Valve Corp’s Steam client may be one of the most popular platforms for buying digital games on the PC, but analyst Sergey Galyonkin, creator of Steam user statistics website SteamSpy, says that you can’t trust the raw numbers when it comes to gauging the size of your audience for a game launch.

“When marketing research is talking about MOBA popularity and MMORPG decline they automatically assume it should work like in other industries — people switching from Gatorade to Coca-Cola, while there is plenty of space for smaller imitators and local brands,” Galyonkin wrote in a recent article for Medium. “But games, especially big ones, aren’t consumer products. I’d go as far as to say that each multiplayer game is a cultural self-reinforcing phenomenon, relying on its perceived popularity more than on its market share.”

He added, “It’s not enough to just offer a superior product with better marketing and brand recognition to convince people to switch from League of Legends to, say, Infinite Crisis. Yes, the latter has DC characters, TV series, comic books and even some movies. But the former was here first and didn’t just create the market — it is the market.”

“A huge success of one big title doesn’t mean much for everyone else”

Galyonkin explained some of the demographics targeted by publishers, such as “female gamers” and “core gamers,” do not technically exist as a cohesive group, as both definitions are too broad to offer any meaningful data.

“The term ‘female gamers’ includes both a woman in her fifties playing Candy Crush Saga on her phone and a college girl enjoying Call of Duty on her Xbox,” Galyonkin wrote. “They’re so far apart from each other, that it makes no sense to try and fit them into the same vaguely defined category. There are many female gamers, they’re different and there are probably dozens of categories you could divide them in.”

According to Galyonkin, too many new games rely on a market that he says only exists because of one or two particularly popular games. For example, the MMORPG market exploded thanks to World of Warcraft, but no other games have been able to capture that level of success in the 10 years of WoW’s existence.

“There were no big successful MMORPGs after World of Warcraft not because WoW took all the audience, but because there were never too many people in ‘MMORPG but not World of Warcraft‘ market to begin with,” he explained.

“I think when you start thinking in terms of audiences for individual games instead of broad vague ‘MMORPG crowd’, ‘MOBA crowd’ you’ll start to realize that sometimes a huge success of one big title doesn’t mean much for everyone else. It doesn’t expand existing market or destroy it, it creates a new one.”

Hardcore gamers own fewer games than you think

There are plenty of niche PC games that could never hope for mass appeal, but they try to capture the so-called “hardcore gamers,” people who spend significantly more time and money on PC games than anyone else. Galyonkin admits that there are plenty of PC gamers, but he says only a very small percentage of Steam users actually own a large number of games.

The top 1 percent of Steam users own 107 or more, but the top 20 percent own four at most.

“Various studies suggest that there are 700–800 million of PC gamers,” Galyonkin wrote. “It’s probably true, but it doesn’t mean much for your game. Because if you’re developing a downloadable game for Steam you’re not even fighting for 135M of its active users, you’re fighting for the attention of 1.3 million gamers that are actually buying lots of games.”

How accurate is SteamSpy’s data?

While SteamSpy gathers data on a large portion of Steam users, it is important to remember that its coverage is not absolute. SteamSpy can only see the game libraries of users who have set their Steam profiles to public, meaning that an indeterminate number of private users could skew demographics for many games one way or another.

The site does include a margin of error for each game however, showing for example that the total number of players for Dota 2, currently listed at over 9.2 million, could be higher or lower by roughly 77,200 players.

Other numbers can also be misleading on SteamSpy, such as the total number of owners for a game. This is especially true of free-to-play games, which can have ownership numbers that are orders of magnitude larger than their actual player numbers. This is due to the fact that someone can easily download a free-to-play game, play it long enough to decide they do not like it, and then never play it again.

There are also several games that frequently show up in bundle deals, such as the Humble Indie Bundle, which may be purchased by gamers who have no interest in playing them simply because they come with other more desirable games.

Despite these limitations, SteamSpy does seem to be fairly accurate—more accurate than any market survey at any rate.

Image by Eric David | SiliconANGLE