|
Ever since the English-speaking StarCraft 2 scene stabilized around July of last year, I been collecting whatever tournament viewership data available to the public. I was curious - can we make a reasonable model to predicting concurrent peaks? The theory was simple: with the fan base of static size Y, only Y/X number of fans would watch a tournament at any given time, with X varies based on (1) general type of the tournament and (2) the number of execution errors incurred.
Roughly speaking, the Anglosphere fan base can maintain a maximum of approximately 120,000 concurrent stream viewership, approximately half North American and half European. Therefore, amongst the Majors there're two theoretical ceilings
- 120,000 concurrent - an International Major that surpass the NASL line
- 66,000 concurrent - an Major with primarily NA- or EU-specific appeal, falling short of the NASL line
Below these, I also identified other ceilings based on format, scale, degree of live-ness, and star power involved (33,000; 16,000; 9,000; 4,800; 2,300; 1,000; 400; 150; 50; 18). Once an ceiling is identified based on macroscopic evaluation of the tournament in question, we simply make an educated guess over major and minor errors that may be incurred based on organizational track record and current news, and estimate percentage reduction over the particular concurrent ceiling from these errors.
For several months, I went on to make regular projections based on this model, sometimes sharing them with those that chat on Skype regularly. The model, however naive or simplistic, has been genuinely reliable. In fact, I was confident enough about my research that I recently begun to draft a formal algorithm to post on TeamLiquid. Then the ONOG Invitational came and went, ruining everything.
Deric Ortiz and Hannah B. Muffins are close friends of mine, fellow BarCraft organizers from the founding days of the BarCraft movement. The BarCraft-funded tournament concept were a distant dream we shared since the beginning, on what the movement can become. Deric and Muffins went on to find the One Nation of Gamers with the purpose of making the dream a reality. Five months of relentless hard work finally culminated in the form of $3,000 ONOG Invitational, paid for in its entirety by the barcrafts under the ONOG flag.
So I threw some numbers at Deric:
[1/26/2012 7:08:59 AM] XiFan "Primadog" Hong:
I hate early projections, but hitting 7.6k concurrent shouldn't be out of question if the stream went without a major fault.
Personally, I thought that number was a tad high. After all, an online invite with suboptimal starpower has a ceiling of 12,000. Considering that this is ONOG's inargural production, major errors will surely occur. 7,600 is generous.
80 hours later, the ONOG invitational hit its peak at 15k (appx) concurrent viewers.
Hype matters.
If you're an event organizer, that message should send shivers down your spine. Hype has to be genuine, because the fans can smell bullshit, especially fickle ones like our scene. Certainly, there are some zero-day hype tricks, but these tricks have high depreciation and real risk of being disingenuous. Broadly speaking, hype is out of your control. You have to count on the community to pull through for you.
Let's be clear, a lot of things went wrong during the ONOG Invitational, but those errors were largely irrelevant in context of the hype around the event. The Community are excited over a novel concept, the BarCraft movement can do more than get nerdballers drunk, that ONOG represented. Next thing you know, the ONOG Invitational had an air about it akin to any Major, featured on every StarCraft site and even on Battle.net!
While proud of what Deric and Muffins accomplished (and what's to come, now that the idea is out there), I can't help but feel a little lost. Months of hard work, suddenly invalidated in a weekend. Such is the tenet of Science, a model lasts only until it become invalidated.
I am a genuine '10er on TeamLiquid, a true freeborn of the new era. The fanaticism over Brood War in that disty corner of TeamLiquid is something I genuinely do not understand, only fearfully respect. For us SC2 fans, the stories of Brood War are stuff of myth and legends. We would never truly understand what losing MSL means, even when the sadness permeates and universal.
One thing I did learned while lurking in the BW forums, thought, is that they genuinely cared a lot about their game, even when KESPA doesn't care about the Foreigners. The few times Korean BW attempted to reach out ended largely in failure. Somehow, despite lacking spoon feeding from KESPA or OSL or MSL (RIP), the BW fans cares. Even when most of them don't even know enough Korean to understand the commentators, they still watched with fervor. That's loyalty.
We StarCraft 2 fans can learn from our older cousins. For one, we are really bad at getting ourselves hyped up. In fact, we have became entitled, blaming tournaments for not giving us interesting matchups. If only we can peer over the fence towards the BW forums. The tournaments' job is to make their product - the games - as good as possible. Hyping is fans' job. As engaged fans, getting excited should be our calling. The games and matchups are boring because we don't care enough to make them to make them interesting.
Think back to your fondest SC2 memory in 2011. Try to remember why it made your stomach butterfly and keep you at the edge of your seat. It probably wasn't anything the tournament organizers did. The hype came from within.
Towards Boxer, all are reverence. He walks as a living god among all of eSports. Having the pleasure to met briefly with him in the Ontario Convention Center last July, I am genuinely happy to able to watch him play regularly again.
- The speech was modeled after one delivered by General MacArthur on his return to Philippines.
- The image base is sourced from this BW hype thread; a special thanks to my good friend Alystair for fancy editing.
- The boxer signature is sourced from this blog with alpha channel applied.
- A special thanks to Steveling for recommending this mood music.
 
|
iNcontroL
USA29048 Posts
great blog, good read.. as per usual with you PrimaDog <3
|
i applaud you for scrapping it based on a single data point, that takes rigor. That said i have to wonder if your model was accurately predicting results in the first place, or if you had a general sense of where tournaments end up in viewership and were unintentionally feeding numbers in with an expected result anchoring them. Starpower for example seems like a pretty vague factor. Did you have a third party evaluating numbers like that?
|
I share my projections with friends to ensure rigor. Calling concurrent peaks several days prior, then wait for a friend to call me out on it, is how I figured we can ensure scientific integrity.
|
On February 12 2012 05:40 Primadog wrote:
One thing I did learned while lurking in the BW forums, thought, is that they genuinely cared a lot about their game, even when KESPA doesn't care about the Foreigners. The few times Korean BW attempted to reach out ended largely in failure. Somehow, despite lacking spoon feeding from KESPA or OSL or MSL (RIP), the BW fans cares. Even when most of them don't even know enough Korean to understand the commentators, they still watched with fervor. That's loyalty.
It's not that KeSPA is antagonistic towards foreigners, it's that without the type of social environment (i.e., RTS gaming culture) that exists in Korea, no foreigners can develop the skills needed to play BW at the Korean professional level. Back in the really old school days, before the scene became the modern ultra-competitive one it is today, there were foreigners (e.g., Grrrr..., Elky, Legionnaire), but the difficulty of leaving the rest of your life behind to move to Korea and play Broodwar all day with no guarantee of sucess is a major deterrent for foreigners--but it's not some barrier imposed by KeSPA or some other force. A very small number of dedicated and highly talented foreigners (Idra, Nony and Ret[sort of]), came kind of close to 'making it' in the modern era, but the life of a BW B-teamer is grueling hard work for anyone (and does not guarantee success), foreigner or not.
As for trying to reach out to (capitalize on) the Chinese fan base, the Korean Air OSL finals held in China was a success, while the 10-11 ProLeague finals that was to be held in Shanghai was canceled by the local government due to weather concerns (which were arguably unwarranted, as the weather didn't turn out to be that bad).
|
I don't mean to imply you were falsifying results or ignoring negative ones, which is what you describe guards against, I'm just wondering if it's possible that given the existence of subjective factors, you were naturally rating these factors to fall in line with a common sense perspective of what the results would be - making the model a formalization of your judgement as a predictor, rather than something that would produce the same result given any operator with the same set of data.
I haven't actually seen your model, this is just a potential pitfall I observe in the process as described above. Please don't take it the wrong way if it's not a fair criticism given information outside of this thread, because I don't have any.
|
Oh, I am completely aware that the "major and minor" error portion of the model is largely subjective. It wasn't intended to be an judgement free model. Instead, I simply want to know if projecting concurrent peaks are even possible. Before this effort, the event organizers I asked largely utilize the "kneel and pray" technique to predict numbers.
To be sure, I haven't gave up on the premise. Right now, i am trying to come up with a formulation that properly captures how hype move the dial as an intermittent step between (1) and (2). This topic was largely an attempt to salvage something interesting out of the retooling period.
|
I had no idea MLG Orlando did that well... I went to that one and met everyone. It was a really good event.
I like the second half too. Pictures of BW progamers make me like blogs in general.
|
very interesting post loved it.
Do we know why Orlando was so much higher than other MLG events? Including Providence? 20k+ more viewers seems strange but maybe it has to do with date or timing of stuff or just hype in general.
|
On February 12 2012 08:58 LuckyFool wrote:
very interesting post loved it.
Do we know why Orlando was so much higher than other MLG events? Including Providence? 20k+ more viewers seems strange but maybe it has to do with date or timing of stuff or just hype in general.
I have a couple ideas
1. Huk is a cutie
2. sup son
3. Boxer in round of 4
4. Idra vs Boxer
5. Idra vs MC
|
On February 12 2012 08:58 LuckyFool wrote:
very interesting post loved it.
Do we know why Orlando was so much higher than other MLG events? Including Providence? 20k+ more viewers seems strange but maybe it has to do with date or timing of stuff or just hype in general.
About 150 barcrafts met during the Providence weekend, compared less than half of that during Orlando. Each barcraft has anywhere from 40 to 700 attendees. My suspicion was watching BarCraft is largely responsible for the artificially depressed concurrent numbers. This is an area of concern for us and a potential area for antagonism between BarCraft and tournaments. Hope Sundance won't read this.
|
On February 12 2012 09:12 Primadog wrote:Show nested quote +On February 12 2012 08:58 LuckyFool wrote:
very interesting post loved it.
Do we know why Orlando was so much higher than other MLG events? Including Providence? 20k+ more viewers seems strange but maybe it has to do with date or timing of stuff or just hype in general. About 150 barcrafts met during the Providence weekend, compared less than half of that during Orlando. Each barcraft has anywhere from 40 to 700 attendees. My suspicion was watching BarCraft is largely responsible for the artificially depressed concurrent numbers. This is an area of concern for us and a potential area for antagonism between BarCraft and tournaments. Hope Sundance won't read this.
Barcrafts are guaranteed paid subscriptions and generate huge hype and more numbers in and of themselves. I think they help more than they hurt.
|
Out of all of those BW names listed, sAviOr is the one with the coolest name... MA JAE YOON!
|
On February 12 2012 09:12 Primadog wrote:Show nested quote +On February 12 2012 08:58 LuckyFool wrote:
very interesting post loved it.
Do we know why Orlando was so much higher than other MLG events? Including Providence? 20k+ more viewers seems strange but maybe it has to do with date or timing of stuff or just hype in general. About 150 barcrafts met during the Providence weekend, compared less than half of that during Orlando. Each barcraft has anywhere from 40 to 700 attendees. My suspicion was watching BarCraft is largely responsible for the artificially depressed concurrent numbers. This is an area of concern for us and a potential area for antagonism between BarCraft and tournaments. Hope Sundance won't read this.
I think his numbers for Providence are off by a bit, his estimate is 95k, MLG announced a peak viewership of 241k. I think StarcraftII accounted for more than 39% of that number. I would guess that the actual figure is significantly higher
MLG Season Numbers Recap Press Release
|
|
|
United States16963 Posts
thread renamed by request.
|
|
|
|
|
|
TeamLiquid StarLeague 5 in 
EPT![[image loading]](http://i.imgur.com/qwNjY.png)
![[image loading]](http://i.imgur.com/SWclL.png)
![[image loading]](http://i.imgur.com/XPbBC.jpg)
