By using this site, you agree to our Privacy Policy and our Terms of Use. Close
Kristof81 on 10 March 2018
Edit Thread
StarDoor said:
Kristof81 said:

Any calendar alignments in my figures start from the beginning of first full fiscal year, aka 1st of April (I should've added some asterix to the chart), or the closest date to it, as we're dealing with weekly figures. So it cancels all launch figures up to nearest April, which in case of Switch and 3DS would be 1 month worth of sales and in case of, let's say, PS4 four-ish months. As it's calendar alignment, all of those systems have had equal amount of holiday season sales accounted for. The calculation method is simply the sum of software sales from their first April to the date shown in the attached table. 

Yes, I know that. It does not counter anything I said.

Let's imagine that Switch had the exact same hardware sales as PS4, calendar aligned as of week 40.

This would mean PS4 and Switch both sold 12 million units between April and the first week of January. With identical hardware sales, you would think that software sales would be comparable too, right?

No.

Thanks to the differences in launch timing, PS4 sold 7 million units prior to the beginning of its first complete fiscal year, while Switch only sold 2.3 million. At the end of this period that you're comparing, this would put PS4's lifetime sales at 19.0 million, while Switch would be at 14.3 million.

In this scenario, PS4 will always have an enormous advantage, regardless of the current time of year, because 19.0 million console owners will obviously be buying more software than 14.3 million console owners. Switch could have the exact same hardware sales as PS4 for its entire life and it would still be behind in software thanks to the extra 4.7 million units PS4 would have because of its earlier launch. That's why a calendar-aligned comparison for software is nonsense.

I guess I misunderstood your question. 

Yes, you're absolutely right. The software figure will be skewed by the hardware difference. It's very clear during first 12-24 months, but this difference will slowly diminish with time. Is it going to be accurate? Of course not, but it should show an overall image of software sales in whatever calendar period. Perhaps I've rushed it and should've waited until around year 2 or so, or ...

To make it somewhat accurate, it'd have to go with my first idea and instead of reporting sums, report estimates based on actual, unaligned tie ratios, up to target date. It would still be an estimate, but far more accurate, especially in early weeks. That's something that I could do in the future, but because it'd require 180 change to my existing models, it'd be quite time consuming. That's why I didn't do it in the first place.  

But hey, that's for the feedback. It's always appreciated. If there's room to improve, it's going to be improved.