As I teased the other week, I’ve been working on calculations and a report on the entire development process for Die Gute Fabrik’s new game Saltsea Chronicles. That report is now complete, as is the the game itself.
This year I’ve been revisiting the emissions calculations I did for my book Digital Games After Climate Change, trying to put a figure on the emissions across different parts of the games industry – the Saltsea report adds some more context, chipping away at the edges of what we don’t know. Working through the same logic used for my earlier estimates, I looked at the footprint of game developmentbuilding off the detailed disclosures from Ubisoft this year, and extrapolating from them as a relatively representative case study. As the latest net zero snapshot showed however (and as I was reminded by Nicholas Hunsinger), Ubisoft are a fair bit above average in their environmental performance so that shifted things and I landed on some revised figures for upstream game development. I ended up settling on a range of emissions from 2 million up to as much as 60 million tonnes of CO2e.
The bottom end of that estimate (2 million tonnes) was clearly, hopelessly optimistic. Thanks to this year’s just completed net zero snapshot, we now know that total disclosed emissions of the purely games businesses in the snapshot alone were at least 14 million tonnes of CO2e. But the middle of that range – say 30 million tonnes – is looking awfully plausible, especially once we consider how much is still not being captured by the disclosures in the snapshot. As I shared in the snapshot itself, too many companies are still only disclosing a partial account of their Scope 3, and in the coming years more stringent regulations will likely start to draw out more of that picture.
But what’s still missing? At the very least, probably a lot of mobile gaming, though I wonder whether mobile game development has a different profile of carbon intensity again from the rest of those in the snapshot. There may be ways to draw that out.
There’s also going to be game development emissions coming from beyond the biggest 36 companies in the industry – but how much? There seems to be a huge host of ‘Double A’ and indie game making these days – but we still only have the faintest idea what is involved in making indie games. Which brings me back around to the climate impacts report of making Saltsea Chronicles. The report adds a bunch of new details to help us flesh out the picture of indie game-making, and (if I do say so myself) does some new things that I think might be useful to others as well.
In the report we focussed exclusively on ‘upstream’ game development – because the game wasn’t released yet everything is upstream. I broadly followed the principles of the GHG Protocol, however because DGF don’t have any of their own offices, they strictly speaking don’t have any Scope 1 or 2 emissions (fossil fuels you burn yourself, or that others burn to make your power).
With an office, you have the benefit of utility bills that cover the whole company, and that makes calculating and attributing emissions nice and simple – but when everyone works from home you don’t get to simply look at all your employees power bills. Because their fridge, hot water, TV, their pool if they have one, and all their other home appliances are mixed in their with the devices they use for work. We could leave Scope 1 & 2 out of the picture, but that makes benchmarking and comparisons hard, because a) those are the emissions organisations tend to have the most control over, and b) those are also the emissions we have the most data on to compare to other games businesses, because they’re so simple to calculate.
So the solution is to estimate work from home office energy use, but not by using power bills. Luckily DGF kept detailed records of their staff’s working time which could be combined with the UK Govt’s figures for home office energy use – which they supply for both office devices, and for heating. That solves the Scope 1 & 2 figures to a sufficient degree, and just leaves Scope 3 which is another kettle of fish.
For Scope 3 we needed the entire output of their accounting software, and a list (organised by categories) of everything that DGF spent money on over the 3.5 years of development. Those categories could then be interpreted (which involved some back and forth, and I got to learn a few Danish accounting terms in the process) and matched to appropriate ‘categories’ from the various databases of economy-wide input/outputs that have been produced to provide emissions factors for different purchased goods and services. Then with some excel formulas, we convert the spend figures in Danish Krone or British Pounds into the relevant amount that the database takes (2013 US Dollars, for the main one).
That pretty a pretty high-level description of the process. The report itself has more detail, including an entire multi-page appendix where I talk through my choices of database for emissions factors for anyone who is thinking about doing this process themselves. There are pitfalls, but the important thing is not letting uncertainty stop us from trying.
The last step in the process was calculating business travel emissions, which is a bit more straightforward. This was done through distance based estimates for flights and ground travel, using emissions factors for the different modes of travel (again from the UK govt database), and per-night emissions factor figures for hotel stays. No matter how much I know it, abstractly, nothing brings home just how emissions intense flights are than seeing them dwarf absolutely everything else, and every single time it still surprises me. A avoiding even one or two flights makes such a difference.
So after all that, when we add up all those components, for the 3.5 years of development of Saltsea Chronicles:
The total footprint of the game is estimated to be 47,196kg CO2e, or just over 47 tonnes of CO2e, or approximately 0.000058% of what the biggest game companies in the world acknowledged emitting in one year (2022).
That percentage is taking the 80million figure, by the way. But it really doesn’t matter much whether you take it as 80 million, 30 million or 14 million – they are still lightyears apart. One entire indie game has a footprint so microscopic as a fraction of just the known and disclosed emissions from the biggest game companies that we actually couldn’t show the DGF per-employee, per-annum metric on the same graph. It’s a sub-pixel footprint compared to a Nintendo or an Activision or a Microsoft.
The scale issue doesn’t give indie games carte blanche however. Its still 47 tonnes of a greenhouse gas in the atmosphere that wasn’t there before. It's obviously not ideal. But a sense of perspective is key. For whatever reasons, and I think they are multiple, making indie games is just fundamentally better for the planet than making games for the biggest platforms and the largest audiences in the games biz. Is that surprising? Maybe not, but it’s nice to be a little more certain of it.
The recommendations I make in the report are fairly simple – that DGF (and other indie companies for that matter) undertake annual emissions accounting and disclose them. That’s a big ask for a small business, I know, but it’s mainly driven by the urgency of the moment – doing this kind of process every 3 years for each project might be more logistically achievable, but it is a long time to wait to make progress.
In any case, some sort of calculations of footprints and identifying the big sources of emissions is almost certainly the first step for any business. It’s not particularly sexy, but it’s essential because we can’t improve what we don’t know. Having a baseline enables setting targets, and measuring what works to hit those targets and what doesn’t. That's the second recommendation.
The final ones are fairly specific to DGF's unique position, but they're the same overarching goal any games company can adopt. Finding ways to reduce direct emissions can be tricky, especially when you don't have an office to control, and have to people all over the world. I don't envy them the task, to be honest. The last recommendation is the one everyone is trying to figure out – how to pick better suppliers with a lower footprint. There's no simple answers here yet, though I do see things getting easier as more disclosures start happening, and more interest begins to focus on suppliers.
This has been a really fascinating project, and a great chance to test the feasibility of the ecolabelling scheme I proposed back in August. I think it shows that, if you do have the data, you could actually put the figure on the box, or make it visible somewhere on a digital storefront and have a reasonable degree of confidence about it. In fact, I’ve found out just recently that the French online games retailer GamesPlanet is doing something like that with its eco-tags.
What we still need for it though is a lot more data points – more benchmarks from which to extrapolate average intensities for certain types of games or sizes of company, in order to be able to say who has done a better or worse job at keeping emissions down. Climate justice needs to be a part of the consideration in this process as well, as teams in developing countries and emerging markets may not have the same access to resources and opportunities to decarbonise. But the challenges are not insurmountable.
I would also really like the chance to try this process out with a much larger scale of game as well. Know a studio or team (big or small) with a game coming out who’d like to do something similar? Get in touch! Reply to this newsletter or use the contact form on the AfterClimate website.
There’s also the huge issue of end-user emissions, and what developers could be doing to help keep gamer’s own emissions down as well. I wrote a bit about that as well back in August – finding areas of wasted energy that can be designed out of the game still has a huge amount of untapped potential. But we may need to find even more between now and whenever our energy systems become fully decarbonised. Until then, there’s plenty developers could do.