and tax residency as you see fit, be sure to make the optimal choices!

Note: There is an error in graphs 5 to 9 causing US Social Security Tax to be understated by approximately $150 across all income levels. At the resolution of the graphs provided, this error is nearly imperceptible.

Tuesday, 19 August 2014

Gamescom and Fluffy Puppies

Originally written for and posted on the Unknown Worlds blog

Travelling from Cologne to Frankfurt via high-speed rail is one of the most pleasant things a human being can do. A comparable experience would be being plopped on a bed of feathers and covered in golden retriever puppies. To anyone that has grown up in a part of the world without brilliant trains, that is to say most of us, the German ICE is a revelation.

Everything about the journey is delightful. The train pulls gently and quietly into the station at precisely the scheduled moment, panels on the carriages display numbers corresponding to one's ticket, and little steps deploy outwards from the base of the doors to ease ingress.

Inside, doors open with a woosh when approached. The attention to detail is brilliant. For example, the route information display is covered in semi-reflective glass that makes the whole panel look like something out of the future.

At the front of the train, it is even possible to see straight through a floor-to-ceiling glass panel into the driver's cabin, and on to the tracks ahead. The whole setup makes the Eurostar look a bit dour, the Thalys gaudy, and the TGV clunky.

The company responsible for putting together this slick piece of kit is Siemens AG. Since building the original ICE 3 'Velaro' for DBahn in Germany, Siemens has inked deals to export the train to China, Russia, Spain, the Netherlands, Turkey, and Eurostar International Ltd.

Siemens has made itself very good at building high-speed rail train sets, and as a result it is selling lots of them. Which brings me to the process of making games. The purpose of my recent trip on the ICE was to escape Gamescom.

Gamescom is a truly superlative experience. For reference, PAX East, another major gaming event, occupies 19,100 square metres of show floor space in Seattle. Gamescom occupies 284,000m², excluding transit halls and outdoor eating areas. To put that in perspective, the entire PAX East show floor would fit inside just one of Gamescom's halls, the biggest of which occupies 22,332m².

All that space means lots of games. And as one walks through the space, it becomes rapidly apparent that there are lots of people on this planet trying to make train sets - And many of them are very good at it. To sell lots of trains, a game developer needs to think carefully about what kind of train to make, lest they accidentally try to take on Siemens in the high-speed rail space.

Walking the Gamescom floor, there are several themes: There are incredible trailers everywhere, astounding graphics abound, press schmoozing is ubiquitous, and the spectacle is universally extreme. Attempting to differentiate a product on the basis of any of those points is a high-risk strategy, because so many developers are doing it so well.

What Gamescom teaches the observant attendee is that to sell lots of trains, a game developer needs to identify their competitive advantage, and pursue it ruthlessly. A developer may assign resources to many aspects of a game - A successful one will assign those resources where they have the advantage over everyone else.

I don't know what it is about Siemens AG that makes it so good at building high-speed rail train sets, but that doesn't really matter: What matters is what makes Unknown Worlds good at making games. What is our competitive advantage? What can possibly make our product stand out, and not on the Gamescom show floor, but the much bigger floor that the Koelnmesse represents: The games market as a whole.

Little examples of the pursuit of competitive advantage abound. DayZ has terrible animations, Titanfall terrible textures, Call of Duty simple mechanics, Civ5 hair-pulling multiplayer, and so forth. None of them care, as they laugh all the way to the bank. For Bohemia Interactive to re-allocate resources away from inventive, iterative gameplay mechanics in DayZ towards improved animations would  be a huge mistake, and so forth.

Historically, Unknown Worlds has had several competitive advantages. We are capable of creating our own flexible and capable game engines and associated technology, balancing hyper-complex competitive multiplayer, attracting attention through open communication, keeping costs low by maintaining small team sizes, achieving high quality graphics, presenting unique and compelling art styles, and being responsive to the needs of a global, not just English-speaking audience, by operating as a globally distributed and culturally diverse team.

We also have disadvantages. Our games tend to perform poorly on lower end hardware, we are pathologically incapable of sticking to deadlines, developments suffer from massive feature creep, global team communication is often patchy, team gender diversity is atrocious, formal press-facing public relations mechanisms are virtually non-existent, and decisions are too-often made by intuition rather than investigation.

Siemens AG is not perfect either. While it is a highly successful in the high-speed rail, competitors have arguably beaten it in slower and higher speed configurations in many markets. Siemens has advantages, and disadvantages. The same is true for everyone on the Gamescom show floor, and across the whole game development space. Everyone has strengths and weaknesses.

It is crucial that Unknown Worlds, and any game developer, consider where their competitive advantage lies and pursue it ruthlessly. If we don't, if we try to fight Siemens in the high-speed rail space, then we are taking on huge risk. Either we must very clearly consider why we think we can build a better product than the ICE 3 Velaro, or must build games that express our advantages - And in so doing, delight customers in ways that our competitors cannot.

Friday, 6 June 2014

Subnautica Pre-Orders or Not?

It is now June 2014, and Subnautica has been in full-scale development for a few months. While much gameplay is still in the research and development stage, other components of this new game are starting to take on strong form. Subnautica has advanced far enough that Unknown Worlds can begin thinking about offering a product to potential paying customers. (Gasp!)

Subnautica at a very early development stage
There is a loose consensus on the team that we wish to release Subnautica in development form sometime in in the third quarter of 2014, which means a day in July, August, or September. We are fairly confident about doing this, having gained much experienced from Natural Selection 2's "Alpha" and "Beta" releases.

But could we offer Subnautica for sale before this date? Charlie and I had an interesting discussion about this option today. Might Subnautica be made available for pre-order? After all, like every game developer, selling stuff is how Unknown Worlds wheels keep turning. Surely selling more stuff earlier is a win?

Look, don't touch.
Not exactly. Step back to the third quarter target: What do we want to release at that time? If we do it, Subnautica will be in a very early development state. Gameplay will be limited to some basic systems, the environment restricted to a subset of its envisioned glory. But at the core, Subnautica must be fun. If it is not fun, we must not release it.

If a customer purchases Subnautica later this year, on Steam Early Access or another platform, we want them to be thrilled by it, enjoy playing it, want to talk about it, be delighted by it. We then want to update and build upon the game regularly, consistently and inexorably until it is ready for release. What we are selling is therefore twofold: A great gameplay experience, and joyride of updates improving and growing the game all the time.

This is the guiding principle. Make customers happy, and everything else follows. Marketing, public-relations, trailers, reviews, they all become a side-show that could very well help, but are not the core of what it means to be a good game developer.

Spot the Jumper...
What does a customer get if they pre-order Subnautica now? Not much. They do not get a great gameplay experience, they do not get updates, they get a receipt. And a wait. Woo-hoo. That's not to say there would be no value in the wait for some. Occasionally, after releasing development news such as pre-Alpha screenshots, we have received comments like 'I am throwing my money at the screen and nothing is happening!'

Comments like this are humbling, exciting, and morale-boosting for us. By offering Subnautica for pre-order, we might well make a few very excited people happy by giving the opportunity to turn that comment into a purchase. But we would not be doing right by them: We would be selling a promise we have not yet fulfilled. Unknown Worlds would have their money, and they would have a confirmation email. That is not a fair exchange of value, and not an exchange that is likely to induce real happiness in a customer.

Submersible concepts
Sometimes, pre-orders offer more than just an outlet for excitement. For example, some projects cannot proceed without early development stage funding. Kickstarters and Natural Selection 2 are great examples of this. Anyone who pre-ordered Natural Selection 2 was not purchasing a receipt and a wait - They were purchasing the chance for a project they felt passionate about to happen at all. The same goes for countless other worthy crowd and pre-order funded games.

This is not the state Unknown Worlds finds itself in now. Natural Selection 2 was a very successful game. Enough people found challenge and joy in it to fund not just its own development, but Unknown Worlds' future projects, including Subnautica, for a reasonable (but not infinite!) amount of time. Anyone pre-ordering Subnautica would not be purchasing the chance for the project to happen. While we can't be sure we can develop the game fast enough for this to stay true, we think we can.

Instead of offering pre-orders, we are going to double down on making sure that later this year, anyone that purchases Subnautica is absolutely delighted with what they find beneath the waves.

Tuesday, 20 May 2014

Steam Store Page Graphics Guide

Setting up Steam Store page graphics can be a bit daunting at first. There are lots of different images to create and upload throughout Steamworks. Here is a cheat sheet that shows examples of where each of the required images goes.

Steam Store page graphics cheat sheet!
Each of the images pop up in a few other places. For example, the Header Image is also used in Grid view when Steam is in desktop mode. Steam will also often re-size images - The Small Capsule Image in particular can get down as low as 120 x 45, so make sure the image and any associated graphics are readable at that level!

The list of required graphics is:

  • Large Capsule Image (467 x 181)
  • Main Capsule Image (616 x 353)
  • Page Background (1004 x 626)
  • Small Capsule Image (231 x 87)
  • Header Image (460 x 215)
  • Community Icon (32 x 32)
  • Client Icon (32 x 32)

And here are Photoshop size templates for the main ones:

This is all up to date as of May 2014. Given how fast Steam evolves, it might not be up to date by the time you read this.

Thursday, 17 April 2014

Resource Allocation: Deciding What Ideas to Pursue

Deciding which projects to fund, staff, and pursue is simple in theoretical financial terms. Once can calculate expected cash flows from a project, discount them at the cost of capital, and then compare the size of resulting returns. The project with the biggest discounted cash flow number wins - And that is a wrap.

Such calculations are a test that large companies often get wrong, and that small companies, such as independent game developers, cannot reasonably undertake. Rather than attempting to concoct myriad unjustifiable input assumptions in an effort to produce a likely useless number, a degree of abstraction is required and a more qualitative decision making process appropriate.

In forming the structure for this decision making, it is prudent to maintain one constant assumption: You know very little. The effectiveness of the decision maker is limited by their necessarily incomplete cognisance of relevant information. The decision making structure must account for this limitation and not depend on unsustainable attempts to defeat it.

A first principles approach may allow a decision maker to work within those limits. Rather than attempting to define individually what components of what ideas are good or bad, those ideas may stand before guiding principles, and be culled or accepted according to their showing. An example set of first principles that I apply when deciding on how to allocate resources to projects is as follows. A project must,

  1. create value for the company and,
  2. create value for customers and,
  3. create value for creators.
Note that creators and company are separate entities for which value creation is required. A company is a non-human corporate form, the people (whether employed by the company or not) creating products are individuals with goals and desires not necessarily perfectly aligned with those of the company.

Any project that a company undertakes for which I am the decision maker must satisfy these first principles. In achieving such satisfaction, the project ensures that it is viable, that it is capable of execution through intrinsic motivation of all parties. No volunteer need be convinced to work on such a project, no customer need be fed marketing, no employee need be cajoled into contribution.

These first principles represent the initial stage of a decision making structure. The second stage is to consider the relative magnitude of value creation across potential projects. Resources are finite. It is a trap for naive leaders to believe that they, or their teams, can successfully execute all viable projects. It is essential that projects creating the most value must be given resources ahead of those creating less.

The difference between more and less valuable projects is an opportunity cost. To allocate any portion of available resources to a less valuable project is to incur an opportunity cost of magnitude equal to the progress foregone on the more valuable project, less the value created by such allocation. Incurring an opportunity cost is the result of bad decision making.

Such an assertion may appear cold-hearted, but it should not. Any project that incurs an opportunity cost under this method is necessarily preventing the creation of greater value. If an opportunity cost is consciously incurred, then a decision maker has foregone greater value in favour of lesser, to the detriment of their customers, their creators, and their company.

Monday, 17 March 2014

Dangerous Waters: The GDC Effect

Amongst game developers, March - June is often referred to as the 'silly season.' Game Developers Conference (GDC), the Penny Arcade Expo East (PAX), and the Electronic Entertainment Expo (E3) all fall in close succession. To further reduce breathing room, UK shows such as Eurogamer/Rezzed often schedule into the same space.

The PAX East 2013 show floor
Any developer that wishes to attempt a showing at any of these events must view each one in the context of the entire event landscape. Whether or they intend to show at all or just one, the presence of each event affects the other. They are an interconnected, in potentially unexpected ways.

Consider a team that intends to target PAX East to reveal a new game. That reveal will clearly require high-quality content, likely a playable demonstration build. That content would in the most ideal situation require focused, accurately scoped, effectively produced, and sufficiently resourced development time.

Of course, no development situation is ideal, and certainly not the lead up to PAX East. Sitting right before PAX, like a landmine on a war torn highway, is GDC. The value of GDC attendance to a development team is various: Conversing with fellow professionals, attending informative lectures, and gaining motivation and morale by working the party circuit.

Blacking out GDC week in any development cycle is prudent, but possibly not enough - The hangover, literal and figurative, can and likely will last well into the following week. Nutrition, sleep, exercise, often take a dive across the development team during this period. In the context of an entire game development, one week might not make a material difference to the value of the final product. In the case of a PAX East development milestone, that week is pure platinum.

The short term loss of productivity that GDC invokes (I am not in any way passing judgement on the long term productivity effect of GDC attendance: Wide consensus appears to be that it is positive) is not limited to your own team. Every partner that you may be working with to produce a PAX East presence may be experiencing a post GDC slowdown. Press, hardware partners, and contracted third party development resources are attempting to get as much out of GDC as your own team!

Effective planning for the GDC landmine is not sufficient. Consider that post PAX, E3 is less than two months away. If you intend to attend that show as well, then the potential intervening development time is limited. Scope must be reduced to compensate for the risk inherent in a fixed schedule (neither PAX nor E3 will not move for you!), and your team productivity will necessarily taper after a PAX push and need to be ramped up again.

With proper planning, a highly motivated team and a tightly controlled scope, it is possible to navigate the dangerous waters of this particularly crowded part of the gaming calendar. Without it, a team risks missed objectives, the destruction of morale, the breaching of budgets and a failure to achieve the very goal that these events are intended to celebrate: The creation of great games.