This came about  from the YTD software sales comparison topic.

What I am looking at is the software sales over hardware sales of the three home consoles and their change in time.

So first up let's see what we are looking at:
First up Monthly Software Sales:

followed by monthly Hardware sales:

Now we start to look at attach rate, to start with a graph which provides no real information Monthly SW over Monthly HW sales:

What this graph shows us is as we expected software sales are increasing faster than hardware sales.

EDIT: Now onto more interesting graphs, the current attach rate, that is monthly sales over total hardware. This gives us an indication of how software is selling to the existing user base:

Here we see that except for the inital start the three consoles actually follow very similiar sales patterns with an almost identical current attachment ratio.

UPDATE: updated till end of August 2010

Someone needs to confirm this but is this the first generation with this high attachment rates all around?

Nice work. I find the Monthly SW / Monthly HW graph and Monthly SW / Total HW graph to be the most interesting for a comparison basis. I'm not sure how useful they are over time though. The Monthly SW / Monthly HW graph should naturally continue to trend upwards as the install base grows. The Monthly SW / Total HW graph should naturally continue to decline as the install base grows.

You may be interested in this thread: http://www.vgchartz.com/forum/thread.php?id=66975

Unfortunately, I haven't updated the numbers in awhile.

@Acevil Unfortunately the previous generations information isn't as available but it would be interesting.

@theRepublic: That thread is quite interesting, And doing that calculation weekly, ouch ;) . What I would really want to do is have a software per unit over life cycle, but obviously that information is not available. I'm not sure I fully understand the TWO calculation so will need to examine it some more.

It is all in a spreadsheet, so it is not too bad.  I just have to put in the numbers for each week.  I just haven't updated in a few months and there have been some adjustments to the numbers that I need to account for.  I also want to do a regional breakdown.

Its a little more complex than that.

You need to know when the software was sold relative to the purchase date of the console.
Then divide the total number of software sold for each month after purchase date of console by the toal number of consoles which have reached that month after purchase.

What you would expect is more software nearer the original purchase of the console pattering out later in life

Tendentially, it should relax toward a constant value characterising the average console owner, as it's basically an intensive property (number of games each console owner buys per month).

It might not work that way. The point of manipulating data is to remove bias, and make conclusions based on the data.


Making conclusions you suggest requires more and different data. I dont know if the OP has enough data in order to figure out whether the sale of software is at it's peak when it is sold and then decreases with the life time of the purchase.

Well the PS3 line can certainly thank me for helping it rise.

Otherwise, nice job on the info

You are right that may not work that way, which is why its the one I would really like to do ;)

But yes that data will never be available in the required form.