And you thought 33% was bad!

[Raesvelg]

It's also a functionally universal practice. Every company does it, to varying extents. Microsoft certainly does it at least as much as Sony.

 

[Tristam Green]

It's also a functionally universal practice. Every company does it, to varying extents. Microsoft certainly does it at least as much as Sony.

http://itsbetterwithwindows.com/

boioioioing

 

[Bustaballs]

I know dozens of people that own 360s and every one of them has had to replace or have it repaired at some point. One of them is working on his 10th 360. The DVD-Rom drive went out in mine a while back. If that number is wrong, it's because it's too low.

 

[Sorrowfulturnip]

Mine RRoD for the first time just last week and I've had it about 2 years this month. It lasted for quite awhile and I was getting my hopes up that I was one of the "lucky" ones. Guess that didn't work out lol. I did use it a lot. It was nothing for me to play a game on it for 2-3 hours and then turn around and watch another 2 movies on it. I shipped it and am currently waiting for its return.

 

[Bajaratt]

http://www.gamespot.com/news/6216691.html

Another study, but one with actual tracking of 360 failure rates over time with the different versions that have been created. This was made by a company that provides the warranties and tracks based on the warranties used, so it's got actually accurate figures rather than just asking random people with no proof that they're being truthful.

Xbox: 23.7% failure
PS3: 10% failure
Wii: 2.7% failure

According to their tracking, the latest version of the 360 has a less than a 1% failure rate. The PS3's failure rate seems high, but the Wii's seems pretty good. Still, 23.7% failure total on the 360's seems a much more believable figure than 54%. Sample size for the study is still too low for my taste, but the margin for error is drastically lower than with the game informer poll.

 

[Raesvelg]

There is a problem with sample bias in that survey, however, since they only took the information from people who purchased add-on warranties.

The truth probably lies somewhere in the oft-quoted 33%.

 

[Chasym]

from gizmodo

FairTrade, the company that's making this claim, bases their little theory on failure rates correlated with purchase date. Before the Jasper chipset, they were atrocious. But after? Not too bad!

"In Q4 2008, Jasper units started arriving, although we believe units purchased during this period continued to be a mix of Falcon and Jasper models. Even with this mix, we projected the 1-year failure rate to drop below 4%. Furthermore, when looking at over 500 units purchased in 2009, fewer than 1% of customers have reported a RROD error as of Aug 2009. It is still too early to definitively assert that Jasper has given RROD a knockout punch, but such an argument may be pronounceable in the coming months."

That's down from an (admittedly) conservative RRoD rate of about 12%, which accounted for more than half of the console's overall failures, which stood at 23.7%. They explicitly say it's too early to call this one, but early signs--even if all the numbers are skewed a little low--are definitely promising.

 

[Bajaratt]

There is a problem with sample bias in that survey, however, since they only took the information from people who purchased add-on warranties.

The truth probably lies somewhere in the oft-quoted 33%.

How is it biased? They have accurate numbers of how many warranties were sold vs. how many were actually used; that's the next best thing to actual warranty usage by the company itself.

More samples would be excellent, but they're working with what they have, and with that significantly higher accuracy, they don't need as big of a sample size and Game Informer's polling method would require.

 

[Raesvelg]

It's biased on behavior of the subjects, obviously. I shouldn't have to spell that out. It's akin to polling the residents of a particular neighborhood as to how effective their city government is, rather than polling a broad spectrum of the city.

For all your ranting about the poll in the OP and how tiny the sample size was, you were awfully quick to claim the 1% failure rate of new consoles, when the sample size is a whopping 500 machines.

The truth of the matter is that you can find a particular survey to support pretty much any stance you care to name.

 

[Bajaratt]

It's biased on behavior of the subjects, obviously. I shouldn't have to spell that out.

Actually, you are going to have to spell out how that creates a bias in the poll, because that doesn't make any sense at all.

It's akin to polling the residents of a particular neighborhood as to how effective their city government is, rather than polling a broad spectrum of the city.

No, it's not, in any way akin to that. Your example is directly introducing a bias into the poll; but there's nothing like that in their numbers, since who the people are have absolutely nothing to do with whether their consoles broke or not under the warranty. If anything, people with warranties are less likely to take care of their products, since they know they can get it fixed if it breaks; meaning the "real" results would be lower.

People who have their console broke who have a warranty are going to get it repaired by whoever warrantied the console. Therefore you have a perfectly accurate, albeit very limited, sample of how often a console fails.

For all your ranting about the poll in the OP and how tiny the sample size was, you were awfully quick to claim the 1% failure rate of new consoles, when the sample size is a whopping 500 machines.

If you re-read what I posted, I didn't claim anything like that. All I said is that they're tracking the different versions of the hardware and that the new hardware is much more stable than the old. I didn't say that it was proof, I even said that the sample size was still quite small(twice, mind you); but because of the accuracy of the figures, that the margin of error is much lower than the game informer poll.

The truth of the matter is that you can find a particular survey to support pretty much any stance you care to name.

I didn't go looking for anything, I just happened to see the poll while checking out another article. While it's true you can use statistics to prove anything you want, some polls are more valid than others. All I did was post a poll with significantly more accurate results.

 

[TheOsyris]

Baj i founs that same example earlier today on the front page of Gamespot. I was going to post it here as you did but i knew i would get ignorant answers. Truth is if there was a survey to come out tomorrow and say 100% of 360 failed.. The people who have had a bad experience or just hate MS will believe it.

 

[Raesvelg]

Actually, you are going to have to spell out how that creates a bias in the poll, because that doesn't make any sense at all.

It's pre-screening for a particular behavior, namely the purchasing of after-market warranties. Hence why their sample size for the Wii was substantially larger than their sample size for the other two consoles.

Get it now?

Aside from that, by it's very nature the survey (study technically) is flawed. It tracks a period of time well after the launch of the 360 (2nd quarter 2007 to first quarter 2009), and ultimately all it will really tell you is the percentage of warranty purchasers who elected to use their extended add-on warranty rather than the generous 3-year warranty that Microsoft enacted after the 360's hardware issues became an embarrassment to them.

I'm not saying that the survey is not truthful, I'm simply saying that its not that incredibly accurate either. In order to get truly accurate numbers... well, you'd have to get Microsoft to talk (they won't) and add in numbers from independent warranty corporations.

It's largely a moot point at this rate anyway. Microsoft has obviously resolved the majority of the hardware issues with the 360 by now, of course. The newest models are reasonably solid machines.

 

[Bajaratt]

It's pre-screening for a particular behavior, namely the purchasing of after-market warranties. Hence why their sample size for the Wii was substantially larger than their sample size for the other two consoles.

Yes, it's only getting results from the people who buy warranties on their consoles; but what you're not explaining is how that creates a bias in the results. It's like only asking people who eat wheat bread what they think of the government(reworking your bad example); the actual behavior is irrelevant to the results.

There is no relevant bias in the data, since the people who buy extended warranties are pretty diverse. Those who buy one also use them; so you have more accurate results versus people who the console broke and they just bought a new one or bought a PS3 instead of having it fixed.

It's largely a moot point at this rate anyway. Microsoft has obviously resolved the majority of the hardware issues with the 360 by now, of course. The newest models are reasonably solid machines.

This is mainly what I was driving at, that they had been tracking the failure rates for the different versions of the hardware and that they had been getting more and more stable; proving that the 54% number was completely wrong.

 

[Raesvelg]

Yes, it's only getting results from the people who buy warranties on their consoles; but what you're not explaining is how that creates a bias in the results. It's like only asking people who eat wheat bread what they think of the government(reworking your bad example); the actual behavior is irrelevant to the results.

Any time you select a group based on a behavior, you introduce bias into a study. Sometimes, that bias is as simple as selecting a group based on their willingness to answer a poll. The range of bias can be subtle, obviously, but it exists nonetheless.

For example, people who only eat wheat bread are obviously more health-conscious than other people. More health-conscious implies a greater level of education. A greater level of education will yield a difference in any opinion poll.

In this instance, you're looking at people who are more willing to make an investment to secure the extended operation of their product. That would tend to pre-select more for people who will actually do things like pay attention to the manual when it tells you how to properly set up the console, people who will keep it perhaps cleaner than someone else, etc.

Except for Wii consoles. Pretty sure that was all just random parents getting talked into it at Best Buy. FIVE TIMES THE SAMPLE SIZE.

This is mainly what I was driving at, that they had been tracking the failure rates for the different versions of the hardware and that they had been getting more and more stable; proving that the 54% number was completely wrong.

Will be completely wrong is the phrase you're actually looking for. Or possibly is now. It's generally conceded by most people, I think, that the latest mark of the hardware (4th or 5th version, whatever it is) is pretty solid. But I don't think you can blame people for being leery of the console's history.

Microsoft openly threw a billion dollars at the problem, because they knew that the original hardware was shite. I have to expect that they knew it was shite when they put it out originally, and did it anyway hoping that it wasn't as bad as they thought, just to steal a march on Sony.

The fact that the hardware is decent now does not mean that it was not bad then.

 

[Bajaratt]

Any time you select a group based on a behavior, you introduce bias into a study. Sometimes, that bias is as simple as selecting a group based on their willingness to answer a poll. The range of bias can be subtle, obviously, but it exists nonetheless.

For example, people who only eat wheat bread are obviously more health-conscious than other people. More health-conscious implies a greater level of education. A greater level of education will yield a difference in any opinion poll.

The problem is it's not a poll, I know I keep saying it is, but it's not. The user's opinion's, backgrounds, behaviors are completely irrelevant.

In this instance, you're looking at people who are more willing to make an investment to secure the extended operation of their product. That would tend to pre-select more for people who will actually do things like pay attention to the manual when it tells you how to properly set up the console, people who will keep it perhaps cleaner than someone else, etc.

That's a big assumption, I could make a case for the exact opposite type of behavior. If you know they'll fix it or send you a new one if it breaks, you could be less likely to actually care about making sure it's clean/dust-free, etc. If you know you're going to be taken care of, that your console will be fixed, you can be less likely to try to take good care of the console; if it breaks you can just use the warranty and get it fixed, right?

Will be completely wrong is the phrase you're actually looking for. Or possibly is now. It's generally conceded by most people, I think, that the latest mark of the hardware (4th or 5th version, whatever it is) is pretty solid. But I don't think you can blame people for being leery of the console's history.

Except that there's still a bunch of people who think that the console fails just as much as ever. Just look in this thread for proof of that.

The fact that the hardware is decent now does not mean that it was not bad then.

And no one is disputing that fact, all that I've said is that the 54% error rate is ridiculous, because of all of the hardware advancements.

 

[masterorion]

54% may sound rediculous. i agree that i may be a bit far fetched. however, you still cant prove against it with hard facts. no number will be rock solid unless microsoft comes out and gives a 100% factual number of broken consoles from release date till today. which, i can pretty much assure you that no company will ever do. so until then, i would assume there is a failure rate ranging from 20-35% based on all of the stats that have been floating around. on those numbers, how many of those are REPEAT problems?

from that, you have to assume that numbers given, depending on where the stats are coming from, are mostly correct. still, release xbox consoles gave MS a bad rep. they are lemons. the only way to get past them is to have the company see the flaw and fix it in later releases. this happens often auto companies either through warranty or later productions.
in all honesty, i forsee problems with the ps3 slim. if i were a new ps3 customer, i would not buy a slim until 6 months down the road, at least.

 

[Elemmire]

people don't seem to realize most of Microsoft products are terribad.. specially upon release.. they almost NEVER test their products. I'm amazed Window 7 has the quality it does lol. It's like it was never made by Microsoft. Microsoft does eventually make the product of decent quality but only later in the life.

360 is a fine example of this. Shitty quality at first but the newer ones HAVE been improving in reliability.