It is not a vague statement at all.  They both explicitly use retail and end user data, sampling and historical trends to extrapolate numbers.

The OP and several of my posts since make it abundantly clear that due to a larger and more representative sample base NPD is more accurate. Indeed, at no point in this entire thread have I (or anyone as far as I can see) critcised NPD. You're arguing with thin air buddy!

Don't condescend to me. That is an incredibly vague statement that doesn't even begin to describe their methods. As per usual you're just making shit up to support your point. I'm sorry you don't realize that ioi has long since given up on caring about the site or it's tracking. 

VGC do say they can track pretty much in real time but it needs to be taken as an estimate. So yeah I agree going to a bi-weekly update is moving away from what makes this site different to NPD, GFK etc. The great thing at the end of the day is they make this data available for us to discuss, just hope it remains weekly :) 

Solution: Ban stupid people.

That could be changed...

On topic:

The two issues I'm seeing really are with Vita last month and with Xbox One over the last couple of months in the USA - both hardware related. In the case of Vita, our system simply didn't tag the Borderlands 2 bundle as hardware (so we were just counting it against software sales for the game) and in the case of Xbox One our samples simply seemed to have sold a disproportionately large number of pieces of hardware. I have been away a lot over the last month or so and have left much of the data work in the hands of others which is why the issues were missed. Things happen I'm afraid.

Part of the original appeal of VGChartz was its ability to offer sales numbers guidance in the West at a greater rate than any other publication. Over the years a very strong community has built up through the sites forums, game database and publications which is fantastic.

In recent times, despite VGChartz consistently providing very accurate short and medium term sales estimates for most consoles, the degree of vitriol it attracts for 'guessing,' 'making the numbers up' or for being 'biased' (in favour of any number of consoles!!) has anecdotally increased dramatically. There is some suggestion that this might be a by-product of a new generation starting, which is in some way supported by the majority of vitriol being directed at Xbox One numbers over the last couple of months.

However, I would suggest that one solution would be to cut back numbers reporting to once every two weeks.  This would have multiple advantages, without sacrificing VGChartz unique selling point in consistent reporting. For starters, it would double the amount of data available for each iteration of numbers, allowing greater opportunity for anomalies to be detected.  Ultimately anyone with an understanding of the numbers knows that for the most part, VGC has a remarkable solid medium and long term strike rate for hardware reporting.  But having a two week reporting schedule would cut back (but of course not eliminate) some of the ill-informed knee-jerk responses week-in and week-out, arguments over the degree of impact 2 day versus 5 day price cuts should have had etc.

This is not an idea I am firm on, but a suggestion that has occured to me. I'd welcome other people's opinions.

Edit: Lets keep it to hardware, software is a whole other kettle of fish.

I should stress that though I am a forum moderator, I have no major links to the editorial team and all (vague) opinions expressed above are my own inane thoughts.

What is the benefit of going to bi-weekly? Just halving the number of times that the vitriol is pointed at the site?

I guess I just don't see the real benefit. Sure, collecting twice the information might mean a reduction of about 30% in error margins (depending on how you do it), if you report the total... but I'm getting the sense that this is more of a "hold off from reporting the information every second week, release two weeks simultaneously" approach that doesn't help with numbers at all.

If I were going to suggest something for the purposes of making what VGC does clearer to reduce the vitriol, it would be to calculate 95% confidence intervals for data (at least for hardware) and display it in hover text, so that people can see just how much certainty there is in the numbers. These intervals should be properly-calculated, rather than just an estimate of "about 5%" or whatever it happens to be based on a priori guesses.

Because they don't use the forums. If you have a genuine issue (beyond just moaning because a game is "undertracked" by a few hundred thousand) then email data AT vgchartz.com.

I get your concern about UNDERTRACKED/OVERTRACKED even back @ the start we got banned for saying undertracked without backing it up without facts remember Doimedes1970 or something?

But that was the charm back then aside from the monthly numbers and later on the weekly numbers numbers, a vocal group of users assessing the data and giving feedback and seeing results on that feedback. This is the very first time for myself even seeing data@vgchartz.com being mentioned. Whoever fixed  the banlinks could also work that mail  in the forum rules maybe?

That is such a vague statement that I"m not sure how you equate it to being the exact same methodology that VGC uses. Sounds like you're making up a bunch of BS to make VGC's horrendous tracking look better. And yes it's horrendous. Defend it all you want but that doesn't stop the numbers from looking like a joke every month.  NPD is taken as *fact* by some of the largest corporations in the world. If it's good enough for billion dollar corporations to base their entire market strategy around then I'm not sure why it's not good enough for you.