Like the title says, i heard people here say when NPD January came out that PS3 was overtracked. Then when NPD February came out, i think ioi said that since NPD undertracked the DS in January they were making up for it in February. So i would like someone to explain to me how can we say these things? What kind of information do we have to be able to say these things (im not disputing the information, im simply trying to understand)?

I just want people to know that i dont really know anything about sales (so what im asking could be really simple), and that im not in America, Europe or Japan (im from Brazil), so there might be some stores and companies that i simply wont know. So, to whomever answers thanks in advance.

can't tell it any better than soriku did.

Just like the ps2,

Vgchartz may say its at 100k for the month in north america, when it actually sold over 350k in NA. That is a huge undertrackage.

First time ever:

Soriku +1

 Thanks soriku, but what i also want to know is why did people say NPD overtracked PS3 and undertracked DS in January. I used to think that it was like what you said, under and overtracked means difference between VGchartz and NPD, but i heard some people say that NPD was wrong, so why do they think it is wrong if NPD has more information (market share )? Thats what i dont understand.

NPD makes estimates based off of some raw data (supposedly they get 50-60% of actual sales reported to them). So theoretically, the real sales could be 100k, yet they could report anywhere from 50-200k if someone really botched the calculations.

VGC also makes estimates based off of some raw data (supposedly 10%). They also could make mistakes, and have to work harder to make sure they don't. However, VGC's sample size is big enough that they could be making more accurate estimates.

When someone says NPD is over/under tracking it is because they believe VGC has provided more accurate information based off of their samples for that period. It has been the case a couple of times that VGC has been more accurate than NPD. This was shown by comparing VGC numbers versus manufacturer shipped numbers (note, shipped, not sold to consumers like NPD and VGC track).

 Now i iunderstand. Thanks a lot, nordlead. That does leave one question, though: Do the people that work for NPD not check this manufacturer numbers, to see potencial errors in thir tracking?

Manufacturer numbers are only released on occasion, typically just with their financial reports, which are once a quarter. These numbers don't show up till after NPD and VGC has reported their estimates. So corrections can only be made at a later date.

NPD does not release their corrections to the public, only to their paying customers. So we can't see if NPD corrected using hardware shipments, but I am sure they do. You can see when VGC gets updated, since the historical weekly sales are always accessible.

 Thanks again, nordlead. You explain very well. Maybe some mod could take what you said and make to a sticky because there might be other people with the same doubt. Thanks again.