Do you gather analytics on your game?

[bbgames]

This is sort of a question for developers who have a game that's "out there", or about to be.

Do you use any analytic/metric-gathering tools in your games? Things like Mixpanel (http://mixpanel.com), KISSmetrics (http://kissmetrics.com), or CrazyEgg (http://crazyegg.com)?

Which ones do you use? Why those ones? How has gathering metrics changed the way you approach adding features to your game?

* If you haven't tried out any of the above systems and would like to, signing up for Mixpanel using this link, will get you an extra 5,000 data points on the free plan.

[jannesiera]

Piwik is also a nice google analytics alternative worth mentioning. I use it for my blog but don't have experience in game statistics.

[Chris]

Custom made + google analytics (+ some public counters to throw the data into advertisers eyes but I do not use them for inernal tracking)

[Fizzadar]

I'm writing my own stats program currently, which is done except for displaying the stats. Tracks everything in 'visits' and tracks each visit across the site, so you can get a good overview of where people are moving exactly.

And for my game I'll throw some custom stats in.

[bbgames]

Do either of you do any Funnel analysis on your games? For a few of the projects I'm working on we've been finding out a lot by setting up and tracking some basic funnels - things like "completed the tutorial", "finished the first quest", and so on. Spotting funnels that users aren't finishing has been a great way to find areas we need to improve on.

[dsheroh]

I just opened my game (Psi Rangers) up as a public alpha yesterday and set it up on Google Analytics as a matter of course.  I hadn't considered funnel tracking, but it seems like it would be a good thing to consider adding once I start writing funnels to track.    But getting core-feature-complete comes first, of course.

[Chris]

I have mixed feelings about funnel analysis. Looking at my games, would it help me if I know where exactly problem is? Unlikely, poor games are poor, good games are good. The games I would want to use funnels for are already too broken/poor for it to make a difference. Besides, would knowledge where the problem is tell me how to fix it? Unlikely. Most things can be checked better by keeping in touch with players anyway.

Funnel analysis would be useful in rather few cases. Like feedback from new people (who quit before finishing tutorial) or like some warnings set up for basic things (for example if registration system did not broke down due to some changes I made in the code).

If I already had funneling analysis system I would probably check it from time to time (mostly out of curiousity), but if I had to write it I would find it a waste of time.

[bbgames]

If I already had funneling analysis system I would probably check it from time to time (mostly out of curiousity), but if I had to write it I would find it a waste of time.

The nice thing is, a service like KISSmetrics will handle that all for you. You paste a bit of javascript into your page, go to your dashboard and mark down which pages are which steps in your funnel, and just watch the data roll in. Then from there, you can make observations like "most users drop off after introduction page 2" or similar. It takes maybe an hour to get set up (I think it took me like 15 minutes when I was checking it out), and you don't have to worry about building anything yourself.