Friday Afternoon Stats Lecture!

Friday Afternoon Stats Lecture!


Maybe I'm just a nerd who's looking for any justification out there as to why he took a 9am "Statistical Methods of Psychology" course in college, but damnit, I think numbers are interesting (ooooh).  As a corollary to that (ahhhh), I think that all video game developers should treat their games in the style of Bungie and Valve: with raw, detailed statistical analyses of even the most minute of details.

I've been enjoying Valve's hardware surveys as of late.  They're an awesome way to get a snapshot of what's important to the gaming community -- although the latest survey's one-million users aren't necessarily reflective of everygamerontheplanet, I'm pretty confident that a survey of that magnitude can tell us a pretty good deal about what's going on hardware-wise.  In this case, nobody seems to run SLI, even more of nobody seems to run Crossfire, and nearly as many nobodies as the first example game on DirectX10-compatible platforms.  Yikes.

You can check out all the fun results for yourself.  Although the survey I'm referencing ended August 2007, there's a new survey that just started a few days ago.  Every website under the sun seems to be offering an article right now that directly compares the newest survey against its direct predecessor, but I think that's rather worthless.  After all, the current-running survey from around 130,000 users at the time of this post. The previous survey had over one million samples, making comparing the two a rather pointless exercise at this stage o' the game.

Even more interesting, however, are Valve's statistics for Half Life 2: Episode Two.  Criticism of the game aside, it's pretty neat to see details like average total time spent playing, average session, et cetera.  Check out the graphs, and you'll note that people have beaten the game in anywhere from 2.5 (!??!?!) to 10+ hours -- that's one heck of a difficulty curve.  That, or some people are just really enjoying (or hating) the scenery.

I also like the Achievements tab -- apparently nobody like sending the game's garden gnome into space (an epic task, for sure).  It's awesome enough that Valve is jumping on the achievement bandwagon with its titles.  And now, you can see just how 'leet you really are.   Seriously, check these stats out; they're geeky fun.

As it stands, 0% of Maximum PC Associate Editors have played Half Life 2: Episode Two.  However, 100% of that sampled group loved the crap out of Portal, and enjoyed themselves some moist, delicious cake.  Yum.



+ Add a Comment


Don't we all take part in this? If so, you'd have to assume it's accurate overall. If not, then who's opting in and out and making a survey like this inaccurate on all levels? If it's opt-in, it's missing mine because I'm ignorant of it.

The numbers show that there ISN'T as many of as we'd like to think. If we have 10K member in the forums, 10% might have some PC's under 1 year old AND backed with some beefy stuff. Just a rough estimate honestly.

I think the point goes to show that hundreds of thousands of others just want to play the awesome games on what they have, or what they can afford...or both.

I played HL1 on a Ti4200 when the 6800 was the beefcake. Episode two on a 7800GT when the 8's were coming out. I finally get to play TFC on hardware actually exciting. It's what we have, and one thing I hope the graphs tell people like valve. Keep older stuff able to run their games for years to come.

That's just graphics cards though. The rest is really numbers.



I, like the Murf love to look at stats like these and compare my rig to other PC's that are out there (although I did drop Statistical Methods of Psychology shortly after the semester started!). I have participated in previous Valve hardware surveys but did not participate in this survey.

I built my rig almost 2 years ago and based it on an AMD socket 939 architecture. I have 2GB of RAM in my machine so I'm in the minority there (0.43%) but I'm squarely with the majority of AMD users because I'm running a single core (75.36%) 3700+ San Diego Core at 2.29GHz (20.77%). So I'm above average in RAM but in the proc area I'm not sure if I made a good choice or not. For the money the chip I have is great, but I do wish I had gotten that FX-60 when I had the chance. I do have a dual core chip that I could put in my machine but I never felt it was an improvement over my old processor.

I think the most surprising stat for me was the video card breakdown. I could not believe how fragmented this area was! I don't see how video card manufacturer's make money when the pie looks to be cut so small. There is basically 2.5% difference in the top 5 video cards that a majority of the users out there are running (on valve games anyway).

Something else that stood out in my mind as I browsed the stats was the breakdown of the bus types. It looks like PCIe x16 is the majority with 51.75% of the PC's out there but AGP 8x systems are still out there chugging along at 28.67% (sadly I'm in this group). I would have thought newer systems would have more RAM, but I guess 1GB is still the sweet spot for most.

To wrap this up I guess in my mind the evaluation of my machine is like this: considering what I paid over 2 years to keep the PC running (aprox $700.00), the cost/performance I get is pretty darn good. I know it could be beter, but my rig can hold it's own in most of this Valve pack. Sure this is just my comparison of the sampling from Valve servers vs. my specs, but isn't that why stats can be fun? In this case they helped me to tale stock of my choices in relation to others who share a common interest. I feel pretty good about my PC in general, but man, I really do wish I had picked up the FX-60 and bought a PCIe 16x motherboard!



I'm thinking that these polls show that even though game developers want push the games to the leading edge of technology, the majority of game players may not have the financial ability to keep up with them. Games that are requiring Vista or DX10 platforms are aimed at a relatively small section of the gaming community.



actually I disagree; there are a lot of hobbyists, Maximum PC magazine should stand test to that - and not many people took the survey when you look at the grand scheme of things.. there are way much more gamers out there than the numbers show..

I'll give you my own personal example, I've been saving up to build a new $5,000+ dream worthy rig myself.



I participated in this.

My computer was top of the line when I got it but now my numbers are lower than most. Maybe it's the combination of parts that I selected, but everything still runs well on my 3 year old comp.
X800XT platinum ftw



I wouldn't boast about the X800XT Platinum if I were you; trust me, I own a X850XT Platinum, and let me just say - I'm glad that I'm building a new computer soon..


zine team279

This was a Triumph.
I'm making a note here -- huge success.

I found these a while back -- neat to see that they're 're-surveying' everyone, to get a better idea on what everyone is running.

Waiting for AMD's Tri-Core to show up on there soon.


Log in to MaximumPC directly or log in using Facebook

Forgot your username or password?
Click here for help.

Login with Facebook
Log in using Facebook to share comments and articles easily with your Facebook feed.