Quoted from YeOldPinPlayer:
Dennis, I don't understand how you correlated ownership growth reported on Pinside to ownership growth overall. If pinball ownership did not grow but Pinside membership did how would your data show that?
Sadly, it falls under one of the flaw categories I listed (self-reported ownership; as you know it introduces a lot of issues). That data is incapable of distinguishing ownership increases' causes (be it just site growth or something else). It's why I do not actually provide a correlation value with a statement on statistical significance, because I can't really detect that without knowing the manufacturing quantities. And, if I had those, I'd just use that for ownership (since all games not destroyed would be owned by someone, and we'd just accept the flaw that we can't track destroyed as a small price to pay), though then we could compare ownership counts here versus production counts and probably get a sense of what percentage of owned games is reflected on this site (I wouldn't have bothered, but it would definitely be interesting).
My non-data based assumption at present is Pinside growth is collector-oriented and should, in theory, benefit all years, particularly since the range of examination is still "modern". Because people getting into the hobby now, or joining the site for the first time, still have access to all the games in the period of examination.
But your point is well taken. It would be like the idea of reporting Pinside's membership growth. The growth itself would be true (obviously), but we'd not be able to prove (statistically) that the growth was due to new hobbyists (in theory it could all be old hobbyists finally finding/joining the site). We'd need some other data source (surveying the members when they joined) or, we could use it to fuel our suspicion and then try to find another data source to help shore up what assumptions the data we had were making us think (what I tried to do with the IFPA data; more on that below).
Where I'm more concerned (and why I noted the flaw that the site seems very collector oriented) is when looking at the ownership count of Stern Pros versus Premium/LE. In several cases, the Pro figures are LOWER than the Premium/LE. I do not think this corresponds to reality (I think Stern sells more Pros than LEs and Premiums combined; don't know it for sure of course, just my suspicion due to costs). However, if the site's bias holds true (towards collector-grade games) then at least I feel relatively safe in assuming that's consistent across years. But I can't prove it, so the best I could do is list the flaw and move on.
I gathered the competitive data after the owner figures here because I was pretty disappointed with how many assumptions I had to make about ownership. Since the exploration was on if the pinball hobby was growing, I thought another dataset would help shore that up (since what I do have on ownership was suggesting things are more produced/owned now than 10 years ago). The IFPA data offers a lot more samples and a lot more consistency, and so it helped reassure me that the ownership trends on Pinside fit with the narrative (that pinball, the hobby, is in a growth phase).