The Carbon Emissions of Gaming — Playing for the Planet’s Carbon Complexities — Part 1: Making Games
It’s a sad fact of life that everything we do emits some sort of greenhouse gas. You can’t even breathe without exhaling some carbon dioxide into the world. Though it obviously goes without saying that some actions have a greater impact than others.
As much as I would like to think that video games are a pure form of escapism from the real world and its carbon shackles, I know deep down that it can’t be true. But how bad could it really be? As I sit on my sofa playing a game, am I justified in letting my carbon concerns be briefly forgotten, or should I feel as guilty as I do when I get in my car?
Thankfully, Playing for the Planet have released a report on the carbon emissions of the video games industry to shed some light on the volume and source of greenhouse gas (GHG) emissions across the video game supply chain. The report looks across manufacturers, developers, publishers, and players to understand where the emissions come from, who should be responsible for them, and what everyone can do to minimise their impact.
As is often the way with studying carbon emissions and their sheer level of complexity, the report isn’t able to provide many clear cut answers. So don’t expect any black and white ‘video games are fine/bad’, ‘here’s how to fix it’ statements — the reality is complex. Nonetheless, the report is a step in the right direction, provides some great guidance for the future, and gives us all some actions we can take to minimise our impact.
The Carbon Complexities report from Playing for the Planet (P4P) is absolutely worth reading if you’re that way inclined. I wanted to highlight some points that I found particularly interesting, but summarising it all in a single article couldn’t do it justice; so I’ve split this into two parts: First, the impact of how we make games, and secondly, how we play them.
Emissions Scopes
In the same way that we generate greenhouse gas (GHG) emissions in our daily lives, businesses generate emissions in their day-to-day work. In many other industries, the usual way of reporting these emissions is to use the GHG Protocol Corporate Standard, which separates emissions across three different scopes. The GHG standard defines these three different scopes as:
- Scope 1: direct emissions from owned or controlled sources (e.g. gas or oil owned by the company burned on site to make heat or electricity).
- Scope 2: indirect emissions from the generation of purchased energy (e.g. the energy you use provided by your energy supplier, or emissions from the vehicle you drive).
- Scope 3: all indirect emissions (not included in Scope 2) that occur in the value chain of the reporting company, including both upstream and downstream emissions (e.g. the emissions generated in producing items you buy).
This supply chain can look quite complicated, and what’s ‘upstream’ or ‘downstream’ varies a lot depending on whether a company is a developer, publisher, or manufacturer (or any combination of the three). The figure below from P4P’s report gives some illustration of what the supply chain might look like though.
What this means in practice
Let’s take these emissions in the context of a video game developer:
- Scope 1 emissions would include anything that emits GHGs on the site of a developer’s office, for example. This might be something like fuel (that the company owns) burned on site to generate heat or electricity. However, since most developers use modern offices which acquire their gas and electricity from an offsite energy supplier, Scope 1 emissions often don’t come into consideration for a developer.
- Scope 2 emissions would include emissions required to generate any electricity and gas (provided from a separate energy supplier) that the developer uses to keep the lights on and the computers running, or the emissions from vehicles their employees use.
- Scope 3 emissions would include those generated in the production and transport of the things the company buys, and the things the company produces. Upstream in the supply chain, this might include producing and shipping the graphics cards the company uses. Downstream, this would include the emissions from the use of the products they sell. For a developer specifically, this is hard to quantify, as software isn’t specifically covered by the Protocol (presumably because the energy used to run software is dependent on the hardware running it). However, it doesn’t mean it can’t be estimated using some averages.
For those in the industry of making video games, then, Scope 1 emissions can probably be ignored. Scope 2 emissions will be important, and will depend on the size of the company, the energy efficiency of their equipment, and who provides their energy (whether they provide power through renewable energy, for example). Scope 3 emissions, as the P4P report points out, are the most significant of all scopes for most companies in the industry, with plenty of companies buying a lot of IT equipment, shipping a lot copies of their games (both physical and digital), and with plenty of people downstream in the supply chain playing their games..
Looking across the different types of companies though, the source of emissions within Scope 3 vary quite widely:
- For companies that manufacture consoles (e.g. Sony, Microsoft, Nintendo), a huge amount of emissions come from the equipment they buy, some of which will be used to make consoles.
- For developers, the equipment they buy to make games will be important, but they probably won’t need to account for the emissions used to make the consoles their games will run on (that’s the manufacturer’s responsibility).
- For publishers, the GHG emissions that come from making a game and the console it runs on probably won’t be considered, but a large amount of emissions might come from services they use, like advertising and business travel, due to the nature of the work publishers do.
Of course, many companies both develop and publish their games themselves, and some cover the whole industry end-to-end, manufacturing consoles, developing games for those consoles, and publishing those games. Nintendo is a good example, and looking at their reported carbon footprint gives a good idea of the relative impact of the different types of emissions within Scope 3.
As the P4P report points out, 87% of Nintendo’s emissions came from ‘Purchased Goods and Services’. This itself isn’t that informative, as it covers everything from purchasing IT equipment through to advertising services. However, given 10% of their emissions come from ‘Use of Sold Products’ (i.e. users playing games on their consoles, like the Switch), and another 2% come from ‘Upstream Transportation and Distribution’, this leaves 1% of their emissions to cover all other Scope 3, Scope 2, and Scope 1 emissions.
The equipment used to make games and play them, then, are the most impactful factors in generating GHG emissions, by a long way.
What can be done to reduce emissions?
There will be things that everyone in the video games industry can do to reduce their GHG emissions, and this reporting is the first step in understanding how to do — if we know now that the vast majority of emissions come from upstream emissions associated with creating the equipment used to develop and play them, we can focus on how to reduce emissions from this category. This might include, for example, using second hand equipment, creating hardware that consumes less energy, finding more efficient ways to use materials, or using lower impact materials to make hardware.
If we look at the ‘Use of Sold Products’ category that accounts for 10% of Nintendo’s emissions, there are obviously opportunities to eliminate a lot of emissions, even if it doesn’t affect the larger ‘Purchased Goods and Services’ category (accounting for 87% of Nintendo’s GHG emissions).
As an example the P4P report highlights a brief case study from the development of Halo Infinite by Microsoft-owned developer 343 Industries. While paused, the screen presented to the player is a clear menu but a blurred background of the game itself. In development, this blurred background was still being rendered in 4K, and it didn’t need to be. By reducing the resolution of the background when blurred to 1080p, the team reduced the load on the GPU and so reduced energy usage by 15%; and the player can’t even see the difference (they even tested this to make sure). It might be a small change, but across millions of players and millions of hours of gameplay, it will have amounted to a great deal of GHGs not being emitted, all without any impact to the player experience.
Voluntary Agreements
Console manufacturers themselves have fortunately also taken matters into their own hands to create more efficient consoles. As the P4P report points out, Microsoft, Nintendo, and Sony are all signatories Games Consoles Voluntary Agreement, which commits them to comply with four actions which seek to minimise the power consumption of games consoles:
- Reduce the power consumption of the consoles to the minimum need to meet their operational specifications (but while not limiting these companies’ ability to improve console functionality).
- Ensure energy and resource efficiency requirements set out in the agreement, which includes features like auto-power down and energy caps.
- Ensure consoles aren’t designed to detect they are being tested (lest we end up with another Volkswagen emissions scandal).
- Openly communicate and engage with the European Commission and its member states regarding the energy and material efficiency of games consoles.
According to P4P, the cumulative effect of these changes has saved 54TWh over the course of the previous console generation (including PS4 and Xbox One), which is equivalent to the electricity consumption by Greece in 2021. For the current generation, the estimate sits at 46TWh.
Additionally, in January 2023, Xbox became the first ‘Carbon Aware’ console. This extends the ‘carbon aware’ features of Windows update, whereby updates and downloads are optimised to run at times when the energy grid local to the console is using the most renewable energy.
For others in the industry of making games, P4P point out that guides like PlayCreateGreen.org and the Green Games Guide exist to provide best practice guidance to maximise sustainability and minimise the negative environmental impact of the industry. The guidance covers everything from the way offices are run, how games are marketed, through to how games are played and the end-of-life of hardware.
Relevancy
There’s an interesting point in P4P’s report about ‘relevancy’: the idea that businesses should only include emissions in their carbon accounting that are ‘relevant’ to their actions or choices. It’s a complicated topic that I’m sure is open to interpretation and argument, but the P4P report lists several factors that should determine whether emissions are ‘relevant’ to include:
- Size of the emissions
- Data Availability
- Essential/non-essential
- Influence
These factors influence what a developer, publisher, or manufacturer should account for in their Scope 3 emissions, whether it be upstream, downstream or direct from their activity. For the case of a developer, their Scope 3 emissions would include the emissions resulting from people playing their games, and the equipment they use to do so. Estimating these emissions should take into account this principle of relevancy, only accounting for emissions based on these four factors:
Size — Emissions from some parts of gaming are small enough that they can be ignored compared to others. The P4P reports give an example of emissions resulting from players using handheld controllers. If 5 million players are each using a controller to play a game for 30 hours, the estimated resulting emissions are 75tonnes CO2e (tCO2e), and compared to the same number of players using a monitor display for the same time is 6,900tCO2e. So while the emissions from controllers certainly exist, these emissions are so small compared to other elements like displays or network use that they aren’t worth expending much effort accounting for, especially if it comes at the cost of accurately reporting far more influential factors.
Data Availability — This is a question of how accurate and reliable data might be for an emissions source. Do you have access to estimates of the number of people playing your game? Can you give a good estimate of the power used by the gaming consoles or computers players are using? Headphones are a power-consuming peripheral that players use, but how many players use them versus devices in-build to their displays or laptops? Without accurate data you can’t provide accurate estimates of emissions, meaning you could significantly overestimate or underestimate their impact. P4P argue that unless data availability is good enough to give a reasonably reliable estimate, their emissions shouldn’t be included at all.
Essential/Non-essential — Is an element of a gaming set-up essential for a player to play the game? If not, don’t account for it. A player needs a monitor, but do they need an ultrawide 4K monitor with built-in drinks dispenser? No, those bells and whistles are an optional choice by the player, and the associated extra emissions are therefore their responsibility, not those of the developer. If you’re a developer, only account for the elements which are essential for a player to play the game.
Influence — As a developer, what elements of carbon-emitting components have you had influence over? If your game requires always-online connectivity, for example, you need to account for carbon emissions resulting from a player’s network.This factor plays into the relationship between developers and publishers, illustrated by P4P with the example of in-game graphics options within Monster Hunter Rise. The default graphics options (using a 4K display and a 60hz framerate cap) draws about 145W, prioritising framerate draws a lower 110W (though obviously that power draw will increase if the framerate cap is lifted), and prioritising graphics draws a whopping 180W. If the publisher gets the final say over what the default graphics settings are, and the developer doesn’t get a say, the associated carbon emissions are on the publisher, not the developer. You should only account for the emissions you had influence over.
Each of these four criteria should be accounted for when developers, publishers and manufacturers are deciding what elements of their emissions they should account for. If an element of Scope 3 carbon emissions can’t meet a couple of these criteria, maybe they aren’t relevant enough to be accounted for in an accurate carbon emissions estimate. Be honest about what’s relevant, and don’t report what isn’t.
Consistency
An overriding theme in the P4P report is a call for consistency. The report shows the carbon emissions from several different developers and manufacturers, but consistent they are not. There’s variation in detail, and variation in factors that have been measured (some measuring emissions from user devices, others ignoring it). This makes it incredibly hard if not impossible to compare the emissions from different companies and to understand emissions from the sector as a whole. The whole industry needs a consistent means of reporting on emissions.
This is important for the consumer, as it means they can make informed choices on the products they buy with confidence they understand the environmental impact of the company they’re buying from. But it’s also essential to decarbonising the industry as a whole, as without reliable figures and a good understanding of the largest factors that comprise Scope 3 emissions, it’s hard to know where to focus on reducing emissions and hard to set targets. Consistency really is key to reducing the carbon emissions from the industry.
Defining a consistent means of measurement will be hard, as estimating emissions from things like software is incredibly difficult (given its dependence on the hardware on which it’s running), and given that updates can frequently alter the power usage of hardware, it can feel like we’re trying to hit a moving target. That doesn’t mean it’s impossible though, and standards in other industries would be a good place to start looking. P4P suggests aiming towards meeting the GHG Protocol Value Chain Standard as a good target, with accommodations and edits made to account for the nuance of the video games industry.
P4P asks for the whole industry to come together to define some consistent, agreed reporting standards that can guide the industry in finding high-emissions areas to focus on reducing, and start to decarbonise the industry as a whole.
Conclusions
There are a few conclusions we can draw from the P4P report then, regarding the impact of those who make games:
- The majority of emissions come from those in Scope 3, which includes upstream emissions stemming from purchasing equipment like IT hardware, and downstream emissions that result from people using the consoles and playing the games these companies make, for example.
- Companies should use the principle of ‘relevancy’ when reporting on their GHG emissions, factoring in only those emissions that they are fully responsible for and have sufficient information on based on the following factors: Emissions sources of a significant size (in proportion to their total emissions) to save effort to report larger emissions sources more accurately. Sources that have accurate data available to report on the emissions (inaccurate figures could only lead to inaccurate or misleading conclusions). Sources that are essential to use their products (e.g. the player doesn’t need a 4K monitor to enjoy a game, so only account for the emissions generated by an average monitor). Emissions that were sufficiently influenced by the company (e.g. if a developer didn’t want a physical release of a game but the publisher demanded it, emissions resulting from generating those physical copies should be reported only by the publisher, not the developer).
- There’s work to be done, but it all starts with accurately reporting on emissions so we know where to focus, and having some consistent means by which these can be reported.
- There’s work being done, through voluntary agreements and collectives that provide best practices and a network within which everyone can collaborate to find effective solutions.
- The steps taken to reduce energy consumption and GHG emission don’t have to come at the cost of a good player experience. In fact, often the player won’t even notice.
If you’re in the video game industry and want to find out more detail about any of these points, I really recommend reading through P4P’s report, particularly section 3. It’s clear by what’s been reported and what’s been achieved just on a voluntary basis that there’s a huge amount of will to reduce the negative environmental impact of the video game industry. And given the amount of innovation and creativity that the industry is famous for, I’ve got no doubt we’ll find solutions to even the most difficult challenges.
