The problem with VGChartz numbers *spin-off

[Crossbar]

Some Sony sales info here.

I would call December something of a trinity, because we sold just over a million PlayStation 3s during the month. That’s across all PAL territories and that’s a sell-through figure.

......

In terms of installed base in some of our major markets, such as France, Spain, Germany, Italy, plus some distributor territories in smaller markets like Switzerland, our current cumulative installed base is now higher than that of Xbox 360.

That is pretty much inline with VGChartz.

It’s worth mentioning that last week we outsold Xbox 360 in PAL territories by a ratio of 3:1.

That is not at all inline with VGChartz. VGChartz seems to be reporting very inflated 360 numbers or way too low PS3 numbers for the previous PAL week if David Reeves is right. Will be interesting if we will see such a correction for the coming week.

 

[Rangers]

Some Sony sales info here.

That is pretty much inline with VGChartz.


That is not at all inline with VGChartz. VGChartz seems to be reporting very inflated 360 numbers or way too low PS3 numbers for the previous PAL week if David Reeves is right. Will be interesting if we will see such a correction for the coming week.

They say Reeves is either referring to PS2+PS3 or PS2+PSP+PS3 sales to make the 3-1 statement.

 

[Crossbar]

They say Reeves is either referring to PS2+PS3 or PS2+PSP+PS3 sales to make the 3-1 statement.

Could be, but if that´s the case Reeves is pretty careless in his answer, see the full context below.

And how do you feel PS3 is doing in terms of ‘catching up’ with your rivals who launched earlier?

We don’t often mention competition but I will give you some interesting data. In terms of installed base in some of our major markets, such as France, Spain, Germany, Italy, plus some distributor territories in smaller markets like Switzerland, our current cumulative installed base is now higher than that of Xbox 360.

It has not quite yet reached that in the UK. But the team here expect to overtake installed base of Xbox 360 across PAL territories in late summer. I would say before, but that is when we think we will overtake Xbox 360.

It’s worth mentioning that last week we outsold Xbox 360 in PAL territories by a ratio of 3:1.

 

[2real4tv]

Europe/other numbers posted:
http://www.vgchartz.com/eweekly.php

Interesting...according to them Burnout on ps3 outsold 360, of course they could be way off...looks at COD4 numbers.

 

[fearsomepirate]

900K is huge...I just assumed their hardware accuracy and software accuracy were similar. They've been pretty decent with most of the software I've looked at. I guess COD4 is just way out there.

 

[fastasleep]

Vgchartz, by their own admission sample only 2-3% of retailers, or so they say. They will not say who they sample. If that isn't a big red warning flag, what is?

The guy who runs the site, as far as I can tell, has no formal training in statistical analysis or information gathering, he's an enthusiast at best and what he seems to be good at, is getting people to mention vgchartz.

Does anybody pay vgchartz for their data? I think, that should tell you everything you need to know.

 

[Rangers]

Could be, but if that´s the case Reeves is pretty careless in his answer, see the full context below.

True, but Sony made a similar misleading statement regarding Wii before, that is why people suspect them.

From what I know 3:1 seems possible to me. They seem about even in UK now and 3:1 seems believable in the rest of Europe.

So far, speaking of, VGchartz has 3 weeks of January, 360 at about 80k a week and PS3 50k a week. That point to a 300k 360, 200k PS3 month prediction, which sounds about right.

Since falling to 399 I've noticed PS3 has sold pretty consistently about 60% of 360 in NPD so far. That seems to be their new baseline.

 

[PARANOiA]

Vgchartz, by their own admission sample only 2-3% of retailers, or so they say. They will not say who they sample. If that isn't a big red warning flag, what is?

The guy who runs the site, as far as I can tell, has no formal training in statistical analysis or information gathering, he's an enthusiast at best and what he seems to be good at, is getting people to mention vgchartz.

Does anybody pay vgchartz for their data? I think, that should tell you everything you need to know.

You don't need a large sample size to get an accurate measure though. You just need to ensure that sample is representative. I don't think that's really the negative spin put on the site though...

I think the main problem with VGChartz and co are that the data is used to form or back-up arguments when there is no way to accurately validate the numbers provided. This goes double for for Europe, when all we hear are ambiguous snippets and whatnot. Using PR statements or, even worse, data collected by similar types of services that are professional and more reliable (NPD, GFK, etc) to modify your own predictions, makes the unverified numbers dubious at best.

As to how the numbers are used on B3D... how can you win or lose an argument when realistically there's no way to prove the reliability of the argument put forward? You're just as likely to win an argument as to whether or not there are aliens in the universe or whether or not gods exist.

The numbers are good however for having a historical reference for true and accurate numbers available historically - they are a single source for NPD's historically, which is better than you'll find most other places unless you have a lot of time on your hands.

 

[PARANOiA]

True, but Sony made a similar misleading statement regarding Wii before, that is why people suspect them.

From what I know 3:1 seems possible to me. They seem about even in UK now and 3:1 seems believable in the rest of Europe.

So far, speaking of, VGchartz has 3 weeks of January, 360 at about 80k a week and PS3 50k a week. That point to a 300k 360, 200k PS3 month prediction, which sounds about right.

Since falling to 399 I've noticed PS3 has sold pretty consistently about 60% of 360 in NPD so far. That seems to be their new baseline.

Also consider the 360 seems to be in shortages worldwide at the moment, with reports of minimal shipments since December. This makes the 3-1 number both accurate and misleading if that's what's driving it - but until GFK or whoever else services the UK comes forward, it's definitely best taken with a bit of salt, since there's no way to know for sure.

 

[fearsomepirate]

What would be "formal training in statistical analysis?" If he's got a BS, he's probably taken a statistics course or two. OTOH, his abuse of sig figs makes it all questionable. Of course, I've known engineers who didn't understand sig figs...

 

[fastasleep]

You don't need a large sample size to get an accurate measure though. You just need to ensure that sample is representative. I don't think that's really the negative spin put on the site though...

I think the main problem with VGChartz and co are that the data is used to form or back-up arguments when there is no way to accurately validate the numbers provided. This goes double for for Europe, when all we hear are ambiguous snippets and whatnot. Using PR statements or, even worse, data collected by similar types of services that are professional and more reliable (NPD, GFK, etc) to modify your own predictions, makes the unverified numbers dubious at best.

As to how the numbers are used on B3D... how can you win or lose an argument when realistically there's no way to prove the reliability of the argument put forward? You're just as likely to win an argument as to whether or not there are aliens in the universe or whether or not gods exist.

The numbers are good however for having a historical reference for true and accurate numbers available historically - they are a single source for NPD's historically, which is better than you'll find most other places unless you have a lot of time on your hands.

Anyway you cut it, their methodology is suspect, they don't want to reveal who exactly it is they actually poll, and until they do, their figures belong in the wastebasket. I can get better hardware predictions from the neogaf gang.

 

[Crossbar]

Anyway you cut it, their methodology is suspect, they don't want to reveal who exactly it is they actually poll, and until they do, their figures belong in the wastebasket. I can get better hardware predictions from the neogaf gang.

If you are not interested in weekly updates and European data, I guess neogaf is just fine. Not saying VGChartz is 100%, but pretty good to be free.

 

[mrcorbo]

If you are not interested in weekly updates and European data, I guess neogaf is just fine. Not saying VGChartz is 100%, but pretty good to be free.

I said this in Nov. of '07 about the problems with the use of VGChartz numbers in forum discussions. It still applies.

I am actually a supporter of VGChartz and what they are trying to do. I just don't think they have shown that they are reliably good at it. If you are willing to accept that they do have independent sources of sales data, this change shouldn't effect their accuracy in a negative way. And they have been providing weekly sales data for quite a while now which is in the ballpark enough to be convincing as more than just pulling numbers out of thin air.

The real problem is that NPD data was a rock solid and almost indisputable reference that people could use to support arguments, projections, etc. VGChartz data is nothing close to an indisputable reference, so any argument using that data as a foundation isn't going to get very far since it would be pretty easy for somebody to just call the numbers bogus turning the thread into a "Is VGChartz accurate?" discussion.

The above was in the context of NPD numbers no longer being made publicly available, prior to their reversal of that decision.

Any argument using VGChartz numbers alone as a foundation is doomed to turn into a circular argument with no resolution. Unfortunate, but true.

 

[fastasleep]

Another stellar month for vgchartz, january 2008, they had the 360 consistently outselling the ps3 for the entire month of january, only it didn't turn out that way. DOH!

What's even funnier, is that no sooner do the npd numbers hit, they completely rewrite weekly hardware sales numbers. This site is so bogus.

 

[bkilian]

Another possibility is that PS4 would be more expensive than 720. The PS3 was a lot more expensive than the 360 for it's entire life, and despite that the PS3 has been consistently outselling the 360 every year since it came out.

That's not entirely true, the 360 has been the #1 selling console worldwide for the last two years.

 

[MrFox]

That's not entirely true, the 360 has been the #1 selling console worldwide for the last two years.

Not according to vgchartz, what's your source?
It's clear that PS3 sold and average of 1 million more every year, I guess it depends if you count the year from Jan-Jan (vgcharts) or Mar-Mar (fiscal year). It also currently has a 1.3 million advance YTD for 2012.

 

[TheWretched]

That's not entirely true, the 360 has been the #1 selling console worldwide for the last two years.

If by global, you mean the US, then yeah... if by global, you mean worldwide, then no, at least not according to VGChartz.

http://www.vgchartz.com/yearly/2012/Global/
http://www.vgchartz.com/yearly/2011/Global/
http://www.vgchartz.com/yearly/2010/Global/
http://www.vgchartz.com/yearly/2009/Global/

http://www.vgchartz.com/yearly/2007/Global/

Neither year shows MS in front of the PS3. And only in 2011 and 2012 did it overtake the Wii in yearly totals. 2008 has 360 leading PS3... that is in fact the only full year the 360 has sold more than PS3 (2006 too, but that was PS3 launch year).

 

[MrFox]

There's a reason why some site don't use vgchartz or find them reliable.

What's the reason, exactly? If you have a better source please say it, I'll be happy to follow it. The numbers directly from MS and Sony from launch to date are matching very closely the vgcharts ones. It's the estimates that are often skewed and later need a correction.

My point is that they are outselling the 360 with a console that always had it's lowest SKU $100 or more above MS. Otherwise the numbers would now be 72M-64M (a tie in sales) instead of 72M-70M (sold 6M more in the last 6 years, so 1M per year on average).

(I apologize for my stupid off topic tangent)

 

[ERP]

My point is that they are outselling the 360 with a console that always had it's lowest SKU $100 or more above MS. Otherwise the numbers would now be 72M-64M (a tie in sales) instead of 72M-70M (sold 6M more in the last 6 years, so 1M per year on average).

This may be true, but the real money is in Software licenses and multiplatform releases (with a few exceptions) sell much better on 360 than PS3, in some cases almost 2:1.
Look at the COD numbers.

 

[AlphaWolf]

What's the reason, exactly?

Any number vgchartz has which is not directly sourced from MS, Sony, Nintendo, NPD or another reliable source may as well be voodoo. They are not working with a huge network of retailers or anything they take official data and guesstimate everything else. They tend to follow official numbers because they adjust their guesstimate off of official numbers.