Like I said in the Thread this guy started specifically for this.....

Where do these people come from?

 Actually, I didn't see it until you mention'd about it... xP

Egads, beta!?!!  Sorry ioi, that million dollar check I sent for your "data" is gonna have to be cancelled...

=)

NPD and Media Create collect a lot of data for their "estimates". VGC just asks some shops how the situation is and estimate the sales. So if there are deviations of 70'000 solds per month....what deviations can happen during a console-life?!?

Xbox 360 still has not yet reached 10 million? Is it still behind the Sega Dreamcast? I mean, I'm sure the Xbox 360 has been out in more time than the Sega Dreamcast has, yet is still has not surpassed it? Are the Xbox 360, and Playstation 3, really that bad? I mean, the first digit, "8," in the number of Wii's sold, is only one digit lesser than the first digit, "9," in the number of Xbox 360's sold. The gap is slowly pinching together. I can't wait until the Wii beats the Xbox 360, because currently, never has a Nintendo console, surpassed a Microsoft console, in terms of sales.

Why would you expect a free website to be 100% accurate?  The front page numbers are never more than 5% off.  70,000/month sounds like alot until you realize that over the course of a 7 year cycle, it is 2.5 million units.  Is that a lot of units? Yes.  Is it alot out of the total market (125 mill Ps2, 24 mill Xbox, 22 mill GC, 9 mill DC) last generation - no.  As a total of last generation 70,000/month off for 7 years would account for total sales less than 2% of 180 million

1. The 70'000 were just for US.

2. No I don't 100 % accurate from a free website, but if they don't even have 80 % they can close the page, because they are wasting their and our time.

3. Sale numbers have to be nearly exact, if not I can make my own ;)

Well, by all means, if you can do a better job, what's stoppin ya? And, as it has been stated time and time and time and time and time and time again, not only are these estimates (which NPD also does), this site is in Beta. So, the site owner (ioi) is continually trying to expand his network to get his hands on more sales numbers so that he can increase his accuracy.

Frankly if you want more accurate numbers, hit the paypal button - the retail coverage will expand faster that way.  In reality, the 70k error was an extrapolation error.  That means it was more like an average of 17.5k/week off, which is less than 25% of 360 worldwide sales on a weekly basis.  In most weeks, I would imagine it is more like 15-20% off worldwide 360 sales.

Fair point about the 70k being just for the USA though.  But the USA is 60% of Xbox sales.  That would mean the numbers would still be off by less than 3%.  With 360 doing consistenly low numbers everywhere but North America, the UK/Ireland, and Australia, I wouldn't expect the margin of error to be to significant, especially since there are no reliable figures for Europe as a whole besides this website.

If you read the methodology section, ioi tries to incorporate additional sales besides USA & Canada for NPD as well - namely Central & South America, which is worth upto 5%. 

Anyway, I think that any example of good analysis will necissitate making adjustments, as new data comes in and as old data is looked through again to get a better picture of what is going on. What is the use of being consistent with your data ( NOT CHANGING) when you end up being wrong for the long term?