You'll almost certianly find they have already had a revision, or two.
IGN: Another Xbox 360 bites the dust.
- Thread starter "Nerve-Damage"
- Start date
You are quite correct. I intended to limit my statement to the size of the sample, not necessarily its composition. I have no knowledge of the kind of systems developers use, but if they are different from retail units, or if they are all from a single "bad batch", as I mentioned above, then the sample may not be useful, despite having more than adequate size.
For a sample size of 300, the margin of error on a confidence interval of 95% is 6%. So whatever EA's failure rate was (be kind and take the lower number, 30%), the real failure rate is almost certainly (i.e. with a 95% probability) between 24% and 36%. And that's horrible.
Bigus Dickus
Regular
For that batch of 300, or for the run that included that batch, yes, that's terrible. But as discussed above you can't extrapolate that to unlimited timeframes. The only thing you can say with confidence is that at some point in MS's manufacturing of 360's they have experienced quality issues. You can't generalize for how long, and can't extrapolate that rate of failure to the general population for any time you wish.
MS has admitted there were problem by mentioning a higher than exptected cost of repairs. They wont issue a recall though, even if there were serious problems, unless the problems were dangerous to the public. You dont want to do a product recall unless you absolutely have to.
There is NO way 360 is operating at a 30-40% failure rate. NO WAY!
It would be costing their division handily and they would of posted substantially more loses then they have over the last 12 months.
And for anyone to think a 300 sample size out of 10 million is representative of anything, that's doubly as insane. Too many factors come into play that can throw off the numbers. Anything less then 10's of thousand would need to be taken with a grain of salt.
It would be costing their division handily and they would of posted substantially more loses then they have over the last 12 months.
And for anyone to think a 300 sample size out of 10 million is representative of anything, that's doubly as insane. Too many factors come into play that can throw off the numbers. Anything less then 10's of thousand would need to be taken with a grain of salt.
Bigus Dickus
Regular
Oh, I'd be happy with 300, or a thousand or two, if they came from randomly selected samples throughout the production run of 360's from start until now.
All that report suggests is that there were at least some bad batches at some point, and MS's statements let us know that they have had some level of production problems (which isn't unexpected for new hardware). We can't assume that 30% (or whatever) holds for the entire lifespan based on the IGN batch. In fact, I'm inclined to agree that such a substantial failure rate across the board would have surfaced more obviously in financial statements, reports, and forecasts.
All that report suggests is that there were at least some bad batches at some point, and MS's statements let us know that they have had some level of production problems (which isn't unexpected for new hardware). We can't assume that 30% (or whatever) holds for the entire lifespan based on the IGN batch. In fact, I'm inclined to agree that such a substantial failure rate across the board would have surfaced more obviously in financial statements, reports, and forecasts.
Actually, the computations for margin of error assume an infinite sample space. You should take a statistics course some time, preferably a calculus-based one so you can get a better grasp of why it works.
Actually, the computations for margin of error assume an infinite sample space. You should take a statistics course some time, preferably a calculus-based one so you can get a better grasp of why it works.
Regarding all the people who get 4 failed X360s in a row, you could assume that MS has horrible quality control on refurbs. But for EA to get a 30% failure rate on a batch of 300 is pretty ridiculous. And like I said before, all you have to do is find out what Wal-Mart's return rate on X360s is, and that will pretty much clinch it.
Mintmaster
Veteran
If it is a random sample!
Excuse my ignorance, but how do we go about doing that?
I don't think so. The confidence interval (CI) is what it is, without regard to whether the sample is selected randomly from the population we wish to measure. What the CI tells us is that if we were to repeat this experiment exactly, that is, if we were to sample 300 developer systems made during the same time period as EA's systems, then we can have 90% (or 95% or whatever) confidence that the repeated experiment will yield results within the CI.
I assume that what you mean to argue is that 300 developer systems from EA are not representative of the population of 10 million retail Xbox 360s, and you may well be right. But there's nothing wrong (that I can see) with frearsomepirate's calculation of the CI, per se. It just may not be useful as evidence for us if EA's systems are not representative of the ones in our homes.
You know, it's strange. I have absolutely no evidence that there's any problem with the PSU, but I tried, twice, unsuccessfully, to get MS to accept return of my PSU and hard drive. My reasoning was that I had had three bad consoles in a row, and that the only things that hadn't been replaced were the power supply and hard drive. I spoke with four supervisors, but not one of them would set-up a return for me. One of them walked me through formatting the hard drive, and said that formatting was all MS could do if I returned it to them anyway, and all of them insisted that as long as the light on the PSU was green, it was OK . . . as though built-in tests have never failed before . . .
I did that, and it came up with the denail story from MS and a note saying the orginal article was pulled.
http://biz.gamedaily.com/industry/feature/?id=13636
I dunno. Get a job in Wal-Mart electronics or find someone who has one? Each store keeps pretty detailed records of its returns.
It is worth noting that pretty much every piece of evidence we've seen indicates a really high defect rate. I'd like to put the burden of proof on the people who say the defect rate is really quite fine. What evidence do we have that it's below 10%? All "My X360 works fine" proves is that the defect rate is below 100%.
Probably not. One of the hardware problems was with the RAM as Bizarre discovered with their devkits two months before launch.
I'm not sure if this problem would end up breaking the box completely though or if EA was referring to this...
Last edited by a moderator:
"I'd like to put the burden of proof on the people who say the defect rate is really quite fine"
Actually the burden of proof is on you to prove that the 360 has a higher than normal failure rate. Things that do have a higher than normal failure rates usually spawn lawsuits, recalls, consumer advisories, and can get the attention of the government. None of this has happened, MS has extended their warranty on launch systems and replaced broken units for free.
Until there are official numbers one way or another the burden of proof is on you as no evidence or a lack of evidence shows that the 360's are within a normal range of failures for a CE device.
Actually the burden of proof is on you to prove that the 360 has a higher than normal failure rate. Things that do have a higher than normal failure rates usually spawn lawsuits, recalls, consumer advisories, and can get the attention of the government. None of this has happened, MS has extended their warranty on launch systems and replaced broken units for free.
Until there are official numbers one way or another the burden of proof is on you as no evidence or a lack of evidence shows that the 360's are within a normal range of failures for a CE device.
We do have the proof. It's called Microsoft's earnings over the last 12 months. If they had a failure rate anywhere near what the naysayers are speculating, Micrsoft would not of finished anywhere close to being in the black as they have.
Forgive me, but please explain how you are determining the failure rate based on MS's financials. Can you break it down?
Similar threads
