Let's count Newegg customer reviews of current SSDs - Who will help?

Discussion in 'Hardware Components and Aftermarket Upgrades' started by Phil, Apr 25, 2011.

Thread Status:
Not open for further replies.
  1. MidnightSun

    MidnightSun Emodicon

    Reputations:
    6,668
    Messages:
    8,224
    Likes Received:
    231
    Trophy Points:
    231
    That's probably orders of magnitude higher than the actual fail rate, since this study has what is called "response bias" in statistics: it's been proven before that users whose products do not function up to expectations are many times more likely to post feedback, whereas those whose products work perfectly are much less likely to take the time to contribute.

    I'm a good example. I have perfectly functioning Intel X25-M and Samsung 470 Series SSDs, however, I've never posted a review before.

    So, bottom line: you can't use the absolute percentages as a meaningful number, because it's not. But, if you compare to other SSDs with a similar time on the market, similar number of reviews, and similar category, then the numbers could potentially give an idea of relative reliability.
     
  2. Peon

    Peon Notebook Virtuoso

    Reputations:
    406
    Messages:
    2,007
    Likes Received:
    128
    Trophy Points:
    81
    I hate to go against the grain, but what is the current hypothesis and what conclusions are we trying to draw?

    There's a good reason why scientific papers contain more than just tables of data. If we're blindly collect data without knowing what to look for, we'll only end up with a difficult to interpret sea of information from which just about any conclusion can be drawn. That's probably what some of the posters have an issue with.

    Perhaps somebody well versed in data mining or business management could start us off by asking some intriguing questions to focus and guide this study?
     
  3. Mr_Mysterious

    Mr_Mysterious Like...duuuuuude

    Reputations:
    1,552
    Messages:
    2,383
    Likes Received:
    15
    Trophy Points:
    56
    The only way to get truly reliable data is to buy 1000's...preferably 10,000 SSDs per brand. And test them out over 10 years in a variety of scenarios.

    Then record the results, do some basic statistical analysis and post the results.

    Anyone want to volunteer? :D

    Mr. Mysterious
     
  4. Phil

    Phil Retired

    Reputations:
    4,415
    Messages:
    17,036
    Likes Received:
    0
    Trophy Points:
    455
    When I click on the 1 star reviews here: http://www.newegg.com/Product/Produ...rue&Keywords=(keywords)&Page=1#scrollFullInfo

    I don't see any 1 star review. According to Newegg there should be 5! Anyone know what's going on.

    It could be interesting but I don't think it will be necessary. There are all kinds of limitations in these numbers and I think most people realize that.

    However, it is possible to draw the following conclusions:

    -Sandforce 1200 drives like Vertex 2 and Corsair Force have significantly higher failure rates than more reliable drives like Crucial C300 or Samsung 470.

    -Intel X25m was often regarded as the most reliable drive of all. These data show that Samsung 470 and Crucial C300 may be at least as reliable or maybe even more reliable.

    I advise a lot of people on which SSD to buy. In the past I have often recommended Sandforce 1200 drives. I won't be doing that anymore.
    I've been saying 'if you want the most reliable get the Intel X25m'. I won't be saying that anymore.
     
  5. davepermen

    davepermen Notebook Nobel Laureate

    Reputations:
    2,972
    Messages:
    7,788
    Likes Received:
    0
    Trophy Points:
    205
    as it's impossible to get sales/returnrates as that's normally not public, it's the only data we have and can do statistics on. and so far, the data proves what we expected: vertex2 being massively out the line, and the others more or less the same. this combined with the issues that ocz went trough, the revisions, the lies, the firmware updates, the dramas on the big pages, etc, together all shows the same: the vertex2 is special, and not in a good way.
     
  6. Phil

    Phil Retired

    Reputations:
    4,415
    Messages:
    17,036
    Likes Received:
    0
    Trophy Points:
    455
    To add, Corsair Force is equally bad as Vertex 2, indicating that Sandforce 1200 has massive problems.
     
  7. davepermen

    davepermen Notebook Nobel Laureate

    Reputations:
    2,972
    Messages:
    7,788
    Likes Received:
    0
    Trophy Points:
    205
    indeed. still, no one was doing so much publicity stunts as ocz. their vertex series has a questionable quality (as we have further proof in this thread), so did their core line.

    i'm still impressed people still buy from them..
     
  8. Aluminum

    Aluminum Notebook Consultant

    Reputations:
    44
    Messages:
    238
    Likes Received:
    2
    Trophy Points:
    31
    The numbers compared to all shipments may be fairly inaccurate, since we all know owner reviews are very subjective and bad news posts a lot more than good news. (and especially no news and boring news, those two guys never post)

    But you would have to really put your head deep in the sand to claim that there isn't a pattern of problems with many OCZ drives compared to most other major models.

    Those posts on newegg and similar sites are not meaningless, even if they aren't from a peer reviewed double blind paper :rolleyes:
     
  9. Peon

    Peon Notebook Virtuoso

    Reputations:
    406
    Messages:
    2,007
    Likes Received:
    128
    Trophy Points:
    81
    There's a lot more to reliability than simple failures, though. If you happen to be one of the unlucky C300 owners whose hardware/software combination happens to trigger the firmware bug, I doubt you'd call the C300 a reliable drive even though it's technically functional. In that sense, Intel and Samsung are still ahead.

    Now that you mention it, I'm starting to get curious about SF-1500. I assume it'd be roughly the same as SF-1200, but that wouldn't be very scientific.

    The Vertex 2 has been the cheapest option in the super popular 60GB ($95) and 120GB ($175) space for months, probably thanks to the 25nm switch.

    Looking at 120GB models, prices for the C300 ($200) have dropped significantly in recent weeks thanks to the M4, but that's a very recent phenomenon - as recently as 2 weeks ago, a 128GB C300 cost $220-$240. The X25-M G2 ($230) remains considerably more expensive than the Vertex 2, and Corsair Force ($230) price levels have been essentially flat since the end of last year.

    Someone who's not familiar with the latest SSD market news could easily pick a Vertex 2 simply based on the price and the (no longer true) idea that all Sandforce drives are the same.

    It's a shame that the no news and boring news groups are extremely underrepresented. They're by far the most statistically interesting groups since they reflect the vast majority of users who are satisfied with their purchase, but not to the point of boasting about it. For the record, I've never had a problem with any of my SSDs and I've never written a review for any of them either, and I've had Intel, Samsung, and Sandforce drives.
     
  10. davepermen

    davepermen Notebook Nobel Laureate

    Reputations:
    2,972
    Messages:
    7,788
    Likes Received:
    0
    Trophy Points:
    205
    true that they are popular and payable. still, the amount of issues where since day one, and didn't reduce over time. popularity or not, it's scarily high.
     
Loading...
Thread Status:
Not open for further replies.

Share This Page