By using this site, you agree to our Privacy Policy and our Terms of Use. Close
Kasz216 on 14 January 2010
Edit Thread
theprof00 said:
tombi123 said:
theprof00 said:
tombi123 said:
theprof00 said:
tombi123 said:
By the way Darth, these error margins mean PS3 fans can claim that the PS3 sold 1M units in a week, two weeks in a row, given a 10% error margin.

vgchartz 10% error margin?

Roughly (which is pretty good).

I agree. I want to know why everyone is saying NPD is a statistical tie when they don't know the margin of error.

NPD doesn't track a high enough percentage of the market to be accurate to within a couple of percent.

Well, it would require about a 2% margin of error to be a statistical tie, but even then, all it means is that the difference between can be covered by the margin. Additionally, this only works in one scenario of comparative data. Statistical tie is misleading because it's comparing two numbers on the same margin. Some percentages may have higher margins of error. For example, the 360 might have an 8% margin of error, and the ps3 might have a 2% margin.

Additionally, that scenario would be only if the 360 was really higher and the ps3 was lower. One could stay still and the other could drop, or go higher, and vice versa.

It's still considered a statistical tie... it's more likely the PS3 beat the 360... however it's within the realm of statistical tie... making the numbers more or less even.

Last i've heard quoted on this site was a 5% margin of error for sales tracking in general with the systems.  Which is actually pretty good. 

There margin of error is the same across the board because the adjust for individual sampling errors in their final numbers.