Tales From Discoverabilityland: A Story Of Three Graphs

[Hi, I’m Simon Carless, and you’re reading the Game Discoverability Now! newsletter, a regular look at how people find - and buy - your video games. Or don’t.]

Welcome to the latest Tales From Discoverabilityland (phew, that’s a long, invented word!), the latest in a series of posts that round up interesting data or trends around the game discovery space.

Having delayed this one from last week due to splitting out a story, there’s now a gigantic amount of interesting things to round up. So let’s go for it.

Steam releases, ‘success’ and bar graphs.

In our ‘lead story’ for this round-up, I wanted to focus on the deep data dive Valve recently provided into how new Steam game releases are performing over the past few years.

I’ve heard a couple of opinions on this, including that it felt like a bit of a PR exercise. Which I partially agree with. (However, Valve is also very transparent, even in showing some negative effects for smaller games between 2018 and 2019.)

And it does reveal that after the ‘restrictive’ access to Steam got opened up starting in 2014, there are definitely more games (in total) earning over certain thresholds of money on the service. See below graph:

Now I’m guessing you’ve read the Steam take on the data. (Or if you haven’t, you have now!) But I thought the follow-on articles and graphs were also interesting, because they flipped the script.

For example, Shaz Yousaf’s excellent blog for Gamasutra inverted how we look at the data, and simply asked: ‘Of all games that launch on Steam, what chance do you have to make $10k just after release?’ (Eevee did similarly on Twitter.) And Shaz got the following graph:

So this certainly makes it look like ‘competition’ got a lot tougher, if you just look at the raw numbers on that chart.

Shaz also suggested that ‘blaming’ Steam for this lack of success as a % may not be completely correct. He talks intelligently about the Pareto Principle in the piece, noting that although Steam are now trying to combat this with Steam Labs:

“Essentially, perhaps because of our innate behaviours of forming hierarchies, worshiping the same idols and copying what those around us do, most of us tend to talk about the same things, play the same games, listen to the same music and [insert activity] as those in our perceived community.

This means you get a small number of “big winners” and a large number of “small winners” if winning means achieving popularity in any given domain. In other words, it's not "Steam's Fault" if the most popular games massivly overshadow the other games.”

Finally, Kyle Orland at Ars Technica generated a bunch more handy analysis from the data, trying to stitch together the multiple graphs that Steam provided to get more of a combined bar graph - see below.

So, there’s three different takes on the same data. And they look almost completely different, right? Or at least, are different perspectives on the same issue.

Specifically: Steam used to be a hand-picked selection of games (that were way more difficult to make, before cheap/free game engines & access to game assets).

Now it’s a storefront that’s much closer to the Amazon or Spotify model of ‘put your media up with a LOT of other media’, including hobbyist & semi-professional media. It’s rare to see this change so transparently observed. And it’s pretty damn interesting.

Short takes, short takes, short takes!

OK, now there’s a lot of other things that you MIGHT be interested in, but don’t have to be. So I’m going to the one-paragraph model:

Shorter takes, shorter takes, shorter takes!

And we’re all the way down to the one/two line commentary here, phew: