Jordan Faas-Bush (GSAS '19)

In this post, we sit down with Jordan Faas-Bush (GSAS '19) to talk about his site, SteamSeer. SteamSeer takes its inspiration from Lars Doucet's now-defunct SteamProphet, which allowed users to predict the success and sales of independent (i.e., indie) games released on Steam. Like SteamProphet, SteamSeer seeks to gamify the grueling process of market research for early-career indie developers, especially those new to thinking about the business side of game development. 

Click on the accordions below to read through our interview with Jordan! 

Note: Some content has been lightly edited for readability. 

My name is Jordan Faas-Bush -- I started at RPI in 2016 and graduated in 2019, just before COVID. I went to RPI knowing that I wanted to be a programmer and do Computer Science, but not knowing where I wanted to focus from there. In fact, I explicitly did not expect to go into game development. I enjoyed programming, though, and eventually found the RPI Game Development Club, which was then hosting monthly game jams. Participating in those let me work with a bunch of different people and experience the whole game development process in this condensed, intense period of time. Even so, I still didn’t really think I was going to go into game development – but then I started taking a couple of classes, then I added it as a dual major, and now here I am in the game development industry, after all. 

Yeah, so after graduation, I went and started a little game development studio [called ToothPike Games] and developed Load Roll Die, which is a rogue-lite dice-building game. It’s like a deck-building game, but instead of shuffling cards, you’re rolling dice and adding different mechanics to the faces of your dice. You combine them, you modify your dice, you reroll everything, and you aim for better combos. 

When I was making the game, I’d previously gone through a number of the programs that RPI offers – like Level Upstate – and did my own research trying to figure out the markets, what games do well and what games do poorly, and narrowing my focus to what I myself am capable of making. I’m not as good at art, so I was trying to find a game genre that I could complete at a reasonable quality level and would hopefully do well. 

I did a decent amount of work trying to figure out what to make -- and Load Roll Die did as well as I predicted, but it wasn’t a sustainable income source. So, I went and got a job in the industry and currently I work as a backend tools engineer for ControlZee, which is a startup competitor to Roblox. 

Before I got this job, I wanted to work on a couple of projects and find a way to give back to the community and also help myself for the next projects I was going to be pursuing. One of the things I discovered was just how absolutely insane the Steam Marketplace is – I’ve got some stats here and in 2022, 12,562 games were released on Steam. That means that every day about thirty games are released on Steam, so if you’re making a game, your game has to compete with all of these. Obviously not all of these games are going to be the tip-top of AAA games, but you still have to find a way to be discovered and to just get your game seen by people in this absolute flood of games.

Previously, one thing I’d discovered was a website called SteamProphet by Lars Doucet. This was basically like a fantasy league to try to guess how well all of the games that are released on Steam will do. It let you pick the ones that you think will do the best as a way to get yourself thinking about what games actually do well and what games do badly. 

Now, Lars ended up shutting down SteamProphet after Steam changed its APIs to make it harder to get solid sales results, but there are still a number of other cool ways you can try to determine how well a game is selling. They’re not as accurate, they’re obviously estimations and can be affected by any of a number of different variables, but there’s a well-known one called the Boxleiter method that lets you take a multiplier of the review count to get a decent estimation of the number of copies sold. For some games and genres, it varies, but the idea is roughly that it’s enough for you to figure out, as a developer, whether a game will sell well enough to allow you to eat dinner at night. 

Exactly. I ended up creating a second version of SteamProphet with permission from Mr. Doucet, which uses that Boxleiter method to estimate how well a game has done. Again, they’re not exact sales numbers, but we can then calculate a rough estimate of the money that game made to determine its sustainability. It’s my goal to be able to support myself by making games, but it’s a hard industry to be in, especially if you’re a smaller company. 

If you take a look at this link, there are currently about 44,000 game developers listed on Steam. Looking at the number of games each developer has made, about 33,000 of them have only made one game. It’s possible they might’ve changed names or something, but it means that three-quarters of game developers weren’t able to make a sustainable living with the game they released. And only about 6,000 developers have released two games under their name. So, if you’re looking to make a company, you’re going to have to release multiple games over time – you’re not necessarily going to get a hit with that first one and it’s important to look at how it’s possible to make a sustainable company and a sustainable source of income so you can eat food everyday. 

There's a pretty robust explanation on the homepage, but basically the list of games is updated every Saturday with releases for the next week. For the next week, you can go through the list and predict the number of reviews each game will get during its first month -- and you can predict a review range, the exact count, or both. You can also record how certain you are about your prediction or any additional thoughts. The idea is that you'll then be able to come back after a month to compare your predicted reviews against the actual reviews received. 

I’ll also mention that the site currently doesn’t require you to make an account. If you don’t make an account, it’ll save your guesses on your local computer; when you come back in a month, it’ll report back on your predictions. If you do make an account, it’ll import those saves and work across any computer you use.

I think one of the amazing things about the industry is that there are so many people who are excited to help you. There are a number of fantastic GDC talks about making a sustainable studio or surviving if you don’t get a hit and if you want to just dip your toes in the water, finding some of those videos on YouTube yourself is a really easy method. 

If you’re trying to move up a level and immerse yourself even further, Level Upstate was super helpful because it talked about not only things like market research and target audience, but also these other aspects of running a game development studio or larger business. You can find out some of it yourself, but it’s so much easier if someone sits you down to tell you about them. It taught me about things like the legal side of game development, working with IP, looking for grants, and even ramping up production if you do experience success. You can certainly find fantastic GDC talks about looking for publishers or finding funding, but sometimes it does help to just be able to talk to these local founders of studios who have knowledge and experience about these things. It also lets you form these connections with local resources, like Tech Valley Game Space or the Albany IGDA or any of these amazing local studios. 

Right now, it has a lot of features, but it’s not as focused as maybe it could be. When I started work on this, I really wasn’t certain of the best way to predict how well a game would do. If you try it yourself, you can predict by making an exact guess or choosing a bucket [range] of reviews, you can give a certainty, and you can write notes if you think a certain thing is good or bad. I added all of these options because I wasn’t sure of the best approach. It’s cool to get close with an exact prediction, but it’s also really hard when you’re starting out to make a prediction with no context. I would love to focus and streamline the process of making a prediction, as well as provide more after-the-fact analysis. 

Currently, if you look at how well games have done, you can see a graph of how many reviews they’ve received over their first month of release, as well as how that compares to your prediction. I’d love to provide a higher-level analysis of your predictions versus actual statistics over all of the predictions you’ve made, perhaps to see if you’re correct with a certain genre or if you consistently overestimate with another genre. 

I think the site is currently good as a reason to immerse yourself in the market by gamifying that experience. It may not give you a reward, but it gives you a framework to keep coming back and learning more – and I want to help it do a better job of providing better feedback and more teachable moments as to how well you are predicting. 

Yeah, I think in many of the classes at RPI, and really in every game development program, you end up taking a lot of roles across different projects. Even if that’s not explicitly the goal of the program, it sometimes feels like you should go into indie development because you’ve done so much. That’s great, and I’ve done it myself, but I think before taking that full leap, it’s important to take a look at the other aspects of making indie games that you may not necessarily learn through your classes. 

You don’t have to market your game in class, you don’t have to necessarily add accessibility features or controller support, you don’t have to make sure it does well enough that you’ll continue to be able to eat. It’s super useful to first take that moment and dive into the other aspects of development that aren’t part of your degree. 

Relevant links are included throughout the interview, but Jordan also passed along a couple articles pertaining to sales and revenue predictions on Steam. Both discuss the extent to which early (i.e., Week 1) sales figures can be used as a means of predicting long-term (i.e., Year 1) success: 

  1. "Can Week-One Steam Sales Predict First Year Sales?" by Jake Birkett
  2. "Steam: The State of 'Long Tail' Revenue in 2021!" by Simon Carless

And on that note, a final thank-you to Jordan Faas-Bush for taking the time to talk to us! 

Back to top