The hugely scientific console reliability poll

[Tap In]

Everybody does seem to agree that the Xbox360 fails at an alarming and unacceptable high rate.

one of the problems with this is that the LAUNCH systems failed at an alarming rate... as MS has admitted. I'd guess 15-20% (maybe as high as 30% even?) but the past year I'd guess it is down to a more acceptable level of 3-5%. I know several people who have gotten replacements that know that the new system build runs quieter and cooler than their launch units. (they had to have changed something in the fabrication process after admitting the launch units had problems, that's a no-brainer)

The difference between the units built in 2005/early 06 and the ones built since are another reason why the results of this POLL are useless for determining what the 360's rate is TODAY.

edit: refined my thoughts

 

[Todd33]

one of the problems with this is that the LAUNCH systems failed at an alarming rate... as MS has admitted. I'd guess 15-20% but the past year I'd guess it is down to a more acceptable level of 3-5%. I know many people who have gotten replacements that know that the new system build runs quieter and cooler than their launch units. (they had to have changed something in the fabrication process after admitting the launch units had problems, that's a no-brainer)

The difference between the units built in 2005/early 06 and the ones built since are another reason why this POLL is useless for determining what the 360's rate is TODAY.

As I understand it MS has changed very little since launch. Even the Elite is almost identical to the original design. I've never read a report of a newer 360 being cooler, quite maybe, but that's a DVD drive issue.

Is there any credible evidence that MS has changed the internal design to make it more reliable? I hope it's true, I've been waiting to trade my 360 in due to read issue, but I figured the longer I wait the better.

 

[Cheezdoodles]

Why the limitation?

Why dont you try reading some posts about this issue in this very thread, it has been explained a gazillion of times.

Even direct quotes from the national Science association and similar organizations regarding just why it's not representative for anything else that the sample population itself.

If anything I'd believe it's the opposite since...

I'm sorry anything you people believe or think is at this point irrelevant. There has been provided a number of explanations to all your thoughts (atleast all i have seen), also quotes from probably some of the people with most knowledge about statistics in the whole world.

At this point, unless you can make a mathematical or statistical argument, don't bother. People need to start either reading posts and try to understand it, ask us to elaborate if the explanations are to bad or you dont understand it, but if you don't know the basics of statistics, frankly, what you think is irrelevant, because its flat out wrong.

 

[Shifty Geezer]

This thread is going nowhere. .

Definitely. Discussion has crashed into correct statistical methodology versus empirical observation and the like. I understand Ostepop's POV, but don't agree with it, but there's no point thrashing out the arguments any more. We've both said our piece and the POVs are up for others to consider. The poll is in and done. Those that want to view the data and consider for themselves what they mean can do so. Those that think it's bunk can ignore them.

 

[mckmas8808]

Failure rates are probably around 15%.

 

[Todd33]

Microsoft's Peter Moore Dodges Complaints About 360 Failure

 

[Ben-Nice]

Microsoft's Peter Moore Dodges Complaints About 360 Failure

What problem? He clearly states

Y’know, things break

Which is true

 

[Bigus Dickus]

Why the limitation?

Do you think B3D users are some kind of super duper console players that break consoles mroe easily than others or something?

If anything I'd believe it's the opposite since I get the impression the average age of people here is fairly high. This means kids families etc. And most of us probably play on PCs too which competes with time played on consoels.

Unless you haven't been paying attention it should be no news to you that kids today game A LOT. And it's cheaper for a parent to buy the kid a console than to buy a top notch PC to game on. Not to mention the kid might want a console more than a PC.

So console use by youth is probably at least comparable to your average B3Der and quite probably even higher.

Ok I'm done now passing off opinions as facts.
Peace.

All that is really irrelevant. The bottom line is that self selective samples are rarely, if ever, stastically sound. Bias is rampant. The results are useless outside the sample, which is precisely what I stated. Period.

 

[Todd33]

All that is really irrelevant. The bottom line is that self selective samples are rarely, if ever, stastically sound. Bias is rampant. The results are useless outside the sample, which is precisely what I stated. Period.

I'd say B3D is biased for a longer 360 lifespan. We play less than the average 360 owner and maintain (ventilate, etc.) our equipment better most would bet. You can write off the results if you want, but the poll is a good metric in the relative sense (we will never get absolute numbers). And yes I know statistics, but no one is claiming this is a scientific poll, so dismissing it based on that is ridiculous.

 

[NovemberMike]

I'd say B3D is biased for a longer 360 lifespan. We play less than the average 360 owner and maintain (ventilate, etc.) our equipment better most would bet. You can write off the results if you want, but the poll is a good metric in the relative sense (we will never get absolute numbers). And yes I know statistics, but no one is claiming this is a scientific poll, so dismissing it based on that is ridiculous.

I don't feel like writing too much since I just got done with the AP Stat test (and need to study for calc), but the largest problem is that this is voluntary response. I can't remember right now, but I think that is a formal form of bias.

 

[Bigus Dickus]

Failure rates are probably around 15%.

Based on?

 

[Bigus Dickus]

I'd say B3D is biased for a longer 360 lifespan.

Based on?

We play less than the average 360 owner and maintain (ventilate, etc.) our equipment better most would bet.

I could argue the opposite equally well. I've seen pictures people post of their setups here, and they are far from optimal. You really think B3D gamers play fewer hours than the joe random consumer? Based on?

You can write off the results if you want, but the poll is a good metric in the relative sense

No, it isn't.

And yes I know statistics

You have yet to show it.

but no one is claiming this is a scientific poll

Claiming the results can be extrapolated any further than the subset of B3D posters who felt like voting is making some claim to scientific validity, which quite a few have tried to do.

OTOH, looking at the results and thinking "hmm, I wonder if 360's have a high failure rate" is fine. This poll does provide an interesting result to prompt one to think and seek further (sound) research. However, the numbers contained herein are essentially useless beyond that thought prodding.

This poll merely poses a question, it does nothing towards answering it. Asking the question is fine. I do. I'm not refuting the question's validity. No one is. I'm not refuting that the answer, based on the only hard numbers available, seems to be "higher than normal." I am refuting that this poll can add to an answer.

 

[Todd33]

I would surmise that the average age of the B3D poster is well above the average age of all 360 owners (we had an age poll recently). We are educated, have careers, families, etc. This all means less play time and thus longer console lifespans.

I'd love to hear the theory on how B3D represents a group that plays more, is rougher and thus has more failing 360s. I also want to hear how this same group somehow is easier on every other console including the notoriously flaky PS2.

 

[fearsomepirate]

Any statistical data, no matter how scientific, can only establish probability. It can never absolutely prove anything. This little poll at best only establishes a high probability that between 28% and 48% of the consoles manufactured in the window which the demographic tended to buy them (probably mostly first-run machines) were defective.

And given that the PS3 and Wii results didn't show the same skew, meaning that the poll wasn't mainly attracting people with broken machines, I think it is safe to say that at the bare minimum, assuming no malicious interference, this poll establishes a high probability that some particular manufacturing run of X360 consoles had a much higher defect rate than 5%.

 

[patsu]

That is has been explained in detail as to why it is important by a number of posts, by me and other posters also by
publicagenda, American Statistical Assosiation , National Science Assosiation, this is where the links are coming from, they all say, You have to have data collected in a proper fashion, or else the results your going to get will be flawed.

Sure... what makes you think that the current approach is totally wrong and useless ? I didn't even say the percentages are correct/precise by and large... but I believe that by understanding where the shortfalls are, we can have better intuition about the numbers. We can also improve our poll next time. So what if the data are only valid for B3D community ? They are still meaningful for me (keeping in mind the assumptions). If we truly want to measure the failure rate, we need to change the poll to measure per-box/SKU rather than per-user breakdown experience. Since we didn't, the poll is already "gimped".

Its irrelevant as to what you are polling. If you do not collect data in a proper fashion, and a specially if your sample is based on volunteers, they are going to get flawed because of personal interest (a person with a broken console is more interested in reporting this, than a person with a working one, for example) and biases.

As I mentioned, this behaviour applies equally to PS3 and Wii. So it's indicative in that manner... like how you acknowledged that 360 reliability sucked compared to PS3 and Wii. That's all. Do I think the failure rate is 30% ? I don't know enough and am not interested to comment on it.

What part of people voluntarily seeking out this topic and voting, and people being randomly selected to participate in a poll dont you understand the difference of? Tell me and i will answer.

The B3D pool can just represent one of the random sample pools. If additional data is collected for the Xbox 360 populations, you can actually qualify the data some more before using them.

It has already been explained in the prior posts, and its also pointed out by the articles linked and quoted from made by the National Science and Statistics Associations: the fact alone that this is a voluntary poll is going to tamper with results.

What you think will affect the poll result or not is not something im going to try to bother spending more time on explaining, as your asking the exact same question regardless of the answer.

This has already been explained in detail several times.

The sample population should be Xbox 360s... not consumer populations. If we ask the voters their 360s' date of purchase/manufacture, we can actually profile and reason about the consoles by qualifying 360s that spread across a wide range of purchase date. I don't think I'm violating whatever scientific article you cited. The poll is just one of the ways to gather raw data. You can still qualify them.

Really? YOU think polling have some advantages? What experience do you have with statistics, if you dont mind me asking?

Took a few courses... worked on a one or two market research projects (I wrote the code, the [chair] marketing professor did the formula). Assisted a few shopping malls to organize and analyze their customer data.... before bringing in the big guns.

Also saw so-called market research companies fudging results... and hiring participants that do not fit the survey profile for $50... just to finish survey quicker and move on to another project.

Like I said, there are advantages and disadvantages in various mechanisms. As long as you are aware of the caveat, and you know what you want... you spend less (more) to get the information quickly (or carefully).

Its not a matter of how hard it is to filter out overlapping data, its having data that is random enough to be accurate (sounds kinda wierd huh?) and that also fits certain (already explained characteristics in earlier posts) if your going to combine sample data. It is crazy hard.

You can still qualify the data after collecting them.
The rest of your article just regurgitate the same basic idea, but I think you're oversimplify the data collection process. We usually need to cleanse and qualify the data before use. This is assuming if we have time in real life

 

[Diamond.G]

Failure rates are probably around 15%.

Is that only of launch units? I like translating the percentages to real numbers. 15% is 1.5 million boxen (i know not a word but I like the way it sounds). Doesn't that seem a trite unrealistic?

 

[Bigus Dickus]

I would surmise that the average age of the B3D poster is well above the average age of all 360 owners (we had an age poll recently). We are educated, have careers, families, etc. This all means less play time and thus longer console lifespans.

I'd love to hear the theory on how B3D represents a group that plays more, is rougher and thus has more failing 360s. I also want to hear how this same group somehow is easier on every other console including the notoriously flaky PS2.

Children are subject to the rules of their parents, and presumably have limited gaming time. Teenagers are likely the roughest, and there are plenty of those on B3D. Adults who frequent B3D obviously enjoy gaming, and one could argue that they would spend more time gaming than an adult of equal age who does not frequenty the forum (i.e., joe casual that plays madden football when some friends come over once a month). And again, I've seen pictures of people's setups and see nothing suggesting B3D members take exceptional care of hardware.

 

[Todd33]

Children are subject to the rules of their parents, and presumably have limited gaming time. Teenagers are likely the roughest, and there are plenty of those on B3D. Adults who frequent B3D obviously enjoy gaming, and one could argue that they would spend more time gaming than an adult of equal age who does not frequenty the forum (i.e., joe casual that plays madden football when some friends come over once a month). And again, I've seen pictures of people's setups and see nothing suggesting B3D members take exceptional care of hardware.

Well that's one theory. B3D breaks 360s at a higher rate than all other consoles.

You might want to head to the AVSforums next and make sure they understand they are at fault too

http://www.avsforum.com/avs-vb/showthread.php?t=845026


Edit: B3D age poll

http://forum.beyond3d.com/showthread.php?t=38917

Kids play lots of games, I know a few teenagers that play(ed) WoW 4-6 hours a day, that's about how much I game a week. The same demographic is the key 360 gamer. So why anyone would think the typical B3D poster plays more or is harder on a console is pure fantasy IMO.

 

[Arwin]

Is that only of launch units? I like translating the percentages to real numbers. 15% is 1.5 million boxen (i know not a word but I like the way it sounds). Doesn't that seem a trite unrealistic?

I think launch units is significantly higher than that, and currently they are doing better. That also skewes this board's results, I think, because the crowd here is much more likely to have a launch unit than the average consumer.

I have to think of the 1up yours show, where a majority of the 360s have died. (Though the last one in a very funny way ... )

 

[Cheezdoodles]

You might want to head to the AVSforums next and make sure they understand they are at fault too

Now, instead of trying to be a smart ass, you might wanna try to actually read and try to understand what me and a few others have been pointing out.

You may also, try to use your brain and re-read what your linking to, your linking to a thread, where people talk about x360 failure rates, and people talk about their personal experiences. Its a "official x360 hardware sucks" bitch thread, (just like the neogaf thread linked to earlier), there is no one reporting that their x360 is working fine, its not even a poll.

There is 30 people there who have said that their x360 has broken down, out of 400 000 members on that forum. That thread of course, is to completely useless to use for any statistics, but they aren't doing that, they are just commenting and complaining about reliability.