I’m biased. No, I’m biased!
This entry was posted on September 4, 2014.
(title with apologies to Spartacus) Pretty much from the start of this survey (results part 1, part 2, part 3, part 5, part 6), there were suspicions of bias in the results of this survey in regards to the ‘well-known’ difference between historical players on the one hand and ‘sci-fant’ gamers on the other. So today I’d like to show where we promoted the survey, what the likely effect was, and try to test the result data for any bias.
First things first: yes, WSS is a historical wargaming magazine (with some leeway – we go by ‘on this earth’) so if there would be a historical bias, that would not be too horrible for us. That’s our target audience after all! However, the difference between historical and sci-fant wargaming is somewhat artificial as was argued by Rick Priestley in WSS issue 61. Knowing that, we actually did try to cast our net as far and wide as possible. Multiple news items were released on TheMiniaturesPage, Bell of Lost Souls and TabletopGamingNews. The BOLS Gamewire service especially ensured that our (three) news items were posted to a very large number of wargaming sites. Apart from those three news sites, Beasts of War posted the survey on their website and in their newsletter, it was on Brückenkopf Online, the WWPD Network, and a large number of wargaming forums (both English-language as in German, French, and with some help, Spanish, Italian and Polish communities). Of course we promoted it through our own website, newsletter, and social networks. It was posted and cross-posted by many bloggers, twitterers and forumites everywhere. In total I think we posted the survey to some 25-30 sites, and it went much further from there. Was it possible to miss it? Absolutely, we didn’t have any nudity (you wouldn’t want to see us anyway), nor were any celebrities involved, but all in all I think we gave it the old college try…
Is a bias borne out by the numbers? Really the only way to distinguish between categories of gamers would be to check preferences for periods and genres. So let’s see what happens if we select only those who indicate they are “not interested” in SciFi or Fantasy or Pulp or Alternative History. Presumably those would be the ‘hardcore’ historical gamers. They are 3444 of 7759 respondents, less than half (!). If you turn it around and select those who indicate that either SciFi, Fantasy, Pulp or Alternative History “is a favorite”, you get 4413 respondents, and 3101 if you select only those who favorited SciFi. Within that last group, 43% also indicates that World War 2 is a favorite period to play (smart move there Warlord Games)!
Regionally speaking the proportions of respondents remain about the same, irrespective of which filter is applied. That’s nice, as it means that we did not get substantially more wargamers from one or the other group from a single region. Where you do see a difference is in the age groups, though the overall picture remains the same, it just shifts down (for SciFi) and up (for Historical) by about 5-10%. There’s also a tiny shift in gender, although the very small numbers of female respondents makes those results not the most reliable. There seem to be slightly more women interested in SciFant gaming than the average, and slightly fewer among the ‘hardcore history’ types.
Is there anything we can conclude from all this? For one, it does look like we managed to get responses from wargamers with very mixed interests. Whether our proportions reflect reality is impossible to check. Maybe we should try this survey again at a later date… It also suggests that Rick Priestley was right, but then, maybe that was predictable…
The three charts below show the interests of different groups of respondents. To the left are the interests of those who indicate “Not interested” at least once for Sci-Fi, Fantasy, Pulp/Steampunk or Alternative History. The middle chart is for those who indicate Sci-Fi is a favorite, and to the right those who indicate Fantasy is a favorite. It’s quite clear that there’s quite some overlap.