Partially how?? I think ioi's sentence is 100% accurate whereas yours is 100% incorrect. VG Chartz doesn't have 95% confidence about values with 5% margin of error. They have 95% confidence about "X - 3X" intervals, which is a hell of a range (in my opinion) to use unrounded values placed into rankings. As for the combination of 2 major errors to be 1 in a million, you know it isn't true. I'm not sure but I think it would be 5%^2 = 0.25% (1 in every 400).

You're right.  I do have common sense, which is why I'd never pay for VGCPro even if I had the money to do so.  Just because it's free, doesn't mean I can't offer criticism about the methodology and the presentation however.  To me, there is large dissonance between what they think they are doing and what they actually are doing - at least in terms of how it's presented. 

And like I said, it's crazy how they publish the precise numbers as they do.  From my understanding, they collect data (if any at all) for a particular game or platform, run it through some equations to extrapolate on regional levels, and then publish whatever number they come up with.  The problem is, what is there starting data and how can they possibly estimate out to the nearest unit.  Just as example, there is no noticable difference between their reported sales of 158,346; 158,907; 159,623; 160,721; or 161,914.  None, and it likely goes further out than that.  What is the starting data they have here between these kinds of sales, and how much of the slight difference is just statistical noise?  How can they possibly say one game sold more than the other here and rank them?  I'll bring it up again because the response was not satisfactory - there should be no reason to publish to the nearest unit when you state they are estimates.  The only difference between publishing 158,346 and 160k is that one number is precise and appears correct and exact while the other is rounded off, appearing much more of an estimate (which it is) with reasonable doubt attached (which there should be).  And that's one main reason why the numbers they publish are misleading. 

Strange is why a member so new and with so little post count is so concerned about the precision of the data here...

Actually I can't see anyone here to have much motive to be upset, besides shareholders of said corporations why anyone else would care so much about the sales that two games in the ranking are separated by some few Ks and because of that one is 8th and the other 9th or something like that?

It might lack class to call the user base stupid but you have to admit that statement is a lot more accurate than early VGC numbers ;)

I thought it was a well known fact that this site was always wrong?

~Mod Edit~

This post has been moderated.

-Smeags

so you believe that VGC has any independant information on markets like Japan?

I didn't look it up which I admitted in my response (I believe).

Is there a statement on the site where we can understand how vgchartz does it's investigation? I would like to read that.

If bloomberg on board then it's hard to argue that vgchartz isn't creditable because bloomberg pretty tight on putting there name something.

http://www.vgchartz.com/methodology.php

Not even close. Based on ioi's statement and regarding the last weekly chart, we can be:

- 60% confident that each game of the top 10 will range from 50% to 150% (yet, this can make the 10th to be 1st and the 1st to be 10th!)

- 20% confident that each game of the top 10 will range from 67% to 133% (yet, this can make the 9th to be 1st and the 1st to be 9th!)

- 3% confident that each game of the top 10 will range from 83% to 117% (yet, this can make a game like GT6 to step up from 8th to 3rd!)

So even if we were in the lucky 3%, the ranking wouldn't have much credibility yet. So my question is: why is there a ranking?? I'm not criticizing the methodology used to get the data (like I said earlier, VG Chartz seems to be doing a good job there), I'm criticizing the way the figures are presented. Obviously when you consider each number individually the confidence is high but when you conciliate them in a ranking the margins of error grow exponentially to the point that there is no way we can trust in such a ranking. Even focusing only on the top 10, the confidence that each game to be in the right place is much less than 3% (once the 83%-117% level could still make a game to jump from 8th to 3rd!).

Therefore, I understand the people that say VG Chartz is misleading rather than calling them stupid. Above all, each person has an opinion (otherwise he/she wouldn't be here on the forum), which should be respected rather than ignored or target for insults. I don't know how a site admin can lower to such a level. If this is his userbase he should be promoting it, not flamebaiting it. If there are agressive comments made by users, why don't we make them less agressive (work on the community) instead of lowering to the same level? I don't have any problem of saying that. I prefer to be honest rather than flattering the admin just because he is powerful and important here. I think I'm even respecting him more this way.

What would you think of a person that calls stupid and idiot other users? A gentleman? Let's be honest, you would think such a person should be banned. But since he is the site admin let's defend him, right? Let's criticize the other side, which only said he lacked of class... so much hypocrisy I see here...

I can relate to the ranking being really incoerent even if we had just 5% of error, when accumulating the error for all items in the ranking it can get misplaced, but is that what really bothers you? And as brett put their top 10 were really close to NPD.