37 million ~ 85%.
(37 / 85)*100 ~ 43 million.
Seems to tally with the idea that NPD are out by 15%.
PS2 outselling XBox 360
- Thread starter SPM
- Start date
(37 / 85)*100 ~ 43 million.
Seems to tally with the idea that NPD are out by 15%.
scooby_dooby
Legend
I'm not quite sure what this is supposed to be arguing. That it's possible for 5 million units to be sold, but NPD only report 2.4 million? You think that the 25% not captured can be responsible for over half of the sales, when that 25% includes the small-fry stores? Or do you think when NPD estimate WalMart's figures, they're out by a factor of 1000% and WalMart can sell two and a half million with NPD guessing they only sell a couple hundred thousand? Or does that 15% include another massive chain or three with huge market presence, which unlike WalMart NPD decided not to estimate?
Just that is very fuzzy math. Not only are the US numbers not accurate, we don't have numbers for the rest of world either, so we're basically extrapolating from numbers that are inaccurate to begin with.
The 5million figure was for WW shipments btw.
scooby_dooby
Legend
That seems fair, but the point is that the NPD numbers can't really be taken as definitiv. Regardless, either way the sales are mediocre.
Actually, it would be more precise to go the other way... even if it gives roughly the same result 
So assume that 43 million is 100% US + 7% Canadian sales; thus US sales should be 40.18 million. NDP reports 34.3 million, which is 85.3% of that.
But we don't know how many consoles ar actually sold in Canada, or what part of the shipments is still travelling or sitting on a store shelf. Also, the 43 million shipped data is from here and you can see that it's date is 2006/06/30. So we can't derive an exact percentage, and we probably can't apply it to the sales of other consoles either.
Nevertheless, the 80-85% average seems to be a very close one.
So assume that 43 million is 100% US + 7% Canadian sales; thus US sales should be 40.18 million. NDP reports 34.3 million, which is 85.3% of that.But we don't know how many consoles ar actually sold in Canada, or what part of the shipments is still travelling or sitting on a store shelf. Also, the 43 million shipped data is from here and you can see that it's date is 2006/06/30. So we can't derive an exact percentage, and we probably can't apply it to the sales of other consoles either.
Nevertheless, the 80-85% average seems to be a very close one.
Shifty, this is exactly what I've been saying - people take NPD sales figures as verbatum, but they are not - they don't capture all sales (and bear in mind the % of what isn't captures is actually outlets, not sales) ergo the information cannot be used to give a total sales figure for the region, as there is still a factor on top of that that isn't, and can't be accounted for. This is why I say that NPD is very good for trending (be that at a static period trend - i.e. comparing relative sales of one console to another for the same month is likely to give you a good overall guide for that month - or seasonal trends), but the sum of the numbers given by NPD do not equate to total sales.
That was gross summation of the types of arguments that you get whenever sales figures from the manufacturer are quoted and someone then quotes the NPD figures for NA and says "how can that be so??"
I think the numbers would go like this...
US NDP is 2.4 million, corrected number is about 3 million. Canada around 0.2.
Europe was 1.3 the last time I've heard, that's already 4.5 million. Japan has a few hundred thousand too. Then there's Australia and NZ, and some stock in the shops... and we're already beyond the 5 million number from June.
So it should add up if our assumptions are correct.
Dave, is there a mechanism for people to know the exact POS sales numbers ? What do the console makers use ?
The NPD mechanism may not be perfect, but if everyone is using them as reference, including financial reports, they will be authoritative. For the X% of uncaptured sales, as Shifty pointed out, the business people will just have to compensate for them (assuming the error margin is relatively well-known).
So you are essentially correct to say that data such as NPD's is not accurate to the last unit of sales. For that matter, unless there are real-time ordering systems... there is always a risk in missing out large disti orders at any point in time.
It seems that for all intent and purposes, the NPD data still represent a key indicator of trends, marketshare and sales calculation/estimation.
The NPD mechanism may not be perfect, but if everyone is using them as reference, including financial reports, they will be authoritative. For the X% of uncaptured sales, as Shifty pointed out, the business people will just have to compensate for them (assuming the error margin is relatively well-known).
So you are essentially correct to say that data such as NPD's is not accurate to the last unit of sales. For that matter, unless there are real-time ordering systems... there is always a risk in missing out large disti orders at any point in time.
It seems that for all intent and purposes, the NPD data still represent a key indicator of trends, marketshare and sales calculation/estimation.
Last edited by a moderator:
Are NPD's numbers the total that they actually counted, or are they an estimate based on what they counted?
It doesn't matter, anyway. If PS2 is outselling Xbox 360 in the 75% or so of the market that NPD actually counts, it's highly, highly unlikely that X360 is just blasting the PS2 in that remaining segment.
It doesn't matter, anyway. If PS2 is outselling Xbox 360 in the 75% or so of the market that NPD actually counts, it's highly, highly unlikely that X360 is just blasting the PS2 in that remaining segment.
Last edited by a moderator:
I don't think anyone is suggesting that the 360 is possibly outselling the PS2 (and in fact, that comparison is valid), just that you can't get a specific number of units sold based on NPD numbers, such as, "OMG M$ said they'd sale 10 million and NPD says they only sold 2. LOL"
What the heck? Are you suggesting NPD-numbers (sold to customers) should show exactly the numbers shipped to retailers? Makes no sense.
Okay, with this and other responses in this thread, I think the argument's getting kinda fuzzy. My only point in all this was to try to identify what info can be got from the NPD figures, and not in any way to contribute to any current arguments that reference those figures!
I can conclude my argument by saying that NPD numbers aren't 100% accurate, but are a fairly good estimation. They cannot be cited as exact figures, but can be more useful than just for looking at trends. Anyone arguing over a dozen percent inaccuracy in the NPD numbers to (dis)prove a point should consider the position of the EU, where such a fragmented and unreported market makes worldwide install base (the figure that really matters) a total nightmare!
Yeah I agree with that except Japan figure is <200K defintely between Media Create and Enterbrain(Famitsu).
Well, you can (with a low margin of error), but you'd have to know other information first.
Case 1:
5 million consoles are shipped.
NPD counts 3.2 million consoles sold by looking at the numbers of retailers to whom a total of 4.5 million consoles were shipped (i.e. 500K consoles were unaccounted for by them), so extrapolating this 71% sales rate, you can say that with a small margin of error, 3.6 million consoles were sold by Day X.
Case 2:
5 million consoles are shipped.
NPD counts 2.9 million consoles sold out of the 4.5 million consoles shipped to the retailers they asked and extrapolated this themselves to 3.2 million consoles, in which case you can say that with a small margin of error, 3.2 million consoles were sold by Day X.
And statistically, that margin of error should be very small, like around 1%. You can't get the exact number, but you can get a very, very good guess. By randomly sampling 75% of a population, you can get a very, very good estimate of what the rest of the populations. Pollsters are able to get very good estimates of election results by sampling less than 10% of the total voting population. That's how statistics works.
Wiki's explanation of margin of error is pretty good: http://en.wikipedia.org/wiki/Margin_of_error
Just an example: Suppose that 70% of the people vote "yes." Now suppose you poll 1000 people. The probability that your poll results show that between 68% and 72% of the people vote "yes" is 85%. The probability that your poll lies between 65% and 75% is a whopping 99.95%! This is the driving reason behind why statistics actually works. You can do the computations yourself using this Maple code (sorry, I didn't get Matlab for free when I was at RHIT):
> N:=1000;
> sum(binomial(N,i)*.7^(i)*.3^(N-i),i=650..750);
Last edited by a moderator:
And if you'd read my posts thats exactly what I've said, except that they will not give an accurate total sales figure for the region - whatever total numbers NPD shows, more will have actually been sold.
Perhaps it should be phrased better then...
It is still useful for estimating overall sales once you compensate for the gap. I was under the impression that you meant people cannot use/trust NPD (or equivalent) data when analyzing or comparing sales numbers. There is a subtle but important difference.
Dave Baumann said:
It is still useful for estimating overall sales once you compensate for the gap. I was under the impression that you meant people cannot use/trust NPD (or equivalent) data when analyzing or comparing sales numbers. There is a subtle but important difference.
It reads fine.
Johnny Awesome
Veteran
And even more will have actually been shipped, which is what MS and Sony report anyway.
Well..its looking more like the NPD Data after these months that past show a sluggish sale through of the X360 console. Check it...
Analyst adjusts hardware sales down, software up
Lazard Capital Markets tells investors that Xbox 360 console sales are lagging--but the system's games are still moving, along with DS and PS2 titles.
By Brendan Sinclair, GameSpot
Posted Oct 24, 2006 12:17 pm PT
".... the hardware is slipping off the pace, the analyst said. While he cites "market expectations" as having 5 million Xbox 360s sold for the year, Sebastian said that the system's sales to date suggest Microsoft is more likely to wind up with 4.5 million units sold for the year...."
Analyst adjusts hardware sales down, software up
Lazard Capital Markets tells investors that Xbox 360 console sales are lagging--but the system's games are still moving, along with DS and PS2 titles.
By Brendan Sinclair, GameSpot
Posted Oct 24, 2006 12:17 pm PT
".... the hardware is slipping off the pace, the analyst said. While he cites "market expectations" as having 5 million Xbox 360s sold for the year, Sebastian said that the system's sales to date suggest Microsoft is more likely to wind up with 4.5 million units sold for the year...."
Those are US sales right? 'Coz I thought we were on 5 million back around June-ish!
Similar threads
