NPD: Percentages of US Switch owners that own PS4/Xbone/both/neither

[Pyro as Bill]

Ka-pi96 said:

Pyro as Bill said:

If you shuffle (randomise) a pack of playing cards, how many cards would you have to turn over before you could accurately predict that they're 50:50 red/black and 25:25:25:25 hearts, diamonds, spades and clubs?

If you shuffled a billion packs of playing cards, how many cards do you think you'd have to turn over before getting a 95%+ or 99%+ accurate result?

(a random sample of 1000 is enough to predict a population size in the tens or hundreds of millions)

You'd have to turn over 0. A pack of playing cards, regardless of how well you shuffle it, is always going to be 50/50 black and 1/4 each suit

Even if you didn't already know the result, there's still the possibility of an extreme result. You could theoretically look at 50% of the cards and find them all to be black, thus the logical (albeit incorrect) conclusion would be that all of the cards are black.

Do you know what the odds of that happening are?

Additionally a random sample that's legitimately representative of the general populace is a statistical improbability. Where exactly did they do the survey? If it was via the NPD site itself, or through a game store or something, then you'd expect the majority of respondents to be more hardcore gamers and thus you'd also expect them to be more likely to own multiple consoles than the average person.

It's very important that the sample is as representative as possible.

Oh and I'm not saying this statement should be dismissed either. I'm just asking for the source to clarify how accurate it's likely to be. If it's revealed that it was a survey of 12 people in one city then this information is basically worthless, while if it was a survey of 120,000 people spread out across the country then it's much more meaningful.

Overkill.