Oh, so you do know the real sales numbers? Or how else can you tell that vgchartz has the Truth?
The problem with VGChartz numbers *spin-off
- Thread starter dkanter
- Start date
Oh, so you do know the real sales numbers? Or how else can you tell that vgchartz has the Truth?
From that same Nintendo conference...
Am I not reading the graphs right?
Are the plotted points the differences in sales between monthsLook at the 360 figures for August-September...
The graph is monthly numbers provided by NPD and just averaged out to weekly figures somehow. It looks stupid in a line graph because it attributes transitions to monthly periods rather than the real causes and effects such as the release of halo3 or price drops.
That Nintendo diagram is based on GfK+Charttrack numbers, IIRC. It's pretty hard to determine the accuracy of VGchartz's European sales numbers in retrospect though, as the site has a nasty habit of "correcting" their numbers after numbers by the usual tracking services (NPD, GfK, Charttrack, MC, etc. ) have been released. So I wouldn't necessarily conclude here that their numbers are actually any good for Europe.
As for for good and regularily publically released numbers for Europe and the rest of the PAL territories we have to rely on the quarterly numbers published by the manufacturers (even though each of the has a somewhat definition of the term sold or of the PAL territories).
I consider GfK & Charttrack sales numbers as solid information about European console hardware sales. Unfortunately, those are only rarely published publically.
ya well the actual NPD number for january was 294k, but it could have been a 5 week NPD month. The 5 week months just muddy up the waters a bit more.
True, but my observations from the fall have shown me that when there have been public numbers to compare, for the European numbers, they are pretty similar, so I argue with those who dismiss them as random numbers.
It´s possible that VGChartz has been far worse previously and some people may be emotionally invested from that time, VGChartz may have improved since then, I don´t know. As I said I didn´t pay much attention to their diagrams before I saw that Nintendo diagram, which also provided weekly numbers.
If you accept that the numbers probably are a few percent off from the main tracking firms, I think VGChartz offers a pretty neat service by giving console sales numbers each week.
I have actually seen a lot of trends there that have been later confirmed when numbers have been provided from elsewhere (e.g. France, Spain, Germany, Italy).
I have seen them correct their weekly numbers two or three times during the time I´ve followed their charts. I am not sure I want to hold that against them, because it probably makes the data more accurate in the long run.
Last edited by a moderator:
A few? How about 20%? Like they were off for the wii in December?
You can use the GAF prediction thread (aggregate) if you just want some numbers every week. They have been about as accurate as VGchartz over the course of last year.
Yeah the Nintendo december numbers were a bit off, but you can hardly claim that is the norm. Having a supply restricted item in December where shipments may come in bursts by both air and boat shipments which perhaps are unevenly distributed is probably the worst case when your sample size is fairly small.
Anyways I wished you paid more attention to the numbers, as VGChartz present numbers for NA and not only the US, the error for the Wii number is probably around 15 % and not 20 %.
Thanks, I didn´t know they kept European numbers there, I will check them out.
They don't do euro prediction, because there is no way to verify the numbers. For that same reason I don't find the vgchartz numbers helpful.
Johnny Awesome
Veteran
Well if you overestimate one console by 20% and underestimate another by 20% the differences can be HUGE. VGChartz is amusing, but innacurate.
Using them as trend charts is something that I am willing to accept (i.e. Xbox sales have picked), but given their usual margins of errors (even in the US! a far more homogenous market than Europe) makes taking them at face value (i.e. absolute numbers) for Europe rather difficult for me, at the very least.
As you don´t bother to read my posts, I will not argue with you.
@alphawolf, European numbers do pop up from time to time, such as this for France and they don´t seem to deviate so much that you can dismiss VGChartz for just having fantasy numbers for Europe. I agree with StefanS that their main strength is to provide trends, and for the diversified market of Europe that is a pretty neat feature.
Last edited by a moderator:
Just stumpled upon some Euro numbers for Scandinavia.
Apparently, after a slow start for the PS3, it took two price drops and some time (october) to have the largest marketshare of weekly console sales.
This would be combletely reverse of what Vgcharts reports for europe, but then again scandinavia is only about 24 million people.
On second read, the story seems to be from Sony HQ, so what they actually measure is unkown.
Apparently, after a slow start for the PS3, it took two price drops and some time (october) to have the largest marketshare of weekly console sales.
This would be combletely reverse of what Vgcharts reports for europe, but then again scandinavia is only about 24 million people.
On second read, the story seems to be from Sony HQ, so what they actually measure is unkown.
Last edited by a moderator:
Mod note:
Please use this thread, if you want to post sales numbers for European / Other region's sales numbers, unless you want to contrast / back up VGChartz claims. In that case posting sales numbers is alright.
Please use this thread, if you want to post sales numbers for European / Other region's sales numbers, unless you want to contrast / back up VGChartz claims. In that case posting sales numbers is alright.
What do you mean by "completely wrong?" I would think that would mean their error is on the order of around 75%. But in real life, their average deviation from NPD is 12%. That means that their first digit and their second +/- 2 are pretty much correct. That means the numbers aren't useful for minute comparisons, but they do give you a reasonable picture of what's going on. It's always a question of how accurate something is and what that level of accuracy allows you to do. I for one find them useful to see if a particular software title is doing moderately well, or if it's totally tanking. It's why I knew Zack & Wiki had bombed before any official numbers came out. They were reporting ~35K in week 3 IIRC.
Supposedly, it's up around 120K now. Word of mouth help this title?
VGChartz software numbers are totally hit or miss right now. In comparison to last month's NPD they shorted COD4 on 360 by 900k units! And the rest of their numbers were way off as well. The real killer is the difference isn't proportional. Some titles were off by a larger percentage than others.
This type of inaccuracy is going to continue to happen until they can manage to acquire more sources of hard sales data to improve their sample size. They are currently doing way too much estimating to be consistently reliable. Even their relative success in matching NPD's hardware sales of late could simply be a fluke. They are going to have to continue to be consistently close for a much longer period of time for me to consider them more than a curiosity.
I really hope they do get it together, though. I am excited about the concept of the site and it would be an invaluable research and analysis tool if it ever is fully realized. But they have a long way to go.
I believe vgchartz does have some USA sources perhaps at a store/regional level. They have stated this numerous times before. I believe 3% coverage is one number they've claimed, which is nothing compared to NPD.
Japan they are simply using couple-day-early sources (certain public blogs, etc)for Famitsu and Media create type data. Since Japanese sales are tracked weekly by public reputable firms anyway, it's fairly a non-issue I guess.
Europe is the kicker, the site creator ioi does live in London so it's not out of the realm he may have retail or gfk/charttrack sources, (and it may vary by country) but my current belief is he's essentially fabricating Euro weekly numbers based on past statistics. I believe this because whenever the occasional news account relating reputable sales for a particular country surface, he retrofits his own data. For example recently some french hardware sales came to light, his own numbers were off, so he changed them to more closely reflect the actual. The thing is that you will get legitimate sales reports on the major Euro countries every so often, so his numbers for any given country are likely no more than a few months out of date at any time and so the life to date for a given Euro country are likely to not be too wildly innacurate, it's the weekly that are highly suspect.
About my greatest current hope for vgchartz is that they become a reasonaby accurate NPD hardware predictor (and therefore their weekly USA sales have some value). That is why I like to throw their monthly hardware "estimates" out before NPD.
Oh another thing interesting they've recently done, is the XBLA sales charts. Their methodology is transparent on that one, so I dont think anybody should have problems. Who knows the accuracy but it is at least interesting and a commendable attempt that crystallizes the good things about the site imo.
Japan they are simply using couple-day-early sources (certain public blogs, etc)for Famitsu and Media create type data. Since Japanese sales are tracked weekly by public reputable firms anyway, it's fairly a non-issue I guess.
Europe is the kicker, the site creator ioi does live in London so it's not out of the realm he may have retail or gfk/charttrack sources, (and it may vary by country) but my current belief is he's essentially fabricating Euro weekly numbers based on past statistics. I believe this because whenever the occasional news account relating reputable sales for a particular country surface, he retrofits his own data. For example recently some french hardware sales came to light, his own numbers were off, so he changed them to more closely reflect the actual. The thing is that you will get legitimate sales reports on the major Euro countries every so often, so his numbers for any given country are likely no more than a few months out of date at any time and so the life to date for a given Euro country are likely to not be too wildly innacurate, it's the weekly that are highly suspect.
About my greatest current hope for vgchartz is that they become a reasonaby accurate NPD hardware predictor (and therefore their weekly USA sales have some value). That is why I like to throw their monthly hardware "estimates" out before NPD.
Oh another thing interesting they've recently done, is the XBLA sales charts. Their methodology is transparent on that one, so I dont think anybody should have problems. Who knows the accuracy but it is at least interesting and a commendable attempt that crystallizes the good things about the site imo.
