The PC gaming market-intelligence resource Steam Spy lives. Its creator, developer Sergey Galyonkin, is shifting the tool to a new algorithm that calculates the sales of PC games on the the Steam distribution platform based on “coincidental data” from around the internet. Galyonkin had to make this change after Valve began hiding its customers’ game libraries by default after years of exposing that information through its API.
In response to that privacy change at the time, Galyonkin said on Twitter that Steam Spy would not have the data it needs to operate. But he is now saying that the site will continue, but he needs some time to dial in his algorithm and integrate it into the site. Galyonkin said he wanted to solve this problem after indie developers from around the world reached out to him.
“I received over two hundred emails and messages from developers telling me how Steam Spy improved their lives,” Galyonkin wrote in a blog post. “There was an indie company from Berlin that managed to secure financing from the government for their niche title because they had the data to prove that this niche is big enough. The title got released and succeeded.”
The need for reliable data to enable smart business decisions is not in question by anyone. The problem, instead, is that the gaming market is so large and dynamic that algorithmic models are often inaccurate.
June 5th: The AI Audit in NYC
Join us next week in NYC to engage with top executive leaders, delving into strategies for auditing AI models to ensure fairness, optimal performance, and ethical compliance across diverse organizations. Secure your attendance for this exclusive invite-only event.
The good news for Steam Spy is that Galyonkin’s new math does seem capable of making estimations within an acceptable margin of error … some of the time.
“Frostpunk devs just announced that the game sold 250,000 copies and the new algorithm estimated it at 252,000 copies,” reads Galyonkin’s blog. “[But overall, it’s] not very accurate, to be honest.”
A small pool of developers have shared their sales data with Galyonkin. He has the numbers for approximately 70 games. For 90 percent of those, the algorithm was able to calculate their sales within a 10 percent margin of error.
“But I also saw some crazy outliers,” said Galyonkin. “Where the difference between the estimates and the real data could be fivefold.”
Steam Spy will still charge ahead. Galyonkin has closed off most of the site’s features while he makes some changes and improves his system. But he does plan to reopen essential information to the public once he gets everything fixed.
