Is Q4 2022 still part of "covid"? Why is there no holiday sales bump for the Series X, but there's one for the Series S? Like, Q3 to Q4 2022 you're talking about a 3X bump in "sales" for the Series S and a very negligible bump (within almost margin of error) for the Series X.
US =/= The World
1 to 2M, yet your first estimate was at 3.7M and the second one at 2.5M (while having a stock shortage!). Don't you see the contradiction here?
Edit: What it seems to me is that you're starting from a number you think makes sense, say 20M, then grabbing the revenue values for hardware (which you don't know if it includes peripherals, as we've established), and then you're extrapolating a sales number based on... Luck?
For example, if the 18.5M sell-through is correct and there's a console shortage, you wouldn't have 2.5M on retail channels, you'd have maybe 1M or so.
Series X availability dried up in December. This isn't some secret knowledge.
Also, Microsoft said they are doubling cloud server capacity over the next 12 months from August 2022, and cloud servers use X chips. That was part of why Series X availability went down in 2021, when in June they announced they were upgrade their data centers to the X and in October Microsoft said they had finished upgrading their cloud servers to Series X
At Xbox, our mission is simple: bring the joy and community of gaming to everyone on the planet. To achieve that, we aspire to empower everyone to play the games you want, with the people you want, anywhere you want. We are creating a future that combines the gaming heritage of Xbox and the...
www.google.com
The Xbox hardware upgrade improves load times and performance.
www.theverge.com
A recap in the cloud.
www.windowscentral.com
Series S is not in short supply, the X is. Remember that Microsoft put the Series S down to $249 in the US for 7 weeks in the holiday season. XB1 sold 2.5M the first holiday they did a $50 cut, and 3M when they went to $299, and 2.5M again when it went down to $249. If Microsoft anticipated such demand for a $249 S (why wouldn't they, no way you can reasonably say they cut the price to only maintain sales from last holiday, when the whole of 2022 before was up from 2021), they'd need to send a similar amount of units to be capable of being sold. XBS as whole did at best 2M in the US, down a lot by what should've occured going by historical data.
I don't start at a random number for hardware and try to get it through revenue. I go quarter by quarter trying to work it all out, using any type of language from Microsoft's reports to help figure it all out. I literally showed with evidence that it's possible to get these numbers from Q3 2014 financial results, which had a known total Xbox figure and calculated hardware revenue.
Using that same methodology, I then take the revenue for the quarter, let's say it's $500M.
I then take the price of the consoles, $500 and $300, and see what combination of units between the two can lead to $500M.
So for that example, you can have 1M X and 0 S, or 600K and <700K S, 400K X and 1M S, etc. What I then do is try using any available data, such as availability or just intuition like I do with NPD, and pick one that looks close to right. If in this example the X and S were about equal in availability, I could go with the second option, and if the X were hard to find while the S is plentiful, I'd be more inclined to go with the third option.
That's why my initial showing of my splits in the NPD thread were what they were. Series X and S availability both went down significantly after 2020 but I went with the outcome where S would start improving sooner than it probably did.
But something that can't be ignored is that the S just objectively did see massive supply improvements after June 2021. There is a spike there in availability, it can't be argued.
Also, like I've mentioned, my methodology is taking what worked when Microsoft gave more data. Even if you don't agree with the revenue aspect, it can lead to mathing out revenue and units in quarters where Microsoft provided unit sales. My methodology also worked out that XBS would be ahead of XB1 starting in Q2 2021, and that it was still ahead as of Q4 2021, and all the way up to Q3 2022. The only iffy bit is Q4 2022 because it's unknown if XBS is actually still ahead, and going by what I provided earlier, I have XB1 at 21.7M and XBS at 22M. That's very close and only slight changes to XB1 holiday quarters 2014 and 2015 would need to be made to put XB1 over XBS if indeed XBS has fallen behind.