This seems like a very cool project, thanks for sharing! I agree that this type of project can be considered a "moonshot", which implies that most of the potential impact lies in the tail end of possible outcomes. Consequently the estimated become very tricky. If the EV is dominated by a few outlier scenarios, reality will most likely turn out to be underwhelming.
I'm not sure if one can really make a good case that working on such a game is worthwhile from an impact perspective. But looking at the state of things and the community as a whole, it does still seem preferable to me that somebody somewhere puts some significant effort into EA games (sorry for the pun).
Also, to add one possible path to impact this might enable: it might be yet another channel to deliberately nudge people into in order to expose them to the key EA ideas in an entertaining way (HPMOR being another such example). So your players might not all end up being "random people", but a ratio of them might be preselected in a way.
Lastly, it seems like at least 5-10 people (and probably considerably more) in EA are interested or involved in game development. I'm not familiar of any way in which this group is currently connected - would probably be worth doing so. Maybe something low on overhead such as a Signal group would work as a start?
Considering that I need to cast a relatively wide net to find even thousands of players, the messaging about Mineralis is not very selective on purpose. I think that the overall demographic will end up being relatively random within the set of people who are gamers and enjoy JRPGs. Whether JRPG fans have a predisposition towards EA - that remains to be seen. :)
I would be super humbled if the game ends up being a success and can be considered useful for the purpose of actually teaching EA concepts. However, I would imagine that a vast majority learns better through reading articles or listening to podcasts.
Setting up a group for discussion sounds like a valuable idea. I added a personal task for that and expect to get to it late next week. I'm thinking Discord might make the most sense, given that many developers already use it anyway.
Cost-effectiveness of making a video game with EA concepts
Note: highly speculative case study, with a lot of uncertainty over the numbers.
In this post, I try to estimate the cost-effectiveness of making Mineralis, an upcoming role-playing video game (JRPG) about travel, altruism, and personal growth.
Although the gameplay is typical, many aspects and ideas are inspired by EA concepts. This post is before the game's release, and I will make some assumptions/predictions. I'm planning a follow up post that examines how well they hold up.
The game tries to push the boundaries of the JRPG genre in an experimental way, by having a strong focus on meaningfulness and cutting out all forms of filler content. It could be thought of as a "moonshot project".
Summary:
Estimated conversion rate for EA participation or long-term altruistic lifestyle changes per player: 0.1% (90% CI 0-11%). In other words, one person per ~1500 sales.
Total development costs: ~1200 hours, ~2000 USD, and some opportunities.
The average similar game sells about 42 750 copies. This has insane variance; there is a 55% chance to sell less than 2500 copies.
Expected cost of ~1793 USD per EA participation or lifestyle change conversion (With an upper bound of ~65 000 USD / conversion.)
Background (spoilers):
Following a bout with Covid, I've spent a vast majority of my free time during the past 9 months making Mineralis.
In Mineralis, the player follows the life of a military researcher, as they seek to end a conflict between two warring nations; one is orderly and highly technological, while the other rugged, poor, and led by rebels.
The protagonist holds a lot of power over the conflict, and makes decisions that impact its outcome. In the process, they can choose to become sympathetic to one or neither of the belligerents.
As they find out, the worldviews that are presented in the beginning are not the entire truth. The player is challenged to update them in a way that is morally more defensible, and may end up feeling conflicted about supporting the impoverished rebel nation instead.
The reason I'm making the game is twofold:
I'm frustrated at the shallowness of NPCs and story in most role-playing games - they're a huge missed opportunity in my opinion.
I want to find out if video games are viable as a medium for exposing people to EA ideas and concepts.
Analysis:
Costs:
Firstly, the time cost. I've invested around 700-800 hours of time into the development work. Additionally, planning/studying has taken around 100, marketing/release preparation around 60, and feedback collection 40. Including miscellaneous tasks, around 1000 hours seems like a fairly reasonable ballpark estimate. I'd estimate I'm about 85% done with the work (90% CI 70-94), which brings the total time of making Mineralis to around 1200 hours.
In that time I've implemented 440 features, written 300 pages of dialogue, designed a deep combat system, created dozens of maps, and made hundreds of events.
Given a median developer salary of 80 000 USD / year, it's possible to come to the conclusion that the time cost of Mineralis is equivalent to 50 000 USD.
There are also monetary costs. The greatest so far are marketing (~800 USD) and dialogue editing/proofreading (~680 USD). With other miscellaneous work and advertising, the total monetary budget is close to 2000 USD. (The project is somewhat funding constrained, due to high uncertainty about its cost-effectiveness and marginal benefit.)
Although I've been very efficient, there have been significant opportunity costs, and some social costs as well. I am grateful to my friends and partner for their patience throughout this experiment.
Benefits:
There are numerous "hints" towards EA concepts, although the game is not labeled or branded as an "EA game". The characters introduce and debate ideas related to cost-effectiveness, wealth inequality, prioritization, marginal effectiveness, opportunity cost, career planning, longtermism, AI risk, nuclear risk, and large-scale conflict.
There also are potentially relevant non-EA ideas related to e.g. philosophy, sources of wisdom, the nature of existence, mental wellbeing and depression, social anxiety, education, loss of loved ones, transhumanism, diversity, equity, equality, racism, societal structures, decision-making, teamwork, and conflict resolution.
The specific subset of discussions/ideas depends on the route (which depends on player decisions).
My hope is that players will be intrigued enough to further explore and learn about some of these ideas.
Cost-effectiveness:
Let's analyze possible upsides first.
Based on initial feedback, I'd expect the EA ideas to be of interest to around 20% of players (90% CI 2-35%). I'd expect some of them to be interested enough to actually look up more material about a topic outside the game sooner than they otherwise would have. One test player indicated such action, and my best guess for the size of that subset is 25% (90% CI 15-45%). Further, some of those players may stumble upon the EA movement and read an article or two, where they otherwise wouldn't have. My speculative guess for that subset is 10% (90% CI 4-85%).
A portion of those people may join the movement (defined as joining one event or writing one post here), or take the GWWC pledge, or change to a more effective career, or start a lifelong habit of philanthropy (defined as donating over 1%p of their net income more than their peers). My wild guess is that the conversion rate from an article into actions taken is 20% (90% CI 5-65%).
Of course, this funnel is full of assumptions. But it's one data point to consider.
Putting all those estimates together (in Fermi Paradox style), we get the following figures (assuming one can simply multiply CIs like this):
Total probability to study outside of the game: 5% (90% CI 0.3-20%) - one for every 30 copies sold.
Total probability to look into EA: 0.5% (90% CI 0.01-17%) - one for every 300 copies sold.
Total probability to participate or make lifestyle changes: 0.1% (90% CI 0-11%) - one for every 1500 copies sold.
The number of sold copies is higher than number of players. Based on Steam playtime data, I estimate that ~15% of purchased games are never played and ~20% are played for less than 15 minutes. Thus I estimate that three sales will result in two players.
Some set of players might be inspired to make a one-time donation, but otherwise not change their views or altruistic disposition. I think the overall impact of those players is negligible.
All in all, a 0.1% conversion rate to lifestyle changes seems like a reasonable ballpark to work with. I've played a few thousand games, and there have been a handful that caused me to update my views and habits.
What about the distribution of sales?
Based on a data set on Steam sales, a similar game is (very roughly) this likely to sell more than a specific number of copies:
copies
>2500
>10 000
>50 000
>100 000
>500 000
likelihood
45%
30%
18%
11%
3%
Based on the data set, the expected value is (very roughly) 42 750 sales. Plugging that into the guesstimate above, we can (in theory) expect ~29 players to make lifestyle changes.
Given the combined cost of 52 000 USD to develop the game, with the average sales value we get an expected cost of ~1793 USD for a single lifestyle change or active EA conversion.
However, you can see that the variance is huge. There is a 55% chance to go practically unnoticed. Considering the low marketing budget of Mineralis, I fear that the visibility may skew towards the lower end of that range.
Based on pre-marketing analytics data, I currently expect to have organic monthly averages of around 48 000 impressions, 13 000 page visits, 240 wishlists, and 50 purchases. That would lead to ~1200 sales over the first two years.
It's very difficult to guess the effect of marketing and events, but we can reasonably use that as a pessimistic lower bound. It would put the expected cost of a conversion at a maximum of~65 000 USD.
Fingers crossed.
Open questions:
Should the game break the fourth wall and direct interested players to an EA resource? Players hate losing immersion and even mentioning the name could be taken the wrong way by someone. Then again, I'd assume that anyone who e.g. clears the game with specific parameters would be interested.
Would EA benefit from funding the advertisement and marketing costs of the game? It would be particularly relevant right now, due to the opportune timing of the early October Steam Next Fest, which Mineralis is taking part in.
Although I personally enjoy working on games as a creative medium, it "feels" potentially very ineffective compared to, say, improving the UX and marketing funnel/tools of the AMF or GWWC websites or something. Any thoughts? What is the cost-effectiveness of typical EA meta-work?
This post presents the executive summary from Giving What We Can’s impact evaluation for 2025. At the end of this post we share links to more information, including the full report and...
I used AI to fix transcription errors, rerrarange the ideas, and suggest tweaks to the title and some sentences.
Three of the most exciting projects to come out of EA in recent years are, in a vague sense, CEA spinouts:
* Kairos is directly a spinout of CEA and now handles most support for university AI safety groups. Basically everyone I've found who knows them is really excited about what they do
* NEST is an opinionated ideas-fi...
This seems like a very cool project, thanks for sharing! I agree that this type of project can be considered a "moonshot", which implies that most of the potential impact lies in the tail end of possible outcomes. Consequently the estimated become very tricky. If the EV is dominated by a few outlier scenarios, reality will most likely turn out to be underwhelming.
I'm not sure if one can really make a good case that working on such a game is worthwhile from an impact perspective. But looking at the state of things and the community as a whole, it does still seem preferable to me that somebody somewhere puts some significant effort into EA games (sorry for the pun).
Also, to add one possible path to impact this might enable: it might be yet another channel to deliberately nudge people into in order to expose them to the key EA ideas in an entertaining way (HPMOR being another such example). So your players might not all end up being "random people", but a ratio of them might be preselected in a way.
Lastly, it seems like at least 5-10 people (and probably considerably more) in EA are interested or involved in game development. I'm not familiar of any way in which this group is currently connected - would probably be worth doing so. Maybe something low on overhead such as a Signal group would work as a start?
Hey Markus, thank you for sharing your thoughts!
Considering that I need to cast a relatively wide net to find even thousands of players, the messaging about Mineralis is not very selective on purpose. I think that the overall demographic will end up being relatively random within the set of people who are gamers and enjoy JRPGs. Whether JRPG fans have a predisposition towards EA - that remains to be seen. :)
I would be super humbled if the game ends up being a success and can be considered useful for the purpose of actually teaching EA concepts. However, I would imagine that a vast majority learns better through reading articles or listening to podcasts.
Setting up a group for discussion sounds like a valuable idea. I added a personal task for that and expect to get to it late next week. I'm thinking Discord might make the most sense, given that many developers already use it anyway.