I just took a quick look at the weekly numbers for the month of May so far. The first week of May of course was in April's official numbers. So After looking over all the numbers so far they come to a little of 60k. Now Taking last month's numbers of 80k then looking at the current number (and not knowing what next week's numbers will be)  it seems to me they are based on the prediction, at least for now, that the PS3 will sell less then last month.

 I want to know how we come to these numbers? Theres a 50/50 chance that the true totals will be above or below 80k. 

 If taking this site's predictions, Id say the Wii will sell 20k more this month then last, the 360 will sell 300k due to the Elite being released, and the DS will sell 500k easy. 

 Tell me.. how are these numbers really about to be trusted? 

Oh and if you havent sign the truce so we can have good discussions/debates instead of raging fanboy wars.

http://www.vgchartz.com/forum/thread.php?id=2397&start=0#end  

Well, without any big releases since March I only expect sales to go down. So if NPD data next month show a number of around 70k I wouldn't be too surprised.

Remember that ioi is still trying to collect more data to make his predictions accurate. If the stores he has data from sell unusually well one week, his numbers might be skewed. And we won't know how close he is to NPD until NPD releases their numbers. But last month his numbers were pretty good. Remember, wanting 100% accuracy on any site won't happen. Even Media Create and Famitsu have differing numbers for Japan's weekly data.

yea.. but arent the Famitsu numbers a prediction of what the Media Create numbers will be?  

Yep, SSJ, it's true that the sales numbers sites and companies rely on each other to a greater or lesser degree: Media Create knows its numbers are largely correct because it aligns closely with Famitsu, VGChartz knows its numbers are close to correct because they align with NPD, and so forth. However, each of these sites has its own counting methods, and while they may use the other sites as a fact-checker, it's highly unlikely that all of them are wrong when they reach similar conclusions. Possible, just highly unlikely.