Well a few things first of all are that NPD is by no means absolutely correct. They too are using estimates based on the same statistical principles as VGC. Where they gain accuracy is in the use of larger data sets. But what any stats professor or student will tell you is that you don't need large sample sizes to have accurate data. A sample size of 50-100 would be sufficient to give accurate data about a population of tens of thousands. You can ask your friend to educate himself on estimation theory. You'll notice similarly that large political polling groups use sample size of about 3,000 to estimate for a population of 300 million people then I think you can begin to see where sample size is no problem.
You can also tell your friend that like every branch of statistics 100% precision is never gaurunteed and is not claimed by VGC in any way. Simply because a single console disagreed with NPD (which is simply another estimate anyways) does not mean that everything is wrong. What he is doing is called chery picking. He has found the data that fits his hypothesis and ignores all the data from the other months that don't support it.
Luckily devilstation recently compiled some numbers and you might find it interesting:
MARCH 08........................Wii...............PS3...............X360
VGC (-10% CAN)..............593,738.......318,254.......319,317
NPD...................................721,000........257,000........262,000
Sim Exchange.................. 593,000......301,000........296,000
Pachter............................700,000.......365,000........310,00
Other Months
JAN 08.............................Wii...............PS3...............X360
VGC (-10% CAN)..............286,946.......214,809.......236,357
NPD................................ 274,000.......269,000........230,000
Sim Exchange..................396,900.......209,300 ........288,600
Pachter..........................????????.......????????........????????
FEB 08.............................Wii...............PS3...............X360
VGC (-10% CAN)..............449,902.......299,606........295,569
NPD................................ 432,000.......280,800.........254,600
Sim Exchange.................. 400,700......269,900........249,200
Pachter...........................????????.......????????........????????
VGC VS NPD yearly Totals
Wii........1330,586.....Vs.....1427,000..... -96,414 .... (-3.50%)
PS3........832,669......Vs......806,800...... +25,869.... (+1.58%)
360........851,243......Vs......746,600..... +104,643.... (+6.55%)
You'll notice that despite one or two disagreeances for a specific console in a specific month the quarter comes out very close and NPD and VGC numbers actually agree. The big difference? VGC numbers are available weekly and are always out before NPD numbers.
There has been a lot of bitterness towards VGC especially from the pro-360 camp over the X-mas 2006 time frame when MS shipped a lot of hardware and VGC didn't represent it. The reason for this was VGC tracks point-of-sale or in other words sales to consumers, and MS had stuffed the supply chain to meet a stated Quarterly goal. So as a result of being at odds with the MS shipped numbers for several months (until the stuffed units eventually sold) many 360 supporters now bash the site, this is especially prevalent with the Neogaf forums which he unsurprisingly mentions specifically.
Either way, there are quite a few major gaming news and even national news outlets that use VGC and it is becoming more respected and recognized by the day. Perhaps in the past the data wasn't as wholly accurate as it could be but ioi and his team have strived over the past year and a half or so to make vast improvements to data tracking and estimation methods. Or in short, its getting better all the time, which is why VGC is more and more excepted.
Ultimately I posted this for you, as I really don't think your friend is going to hear anything of it.
 |
| To Each Man, Responsibility |
