Real failure rate?

[Boogle]

I posted this in another thread, and found it profound enough to share with ppl not participating in said thread.

A washing machine fails in this country every 5 minutes. Washing machines must have an awfully high failure rate. (60 / 5) * 24 * 365 = 105,120 wachine machines dead each year.

When you say one fails every 5 minutes, it sounds like a massive number doesn't it? But when you work it out, it ends up being 'only' 105,120 machines in a whole year.

Now given that we know 500,000 X360s were sold in Europe in December, that means 12,500 should have failed (2.5% failure rate).

12500 / 31 (days) = 403.2 failures each day.
Thats 16.8 X360s failing each hour. Or one X360 failing every 3.57 minutes (60 / 16.8).

Now when you consider to have a failure rate HALF of the electronic standard (standard is 5%, MS once said it was a 2.5% return rate iirc), an X360 had to fail every 3.57 minutes during December, is it a wonder so many threads popped up?

 

[Boogle]

Well considering theres absolutely millions upon millions of washing machines churned out every day, then yes the failure rate would be high, when the 360 was released there was something like what a couple of hundred thousand, and yet look at the failuire rate, 35 on here alone, god forbid if they made about 10 million to ship on launch day, we would probably have about 5 million users just on here alone with faulty ones.

As Firestar_3x said, we're looking at percentages here, it doesn't matter if a billion X360s were in circulation, the failure rate in % would be the same. If you looked at my post you would see:

For a 2.5% failure rate, an X360 had to fail every 3.57 minutes during December in Europe alone

 

[Boogle]

Boogle there are a number of flaws in your mathematical reasoning but I can't be bothered to take the time to point them out, perhaps someone else would like to

I can explain a few if you want

First it assumes the 2.5% failure rate is accurate, and was an accurate figure for December alone. Chances are not as many failed in December, as say January or February. It only takes into account a single month, when for an accurate picture you want at least a year. It assumes that there is an even spacing between failures - when obviously they aren't.

However, total accuracy isn't of significant importance here - what is important is bringing the failure rate into perspective. Its easy to say 'wow it must be so much higher than 2.5% with X360s going down each day', when in fact, there should be far more than a 'couple' each day, even with a low failure rate such as 2.5%.

 

[Boogle]

I think for the amount of 360's that were released the failure rate is appaling, as i said above what would it be like if they managed to churn out 10 million for release day, 360's are going down day in day out, and thats out out of a couple of hundred thousand, there still in shortage, you still can't just walk in and pick one up, yet if you want a washing machine .........

Percentages mean diddly squat to me, whos to say what the failure rate would be if they released a billion, that would mean more going down, so the failure rate could be the same, or it could rise as they may be more faulty ones churned out, we just don't know as they only released a piddly amount.

I take it you don't understand the point of percentages Percentages 'normalize', for lack of a better word, data. By 'normalizing' different sets of data they can be compared directly. For example if I said 250 widgets were faulty, but 900 wooden legs were also faulty - you would think the wooden legs were awfully unreliable. But what if I said only 251 widgets were sold, and 9 billion wooden legs were sold?

For a bit of fun, if they did sell 10 million consoles in December, then 224 consoles would fail every minute (thats ~4 every second). That makes the X360 seem awfully unreliable, but the % of failures is the same - still 2.5%. The reason it seems so excessive is because a massive (absolutely massive) surge of products arrived on the scene at once.

Conversely, going back to 500,000 being sold in Europe - that means you have a 97.5% chance of getting a good console. Thats 487, 500 good consoles, vs. 12,500 bad ones. The chances of getting a faulty console are very, very small.

 

[Boogle]

We are as good a test sample as any so what makes your 2.5% (from somewhere) more accurate than the 18% (which we have polled ourselves) for here??

Because the test sample here isn't a fair sample. If I wanted to find out the % makeup of men/women/children in the country - would I get an accurate example if I stood in the middle of a town shortly after home time for either school or work? Would I get an accurate figure outside of a maternity clinic?

Same thing is happening here, if you have a problem (with anything technical) would you go to a technical forum... like this one? A quick google search for 'faulty Xbox 360' restricted to the UK quickly brings up the avforums thread. Do you think people with working Xbox 360s are going to go out of their way trying to find forums with failure rates for X360s? Conversely, do you think that people will go out of their way to find failure rates / help if their Xbox 360 has stopped working? The only place where you can get an accurate figure is to count the number of X360s going in and out of MS (out as in sold, in as in failed). According to MS a while back it was 2.5%, now they're saying its below the Industry average of 5% (presumably the figure will slowly increase to somewhere around the industry average once the X360 has been out for a while, which it seems to be doing if MS no longer give out the 2.5% figure). Its like trying to find out the % make up of Male/Female/Child when standing outside of a school at 8:30 - 9:00, you'll have a significantly higher proportion if Female/Child than is actually the case. The national census would probably be the best place to look.

 

[Boogle]

Would be interesting to have some release stats for the ps2 and % faulty units.

Did a little searching and came up with this: http://boardsus.playstation.com/pla....id=psx2&message.id=1296372&page=1&format=all

I'm sure it would have been lower than 6% at launch (the faulty optical drives would be what increased it to 6%). For the people scared about lots of faulty X360s - assuming 100 million PS2s sold (as in that linked thread) - that means 6 million PS2s out there are faulty.

Comparitively, if you go by the number of complaints to Watchdog (regarding the PS2), the % failure rate is nearer 0.0036%.

Sony's official line is that the PS2 has a lower failure rate than the PSX, which means the PSX had a failure rate (potentially) higher than 6%.

The stories of the X360 failure rate are definitely being exaggerated. If there are any PS3 fanboys, they must be having a field day.

 

[Boogle]

You are trying to compair two totally different polling samples styles. The samples are acurate if used in the correct manner, which this one was about as accurate a sample as can be.
The poll didnt give a percentage of the population that owned an xbox360, but can give a good reflection on the xbox 360 owners who have had problems. The poll also isnt opened to the public so people jsut doing a google search are not eligable to participate, only if they are registered with ocuk, a process which can take a while.
In that respect it is about as acurate as standing in Glasgow town center asking random people if they own an xbox then if they say yes, take the sample data from there answer after that.
So i do think it gives about as acurate a sample as you can get, even if you +-5% it could be as low as 15% but then again it could be as high as 25%.
I for one would take this sample over a 'somewhere below industry standard' for a failure rate (did they even mention the industry? the 360 can cover several, home electronics, home digital entertainment, household good, computer system, any, for which if you dont know then that information is useless

5% comes from MS, in a couple of interviews. First Google result to come up was Wikipedia: http://wiki.xboxic.com/Xbox_360. Other searches came up with many 3% figures, some 5% figures, and so forth, none above 5% thus far though.

The poll is still not reliable. There are thousands of people regged on this forum (probably over 100k by now, before the wipe ages back the user count was way over 100k). Some will have multiple accounts, some will hate MS, some will be serious PS3 fanboys. It only takes a small number of votes against the leader to make the lagging option catch up somewhat.

I'm sure the results would be quite different if we required the person to take a picture of the Xbox they own, and also note the serial number. That way you guarantee that almost all votes are valid.

I doubt MS are lying either, it wouldn't take much for a 'rumour' that the failure rate is 15% to spread like wildfire. Just need one person on the inside to idly mention the real figure, and MS will suddenly look very bad - assuming of course the real figure was in fact 15%. Press would love it, Sony would love it, MS would hate it

 

[Boogle]

If my 360 fails (which it may, it has a very early manufacturing date!) i'll ring up and get it replaced. Simple.

Doubt it, by going by the number of consoles that have really failed, then the odds are heavily stacked in your favour that it'll be fine