I deny nothing, unlike some people here. I mean, do you actually read my posts, or do you just skim through them? Because it seems that you're more intent on mocking me than in accurately representing what I say, and basing your "rebuttals" on that.
Overall libraries are important. Strong support over the system's life is important, as rapid loss of support close to the system's replacement coming out is associated with more rapid later-life declines. In other words, a strong selection of games released across a system's primary life cycle at minimum is a major factor that increases the perceived value of a system.
But we're talking individual games here. Individual. Games. And while a single game can and often does move hardware, not all software causes appreciable hardware sales growth. Sure, a popular game might sell millions of copies lifetime, but how many of those millions actually bought the system for that game? How many wouldn't have bought the system had that game not existed?
Tell me, where do you see substantial hardware growth from Super Mario Party in this graph?
Now here's what a real system-seller looks like:
Monster Hunter Rise, the most recent major system-seller on the Switch, was part of a series that consistently causes pronounced growth the week of whenever a new entry releases (barring a few counter-examples). It sold 1.3M copies at retail in Japan (possibly upwards of 2M or more if we include digital) and caused Switch sales to triple from what they were the week before. In absolute terms, the Switch sold over 178k more units than it did the previous week. Now, MH games are the kind of games that burn very bright but sputter out quickly, the exact opposite of a game like Mario Party. It dropped to 277k copies in its second week and was under 100k/week by its fifth week (with a one-time bump to over 100k during Golden Week). After just one month, it was selling at a low enough level where nobody could seriously argue that it was a continuing driver of Switch sales. Three months after its release it had dropped to under 10k copies per week. While it did have a residual effect in its second week, perhaps aided by the limited edition console released alongside the game, by Rise's third week the Switch was back down to around where it was the week before Rise released, and, aside from the Golden Week bump, the Switch has been on a downward trajectory since, even before the OLED was announced. MH Rise was a system-seller, but also a clear demonstration that the ability of a game to sell systems is not a long-term factor.
If I had time, I could drop many more examples (this post has already taken me over an hour to compose if you include creating charts), but you get the point. While there are a fair number of games that cause obvious (and always relatively short-term) growth in hardware sale, the vast majority of games have minimal to no ability to, by themselves, cause sales growth, and for most of them it's unlikely that their absence alone would have resulted in measurably worse sales for the system.
As for why some systems peak earlier or later than others, that has to do with a lot of other factors besides the games on the slate. The 360 & PS3 peaked after their slimline models were released. Those models caused immediate and pronounced spikes in hardware sales (the 360S more so in the U.S. than elsewhere) and a sustained increase in baseline sales (the PS3 Slim was also concurrent with a price cut from $400 to $300). Neither system had lineups that year that were mind-blowing relative to previous years. The PS2 is another example of a non-software factor leading to sustained growth, in its case its first price cut in May 2002. Speaking of price, the PS3 & 360 both took longer than earlier systems to get down to the prices they were at when they peaked, and at any given point in their lives were more expensive than the PS2. Nobody can look at any specific game on the PS2, even any assortment of games released in a relatively narrow time frame (say, a few months) and say "That's the cause." No, the PS2's sales grew because a lot of people that were holding back decided to take the plunge once Sony slashed its price by a third.
Individual games can and often do drive sales growth... in the short term. Most do not, however.
Price cuts frequently drive sales growth, and is one of two factors proven to produce such growth in the long term (i.e., many months).
Hardware revisions frequently drive sales growth, and is the other factor proven to be a long-term driver.
Limited edition systems often cause short-term sales growth, and have probably assisted most games they're associated with (when they are released alongside a game).
The overall long-term health of a system is determined by the interplay of overall games library, initial pricing, marketing, regional differences in spending habits, the occasional external factor, and a host of other factors that influence how potential customers perceive it as a value proposition.
Saying that Super Mario Party doesn't cause any obvious sales growth does not negate any of that.
And with that, I'm done with this conversation. I've said all I need to say and all I care to say, and I have other stuff to do.
"I mean, do you actually read my posts, or do you just skim through them?"
No, I read your posts, but I often don't bother responding to them, because they lack focus. You usually post a giant wall of text and charts and it can be hard to discern what your point is. I think I can address your point about Mario Party here though, so I'll try to keep this post just to that.
The issue I have with how you analyze Mario Party is that you only look at first week sales to determine if it's a "system seller". However, Mario Party sales are mostly from legs, while a game like Monster Hunter Rise is extremely front loaded. VGChartz has first week sales of Mario Party (worldwide) at 656k, while lifetime shipments are 15.72m. First week sales are about 4% of lifetime shipments so far. Meanwhile MH Rise sold over 4m the first week and its lifetime shipments are over 7.5m. So, it's first week sales are around 53% of lifetime shipments. So, first week sales are only 4% of lifetime sales (so far) for Super Mario Party while they are 53% for Monster Hunter Rise.
You only seem to be counting data from the first few weeks of sales and concluding that is the only way that a game can contribute to lifetime hardware sales. This is invalid reasoning. In fact, you did observe that hardware sales increased during the week of Mario Party's launch. "We saw an increase of only 6.5%, or 2805 units, from the previous week, indistinguishable from the normal week-to-week fluctuations." While that is true, you concluded that Mario Party did not cause an increase in Switch sales. That is the part that is invalid reasoning. Valid reasoning would be to say "we cannot conclude one way or the other if Mario Party caused an increase in Switch sales". After all, weekly fluctuations can go both ways. The could overrepresent the impact of Mario Party or they could just as likely underrepresent it.
However, there is another point that myself and others are concluding about games like Mario Party. All of us observe that large frontloaded games, like Monster Hunter Rise, cause an increase in hardware sales. When I observe this, as well as others who believe that Mario Party is a system seller, we think like this.
1. Big selling front loaded games show us that games which sell a lot of units cause an increase in hardware sales (as long as they launch early enough in a system's life).
2. Mario Party is a game that has sold a lot of units.
Conclusion: Mario Party has caused an increase in hardware sales.
You observe the same phenomenon and think like this.
1. Big selling front loaded games show us that games which sell an enormous amount of units in one week cause an increase in hardware sales, but only during the weeks which have an enormously large amount of sales.
2. Mario Party is not a front loaded game, and its sales are spaced out over time.
Conclusion: Mario Party did not cause an increase in hardware sales.
It's your #1 assumption in your thinking that I have the issue with. Scientific reasoning is to look at data when a phenomenon is easy to observe, such as when noise and random variance are small compared to the phenomenon in question, and then use that to make conclusions when the noise and variance make the phenomenon harder to observe. I.e. we look at front loaded sales and conclude that the software is still selling hardware even during later weeks when weekly sales aren't as high or during first week sales for a game that isn't front loaded. Poor reasoning is to say the phenomenon doesn't exist when noise and variance make it difficult to observe. I.e. Just because Mario Party isn't putting up huge software numbers on any given week, it can still be a system seller because it has large lifetime sales.