Valve spied on me! Cool!
After answering @SheBeeGee’s short notice plea and writing a rather niche opinionated article “TF2: A journey to above average” for her new project Nerdmag yesterday I noticed that I never write anything about gaming. I recently stumbled onto some pretty interesting stats that Valve have been capturing via their Steam platform. Combine these two findings and you get my first Geekissues gaming article!
To quote the team themselves: “At Valve, we’re always striving to make our products better. We’ve traditionally relied on things like written feedback from players to help decide which improvements to focus on. More recently, Steam has allowed us to collect more information than was previously possible. Episode Two includes a reporting mechanism which tells us details about how people are playing the game. We’re sharing the data we collect because we think people will find it interesting, and because we expect to spot emergent problems earlier, and ultimately build better products and experiences as a result.”
At first I was a little taken aback – you mean you’ve been SPYING on me? Yeah like that’s a huge surprise these days. But then I had a browse through what they’ve been capturing (at least what they’re making public) and it’s very interesting high level information; all of which having a very practical use. I’m going to briefly discuss a few of their stats for Half-Life 2: Episode One & Half-Life 2: Episode Two. The entire collection can be found here.
First up: “Sessions played“:
Episode 1: 4′650′952
Episode 2: 4′149′046
Unfortunately they don’t give an exact number of units sold, but this number gives a great indication of the “legs” for their product. Imagine you sell tennis rackets. I bet you have solid figures for the number of units sold; do you know how many games of tennis the average racket gets to see? Nope!
“Average session time”
Episode 1: 0h 35m
Episode 2: 0h 41m
Each time average Joe gamer sat down to play episode 2, he did so for a full 6 minutes more than he did when he played episode 1. That’s an increase of ~17%. At first I wrote this sentence next: “One could argue that this bump in average playing time makes up for the drop in “sessions played” and it’d be an interesting debate, but I don’t want to commit to that long shot without a little more data.”. Then I decided to stop being silly. If you take “sessions played” times “session time” you get the following:
Episode 1: 162′783′320m
Episode 2: 170′110′886m
How awesome must it be to be Valve and have the ability to sit back and say – “ep2 was played more than ep1 so we did a good job” without relying on opinionated reviewers or sporadic tester information.
Note: I know the session time is an average so these numbers won’t be exact, but it’s to prove a point.
Moving on to less back patting and more useful information have a look at the average number of deaths graph for ep2 along with the death maps. This sort of information will become invaluable for game developers as they release updates, fix bugs and try to keep the game-play experience as balanced as possible. You can see a clear trend with the number of deaths across the whole graph except for 2 places with noticeable jumps. Correlate those jumps with the fact that there’s a noticeable drop off of percentage of players that made it past ep2_outland_12. What’s the money that this information will bring changes to those specific maps? No one wants to play a game that is SO hard you don’t want to finish. You want a steady challenge.
With all this useful data to play with – I really like the fact they spied on me! Go Valve, go make my gaming experience better.
