Your graph itself does not support the assumption that the drop is due to NSO, as NSO requirement is known way before before the actual implementation and would have an earlier effect. By your own admission that new content did not noticeably drive sales at all.
.
I fail to see the connection between the fact that a single player DLC did not drive the sale, and the fact that NSO's introduction perfectly matches a significant drop in the sales of the game
What you are seeing is a typical summer spike because kids get the summer off and purchasing starts, then a fall when kids go back to school. There is little way of separating the "end summer" effect with the "NSO effect".
This does not correspond to the data.
Summer holidays typically finish around the 20th of August. Most purchases after that are school related. You can see a significant drop at the end of Obon in the graph. But the drop due to NSO occurs more than a month later.
What is heavily ignored in this analysis is that the game would face natural barriers to increasing its user base since it is approaching 50% attach rate with total existing machines, one year on. Basically, by this time, anyone who wants Splatoon would have brought Splatoon.
And the market would saturate abruptly specifically on the week NSO is released? Seems far fetched to me.
You will not be able to separate the "lack of new updates going forward" effect from "natural decline as potential market dries up". Not every game can be GTAV where every new user picks it up, in fact, most games do not behave like GTAV at all.
This remains to be seen. If the baseline after the holidays is below 5000 per week for example, without any significant increase during holidays, it'll be difficult to find a more consistent explanation than the fact that people are less interested in purchasing a game as service without the service. But this is pure speculation, Neither you nor I know what Splatoon's sales are gonna be after the holidays.
Not to mention if NSO is the major barrier, then a new Splatoon will certainly do worse than Splatoon 2.
Cartidge_games already answered on that point. Larger install base (we both make the assumption that the next Splatoon will absolutely work on the current Switch as well), larger number of subscribers at that time, incentive for people to pay for NSO...
Furthermore, if Nintendo wants to push the IP going forward, it will need to spend heavily on aggresive marketing going forward to reach people they have not yet reached (and they already reached 3 million Japanese by conservative estimates), and those marketing plans can take years to be ready.
No one said the game would release in the next 6 months. I said summer 2020. That's longer than the duration during which Splatoon 2 has been on the market.
Can it happen in 2020? Maybe, but the marketing plans will have to be ready by then, and if there is a new generation of Switch ready then. My own take is that 2021 sounds more plausible.
So overall, you're not that far from 2020 either.
2021 is a possibility if Nintendo is concerned with franchise fatigue (Splatoon 1 and 2 were really close in terms of release). But in this case, stopping the content updates this month is a mistake in my opinion.
Finally, I do not expect Nintendo to release any game not compatible with the current switch before at least 2022. It's a successful machine, with an ever growing install base, and they would alienate many people if a new hardware was to be released with exclusive games within 4 years.
This Splatoon thing is slowly becoming a matter of "someone is wrong on the internet". I'll absolutely bring back the subject next week or the ones after if I feel it to be relevant, but I'm gonna stop here for now. Thanks for the insights.