Alright got off work, here are the stats:
May: 49.53 million views (pre Old Gods/new Rotation/new Balance patch)
April: 41 million (Old Gods)
June: 35.4 million
July: 31.7 million
August: 34.5 million
September: 33.52 million (Karazhan so you would expect a significant bump like with Old Gods but it actually went down)
Ocober isn't completed so no stats on it yet.
Again it doesn't matter that HS is #2 or #4 relative to other games... it's about how it has performed since the rotation and whether it has improved or declined. And going off of these stats and what other tournament organizers/commentators have been saying, HS has been on the decline viewership wise especially in the E sports scene (and your other link proved it with the lower percentage compared to other games).
Hearthstone isn't dead and is still very health, I never stated anything to the contrary. In fact for the earlier part of the year it had gone up year to year but not so the case later in the year after summer. But Karazhan failed to generate major interest in the game, it should've been a bump up for the game but instead to pick up the slack Blizzard has had to put out 3 significant patches in a very short amount of time which is UNPRECEDENTED since the game's formal release.
I think it does matter how it performs compared to other games because we're looking at twitch as the market and how much of that market does hearthstone own; it also the best information we have. Each month the amount of events and hours of broadcast is different, which can easily explain a downtrend. Each month what counts as an esports broadcast is also a factor. This makes the best indicator of growth or decline is it's overall standing compared to other games; it's simply the best information we have available.
Karazhan is not an expansion like OG is. It's an adventure. I wouldn't expect that big of a jump if any, but I would also expect the impact to be in August, not September when it's a month or longer after the first wing came out (where I would expect the hype to be highest is the first wing). OG came out in June (technically April) but June's viewership is lower than April and April's is lower than May. We know old gods caused the highest active users users via nirolak right? If even old gods failed to produce a spike in total hours viewed, then total hours viewed is simply not an accurate indication of the active user base.
And again, the % column doesn't mean the esports is declining, it means that the non-esports scene is simply bigger relatively speaking. If the game had 20 million less total viewing hours but remained with 5 million esport hours, it would have a 33% esport share but it wouldn't be more successful as an esport. That would just mean the non-esport scene has shrunk a ton. This leads me back to the larger point in that each month has differing amounts of events and hours broadcast. I haven't seen any
viewer hours per broadcast hour, which would be much more indicative of growth or decline.
-
Regarding your talk about the patches, if I am correct in understanding your logic, you are saying these patches show the developers think the game is dead or dying and therefore are rushing to put out patches. That is not a compelling argument for two reasons. First, the patches are not unprecedented. 1 was a balance patch we get at least once a year. In 2015 we got two balance patches, one for undertaker and one for patron warrior (maybe more, but those 2 I recall clearly). The next was a "content" patch, more quests basically, and a few adjustments to cards so insignificant people barely cared about them. In 2015 we got tavern brawl. I am assuming the third patch relates to arena? Okay, but so what? This was clearly in response to community demand which leads me to the second reason.
The second reason the argument is not compelling is because as previously alluded to, these patches may have and even likely are in response to motivations other than player count. I think your argument is just heavily diluted by speculation, which flies in the face of more obvious reasons; nor would it really show the larger trend. If hearthstone sees player count dipping one week and they respond and player count is restored, then there is no extra action needed thus you could conclude that player count is no longer dropping but is actually rising.
So perhaps the two biggest points I am making is that relative viewing hours performance is our best indicator of success we have available to us, and that total hour viewership is not correlated enough with monthly active users to indicate growth or decline in the player base.