Amazon U.S. February bestsellers and general Amazon-based discussion

[zorg1000]

SWORDF1SH said:

zorg1000 said:

Like i just said, for all we know #1 sold 10,000 last month while #50 sold 5,000 so using the other items around it means nothing.

Ill repost this

 

January Amazon rankings
NSW-13, 17
PS4-34, 53, 63
XBO-52, 68, 85

13 & 17 could be 8000 & 7000 units for a total of 15,000

34, 53 & 63 could be 5500, 4500 & 2500 for a total of 12,500

52, 68 & 85 could be 4700, 2300 & 2000 for a total of 9000.

Now lets say NPD figures for January are

NSW-300k
PS4-250k
XBO-180k

That would make the Amazon % of NPD totals the same for each console which would result in no bias.

So if a console bundle is wedged in the charts around  a bunch of games that you expect to sell around, let's say, 50k for the month. You wouldn't assume that bundle sold about 50k for the month?

I agree with what you're saying. I just don't agree with it being a deal breaker but rather something to consider. 

I've also mentioned that it's hard to prove because of how many ways the charts can be interpreted. I also accept that we need a clear and closer chart to prove the theory right or wrong. Like a scenario where ps4 is at 19 and 20 and Switch at 10 and 11.

Well the thing is we dont know if Amazon represents a similar percentage of overall hardware, software & accessories.

What i mean is that Amazon could make up 10% of software sales in the US but only 5% for hardware and a whopping 20% for accessories which makes it even harder to predict sales based on what is ranking near it.

 

 

Basically overall we should just continue to do what we have been doing and just use Amazon to predict rankings, not actual sales.