Analysis: Steamworks sales estimates "hack"

Long story short, a guy managed to estimate very accurately the sales for a lot of games on Steam using the Steam API. This method no longer works because Valve quickly made a patch but a lot of useful information was already downloaded. This caught significant attention from the game dev community because there isn't much data about games sales, specially now that valve is restricting more information to the public leading to the painful death of the best source available out there: Steam Spy.

In this article I'm going to write about my analysis on this data. I've worked on a game published on Steam and I liked the experience so I might continue publishing more on the future. That's why this data is important for me.

What was leaked?

Sales data containing an estimate on how many units were sold for about 13k games (games that integrated Steamworks achievements). You can check all this data on a raw CSV file.

This is how the data looks like

What was not leaked?

Please don't get confused, this doesn't gives us an exact amount of money earned by game. Please don't multiply the number of sales times the price of the game, that number wont mean a thing. You have to take into account the following:

  • Most sales occur on discounts
  • Valve's cut
  • Taxes, they are different depending on the country of purchase
  • Refunds? I don't know if this applies though

So, why is this important?

I think this data gives us something I find way more important than money: Users. For me, being able to build a big community of users is something very remarkable. So I did this graph to help me understand better:

y represents the amount of users for each game, x is just a users count "ranking"

This graph shows that just a few games have a huge amount of users compared to the vast majority. This might sound discouraging for game devs because it looks like userbase is rigid and centralized. Getting into the top 100 or 1000 looks distant.

However, things look different if we break it down:

This is the same but only showing the top 100 games

Same but only from 101 to 1000

Same but only from 1001 to 5000

Same but only from 5000 to 13000

If we break it down, we can see that the growth is always exponential. This means that making good games pays better. This applies to all. Indies and AAA. Small or big budget. Niche or not. Sure, Team Fortress is in the top of that list, but it looks like we haven't even find the limit of this exponential growth.

Let me put it this way: Making one good game pays better than making two crappy ones.

I think this is great because it means game devs are encouraged to make better games and this leads us to a better ecosystem.

I'm not always happy with Steam tools and decisions. But I think this numbers reflect how great the Steam community is and how good Steam is to find the good games. Kudos for that.

Ahmed Castro

Read more posts by this author.

Subscribe to Rosalila Studio

Get the latest posts delivered right to your inbox.

or subscribe via RSS with Feedly!