EPTOver the last year, we’ve ran a few viewer surveys during our events. Although most of these numbers are pretty unsurprising, being able to back them up with real data is very cool! Plus, this is the first SC2 event we’ve ran since 2011, so we can compare demographics between current League of Legends and StarCraft 2 viewers.
The three surveys I’ll be taking a look at today are for the CLG Premier Series (League of Legends, February 2013), the Spring Promotion Tournament (League of Legends, December 2013), and of course, the Ender's Game on Blu-Ray Tournament (StarCraft 2, February 2014). It's probably worth noting that we only did the Online Group Stage for the Spring Promotion Tournament, which lasted 3 days: December 6th - 8th. In the future, I’m going to try to gather as many possible points of data possible, ideally over a smaller interval of time, with as diverse an audience possible (so, multiple games in multiple regions)
Some notes on each of the surveys we did for these tournaments:
Ender’s Game Tournament is still technically in progress. We will be keeping the Survey URL live until a little after the showmatch on the 22nd is done.
CPS had an amazing 1254 total surveys submitted, far and away the highest number sent in, despite being in the middle of the three events in terms of viewership.
SPT had an abnormally low survey submission rate, likely because we were sharing so much time between the link to the survey and the link to the sponsor. The event itself was also the shortest of the three, although it did have the highest peak viewership (~50k)
I’m also making several assumptions here that seem pretty reasonable to me. First, that lying/data skewing/joke responses rate is approximately consistent between all three event surveys. Second, I’d like to acknowledge that the very ability to watch streams on Twitch means that the majority of these viewers are generally in a specific demographic- namely, young, tech savvy males (obviously this will never be 100% true, but it might skew these numbers a little bit). Lastly, I’m going to ignore the possibilities of major calendar-based viewership shifts, mostly because that’s really hard to thoroughly address and I’m not knowledgeable enough to take a stab at it.
Unfortunately, some inconsistencies in how we asked the questions over the last year has made me question a few of our results, (mostly because we may have unintentionally skewed things one way or the other in at least one event’s survey) so I’ll be limiting the information I share to the following topics:
Gender
Age
Education
Income
VOD viewing frequency
Live stream viewing frequency
eSports Sponsor Support history
Adblock usage
Okay, let’s begin with Gender. It’ll be no surprise that the vast majority of viewers are male in both games, and all three surveys reflected that.
+ Show Spoiler +
![[image loading]](http://puu.sh/709HF/b6323b4e15.png)
![[image loading]](http://puu.sh/709Jr/855566ffb8.png)
Yeah, a whopping 90-94% of the viewers were male, and interestingly enough, only about half of the remaining survey takers felt comfortable being identified as female. Not much else to say here, so moving on to…
Age is a bit of a sensitive topic, depending on who you talk to. Surprisingly, SC2 has the highest percentage of both very young (<12) and what I’ll call “SirScoots”-tier old (> 40):
+ Show Spoiler +
![[image loading]](http://puu.sh/709ND/be5cc7d5ca.png)
![[image loading]](http://puu.sh/709Oc/b695057d26.png)
To make up for this disparity, League is averaging about 10% more in both the 13-18 range and the 19-25 range, with SC2 more than doubling the percentage of 26 to 39 year olds. What was incredibly surprising to me is that there were more 26-39 year olds watching our Ender’s Game on Blu-Ray Tournament (peak viewership ~16k) than watched the CLG Premier Series (peak viewership ~31k). For these two events in particular, the survey submission rate approximately reflects the peak viewership- CPS’s 1254 surveys compared to Ender’s Game’s 563 is a little more than double, with peak viewership obviously being a little less than double.
The youngest age bracket seems possibly like an anomaly, and with a difference of less than 0.5% between the two games, it’s hard to read in to. On the other hand, it’s fairly easy to understand why SC2 has more older viewers (by the way, it’s a bit odd calling 26 year olds “older viewers”). SC2 has a storied history from the Brood War days, and many SC2 fans are old enough, and invested enough into their game, to remember a time when saviOr wasn’t a synonym for Voldemort. If League lasts as long as SCBW did, there’s no doubt that League fans will have similar numbers of older fans in the coming years.
However, in the meantime, SC2’s fanbase has a significant margin in a few areas, including Education.
+ Show Spoiler +
![[image loading]](http://puu.sh/709RJ/5536f666b9.png)
![[image loading]](http://puu.sh/709S4/35cacec3bb.png)
Yes, the plurality in all three surveys is Some college, and it’s close enough that I have no issue calling it equal. However, the 13-18 age bracket that League has a 10% margin over SC2 pops up again, with 5% and 10% more High School students. SC2 pretty obviously has a higher rate of Bachelor’s and Master’s degrees due to the higher concentration of older fans. SC2 also appears to have a higher proportional rate of college graduates. Among viewers older than 19, there were 225 Ender’s Game Tournament viewers that had graduated college (out of 459 viewers older than 19), compared to the Spring Promotion Tournament’s 73 (out of 159) and the CLG Premier Series’ 369 (out of 890). These ratios amount to 49% for SC2, 45.9% for SPT, and 41.5% for CPS, amounting to a 3% and 7.5% margin in favor of SC2.
Now, these surveys aren’t as specific as I’d like them to be- something I’ll see about fixing for future events, but it appears that SC2 viewers are (generally) a 2-3 year older group as a whole, at least in terms of viewers around their 20s. Not quite as big a gap as I first anticipated, and although minor, the results are more pronounced in the questions we asked in these surveys. For one, there is no option for viewers in college to put their Year, and 19-26 is a significant stretch of age that includes both recent high school graduates (who are very, very unlikely to have a degree) and particularly motivated Master’s Degree holders (theoretically, at least). This problem would be minimized if we were a bit more specific with the age ranges, particularly if we split the 19-26 age group into two sub ranges, starting with 19-21 or 19-22, when it’s most likely for college students to still be enrolled in a traditional 4-year college.
At this point, I’m tired and don’t feel like typing much more. I’ll continue with the 5 remaining topics in another blog later this week.
