Above we have GW's reported revenue dating all the way back to 1990, adjusted to today's dollars (ok, in this case British pounds - and I did use a British dollar inflation adjuster not an American one). I will point out, that I eye-balled the data points to make this graph based on another one I found on the web. I was not going dig out 25 years of GW reports, if I could even find them all, and pick out the revenue line items. I did however confirm that 3 of the data points in the original plot were correct. His was also not adjusted for inflation, so if you search and find it, it will still look a bit different.
Anyway, back on topic. This does not really tell us anything. It is assumption time!
Assumption: Since revenue is sales, if we happened to know the average amount of money that a GW consumer spent on their products on a given year we could back out the number of consumers! Wait, we don't know that! But, I think for this exercise we could assume that if we knew that number it would be fairly steady year by year: ie, averaging it out over the entire sample size, I think we'd see a fairly consistent year to year number for expenditures. It is at least good enough for this I think. But, I still don't have this number and no way to get it. Or do I... Let's make it arbitrary (bear with me, math will make it not matter). Hell, let's make it 1 British Pound the inflation adjusted average expenditure of all of GW's consumers.
This assumption now converts the chart above into arbitrary number of GW consumers by year but still doesn't show/tell us anything. Yet. Let's pick a reference point, 2013 the year that GW killed off specialist games. We can now calculate the % change in GW consumers, for the years after, relative to that reference point and by such remove the arbitrary "average expenditure of GW consumers" from the analysis!
So there you have it. Yes, it is some fuzzy math but a lot better than anecdotal evidence in my opinion. Because of my lazy plotting, the percents stay 0 until after the year of reference, it was just easier that way. Also, why did I choose those specific years of reference?
- 2001: GW acquired the Lord of the Rings license. More on this shortly but I thought it was important to see a relative point prior to that.
- 2008: GW halted support and development of Specialist Games
- 2012: Zombicide was kickstarted and released a flood wave upon this entire industry and hobby. Also, GW killed off it's historical game line.
- 2013: GW killed off Specialist Games
Unfortunately, not really enough data to look at the impact of Age of Sigmar and the destruction of the old world yet.
The Lord of the Rings Problem: Oh yes, that huge spike leading up to 2004. Boy if we used that as a reference it would show a very dim picture. And that is really an issue: How to handle the acquisition of the Lord of the Rings license. Obviously the growth leading into 2004 could never have been expected to be sustainable. And I don't think even sustaining the peak they achieved was a consideration, it was a cash grab pulling in a fair number of non-gamers and LotR collectors while there was momentum from the movies. Additionally, with the release of entirely new product line - I bet a lot of existing GW consumers "found" some extra hobby money those years, violating the constant expenditure averaged over years assumption , and there is no good way for me to reconcile that. I admit that. What I can do though is fucking ignore it! Instead grab a reference pre-LotR and look at other significant references that happened after that era, that are often cited by Chicken Littles ("every since blah, GW has been spiraling down"). Generally speaking, the Chicken Littles don't seem to be referencing anything back beyond 10 years anyway, so they don't get to use the that data point to support their anecdotal evidence IMO. It is really to bad that LotR is messing up that era though, because Warmachine came out in 2003 and from my "perspective" was the first game to really start pulling players away from GW.
Anyway, that is the meat. Part 3 I will discuss how I view this data, some other thoughts (dangerously close to anecdotal) on this topic and final thoughts. Hopefully it will be shorter.
The analysis assumes that per-capita average year-on-year consumption is constant, whereas it should be variable. But I suspect that the trends are significant enough to push the effect of variable consumption into the noise.
ReplyDeleteThe 2001 Ref is the only one that is needed. All of the others only produce y-axis shifts. The function should yield the same curve, just shifted because of a different value for the initial condition. I think. It's been a long time since I've done any real math. Feel free to correct me.
I would really like to see what the curve looks like across that entire history of revenue from the first post.
You mean as 1990 as the reference point, so to speak? Certainly doable but slightly misleading specific to the topic. It is easy to look at that the resulting plot and say "Wow, 300% increase in 25 years!" and ignore it has been sloping down to 300% for several years. Also, I worry that the assumptions become more inconsistent in the 90's era: Practical no competition for GW, the hard-core grand tournament scene hadn't developed, etc. Also, at least as I recall, typical game size was much smaller back then. Etc.
ReplyDeleteI will send you the plot though, eventually. Excel and I aren't getting along right now. LOL