Xbox 360 PR Announcment Question..

[archangelmorph]

I was on gamesindustry.biz when I came across this:-

http://www.gamesindustry.biz/content_page.php?aid=22310

Chris Satchell has revealed that "over half" of the people who have bought Xbox 360 since its November 2005 launch did not buy an Xbox 1.

"We've sold 10.4 million, but the stat you may not have heard is that over half of those sales are from people that didn't own an Xbox 1," Satchell told GamesIndustry.biz in an interview published today.

This announcement came across to me as a difficult one to hold down..

It made me wonder..

How in the hell do they know that??

Being an owner of the previous Xbox I definitely remember not having informed MS specifically that I purchased there console nor do I remember giving the retailer any personal details on purchase..
With this in mind, when I eventually go on to the store to purchase my Xbox360, I will definitely not be giving the retailer any of my personal details once more and even if I did, that fact that I didn't the last time would mean that they have no way of knowing that the Xbox I bought in 2003 was purchased by the same person as the guy who bought an Xbox360 in 2007..

I considered that maybe they were estimating from the total install base figures but this smells like crap because didn't the orignal xbox rack up 80 million units sold? whereas the 360's current install base is only 10.4 million according to the article and therefore it's impossible to estimate the proportion of 360 owners who owned an Xbox from these numbers..

Then I considered maybe they are using the Xbox Live figures (since the service records users details in a way that can track user migration from Xbox1 to 360..) but this still doesn't make sense in the context of the announcement since i'm sure the original XBL install base was only around 2 million right? (and a significantly small proportion of the install base, whereby any kind of interpolation can't be accurately done..)

It just seems like they "made up" these so-called "findings" unless they did surveys and interpolated over the entire installbase but depending on the size of the installbase it is probably just as accurate as pulling figures from ones behind..

For all you marketing folks here at B3D, any thoughts on how MS came to this seemingly-imaginary conclusion?

 

[Qroach]

I think they are probably judging by xbox live carry overs. Keeping your xbox live account from before.

 

[J_Saint]

Surveys? I've received them from time to time in the past few years from MS and Sony. Both will usually include a question like "Please check all consoles you own." While not every 360 owner submitted one perhaps they were able to conclude a general number of people who did not buy the Xbox.

Could it also be the number of people who registered their 360 but not an Xbox?

 

[archangelmorph]

First readjust your original xbox numbers, it sold NOWHERE near 80 million units, I'm not sure where you derived that. I believe it is around 25-28 million. Which would then put the 360 at almost half the install base of the original Xbox, in about a year or just over, one would think that if the 360 does continue at such a yearly pace that people would have to be new owners.

I don't remember the figures but my point is this..

They aren't talking about having over half the install base new to the Xbox brand in a few yrs time..

They are talking about that being the case now..

Read the article!

(So argueing numbers i clearly showed i'm not clear on is a bit of a pointless activity to be fair..)

 

[archangelmorph]

Surveys? I've received them from time to time in the past few years from MS and Sony. Both will usually include a question like "Please check all consoles you own." While not every 360 owner submitted one perhaps they were able to conclude a general number of people who did not buy the Xbox.

Could it also be the number of people who registered their 360 but not an Xbox?

But how many people could have realistically participated in such surveys?

And would that number "realistically" represent the entire install base?

Also I don't kno about the US but here in the UK i'm sure the vast majority of people who buy games consoles have never registered them.. Maybe those who got the Live subscription (I would imagine the Xbox would register your details automatically when you setup the account) but again, judging from the minute Live install base of the original xbox as a proportion of the total install base, does this paint a "realistic picture" of the total install base?

My guess is any statistician would probably tell you that they don't..

Oh and does this take into account the following:-

Resale of the console from one customer to another - which one would be classed as the owner of the unit? and if either one of them bought a 360 which would be classed as a carry over cmr?

Sales of consoles bought by individuals from wholesalers who sell the units on Ebay?

Units given away as prizes in competiton's (I remember Amazon did some a while back)..?

It's all quite confusing..

 

[Johnny Awesome]

MS regularly polls their market. All they need is a sample size of 2000 people to confirm these numbers true. If it's true for 2000 it's likely true for 10 million. That's statistics.

 

[Cheezdoodles]

But how many people could have realistically participated in such surveys?

And would that number "realistically" represent the entire install base?
:

You dont need to poll every single user to get a result that has a very high probability of being correct, statistically. Its the same when you see all the surveys on who's going to vote for the republicans or the democrats before the elections, nobody is going to poll 200million americans, you only need a couple thusand to get a number that is very close to the truth.

 

[Arwin]

I can think of two main reasons

1. Say that the demographic for the xbox is 20-30 year olds (just an example). Only half of those still fit that demographic 5 years later.
2. I noticed that a fair few people with HDtvs bought 360s for them so they could play games in HD

 

[PARANOiA]

You dont need to poll every single user to get a result that has a very high probability of being correct, statistically.

This is the correct answer. You don't have to ask 10 million people to get this data.

 

[Nesh]

MS regularly polls their market. All they need is a sample size of 2000 people to confirm these numbers true. If it's true for 2000 it's likely true for 10 million. That's statistics.

Thats an extremely small sample. Thats not statistics. Thats poor statistics with innacurate results. If that was a serious research done by serious researchers and had only a sample of 2000 among 10 million, they wouldnt have annouced that result. Or atleast they would have done it with a notice. But as usual like all the companies and politicians thats PR statistics. I think Nintendo is the only safe among the three.

They dont have to announce anything. Wii is selling so fast, people cant ignore whats going on so they dont have to go PR.

They are the quitetest of all while MS and Sony go crazy in a PR war trying to persuade which is best. Nintendo should be laughing by now

btw: I think it could be true in some exctent though since 360 shows a completely different but better face than what XBOX1 has shown. Still their "statistics" should be poor

 

[Graham]

Thats an extremely small sample. Thats not statistics. Thats poor statistics with innacurate results. If that was a serious research done by serious researchers and had only a sample of 2000 among 10 million, they wouldnt have annouced that result. Or atleast they would have done it with a notice. But as usual like all the companies and politicians thats PR statistics. I think Nintendo is the only safe among the three.

I think he just pulled the number 2000 out of his cat... That said, a random sample of 2000 would be fine. Provided the... er... standard deviation (?? - been ages since I did stat.. ) is small, if, say, 65% of your sample didn't own an xbox1, and the SD showed a 99% confidence to within +/- 10%, then you could quite easily say that half your user base did not own an xbox1. Stat is an odd thing, but it is accurate.
I personally do not own an xbox1, yet it seems there has been at least one (currently 5) in every flat I've lived in. I own a 360...

 

[Nesh]

I think he just pulled the number 2000 out of his cat... That said, a random sample of 2000 would be fine. Provided the... er... standard deviation (?? - been ages since I did stat.. ) is small, if, say, 65% of your sample didn't own an xbox1, and the SD showed a 99% confidence to within +/- 10%, then you could quite easily say that half your user base did not own an xbox1. Stat is an odd thing, but it is accurate.
I personally do not own an xbox1, yet it seems there has been at least one (currently 5) in every flat I've lived in. I own a 360...

It gets accurate the better and bigger the sample thats for sure. The nature of the sample plays a role also. Which makes you wonder how "random" their sample was. I dont know if they used a real actual model to test the significance of the variables (whatever they are if they indeed used any)

But if what they did was just take the sample, threw assumptions here and there that favor the result they wanted to get and just derived percentages then this is no real statistic

 

[Cheezdoodles]

I think he just pulled the number 2000 out of his cat... That said, a random sample of 2000 would be fine. Provided the... er... standard deviation (?? - been ages since I did stat.. ) is small, if, say, 65% of your sample didn't own an xbox1, and the SD showed a 99% confidence to within +/- 10%, then you could quite easily say that half your user base did not own an xbox1. Stat is an odd thing, but it is accurate.
.

For a 99% confidence interval you only need to poll c=1\2(n+1)-z(æ)* (1\2 sq N) take that number and substract it on both sides of sides... Meaning 4070 people

(ofcourse i havent had Statistics in about 2 years, and i just snapped the formula from my old book, so i might be wrong)

 

[Johnny Awesome]

Mystery solved.

 

[baten]

I thinks Satchell is just extrapolating based on some sales data they have. If there were realy statistical work done, I'm sure they would have provided more info about it, since in my opinion, this is a great achevment for MS, with more implications than is first appear, and could also provide the answer for an even more interesting/intriguing question: Where do the other 50% come from?
If they are from the PS2 heritage, then MS is on a very good path with 360.

 

[Tap In]

Robbie Bach said at CES (during the keynote) that this data was collected from SURVEYS of 360 owners.

see informative posts above for how that data is handled. :smile:

 

[baten]

Pardon for bumping this - but it just came to me that, if indeed 50% of the 360 owners never got an xbox 1, then there is a rather intriguing question: what are the other 20 mil former xbox1 owners waiting for?
There were 25 mil (roughly) xbox1 sold. 5 mil - according to MS - have made the upgrade to 360. But Xbox1 is basically shut down, there are very few games coming out, if any, 360 is out for over a year with lots of great games - what is happening with these 20 mil xboxers, what are they waiting for? Price drops? Halo3?

 

[assen]

what is happening with these 20 mil xboxers, what are they waiting for? Price drops? Halo3?

Is that *that* surprising? Most of the 25 mil xbox1 owners bought theirs at $200 or below.

 

[baten]

Is that *that* surprising? Most of the 25 mil xbox1 owners bought theirs at $200 or below.

Well, you can find good offers and bundles on core systems that are almost there on price.
Anyway, what are they doing meanwhile? What are they playing?

 

[assen]

Well, you can find good offers and bundles on core systems that are almost there on price.
Anyway, what are they doing meanwhile? What are they playing?

I'm waiting for the price to hit 200 EUR. I'm playing my PC and PS2 in the meantime. (And occasionally playing on the 360 in the office, I must admit.) Don't know about the others, but I suspect I'm not the only one