Linked by Thom Holwerda on Fri 19th Oct 2012 13:56 UTC
Apple "Looking at the stats for all our apps (we have 4 universal apps with almost 3 million downloads) and for all versions after a month gives an interesting picture." Close to 70% is on iOS 6 already. In the meantime, Sony just announced it's going to leave loads of users in the cold by not upgrading their phones to Jelly Bean. Those that do have an update in the pipeline will have to wait until next year.
Permalink for comment 539231
To read all comments associated with this story, please click here.
RE[8]: Interesting but...
by gpsnoopy on Fri 19th Oct 2012 21:48 UTC in reply to "RE[7]: Interesting but..."
Member since:

No, you are not. Don't throw around words like "highly biased". Your only initial basis for such a claim is that it measured updaters so people who update are likely to update other things as well.

Their oldest post could only lead to this conclusion, giving that that they only considered the new version.

I see no bias in one source providing data from the only source available to them.

I do. It's called sampling bias. Even if it's not intentional.

You haven't shown huge variation.

They have a >16% relative error between their two posts (70% to 60%). This is quite a large variation considering the sampled populations between the two charts are highly correlated.

That's why we need a more solid source to have a better and more accurate view. Something the Google Dashboard provides (yet you seem to dismiss it).

EDIT: looking more closely at the charts and the blog dates, their absolute error between the chart is 17%, the relative error is 30%.

You have not presented any evidence, any realism, or any logic to support the claim that this data set is disparate from any other data set.

I just have. I've just shown you that there is a 16% variation within their own sampled population. Given the high correlation between their two charts, it means that a less correlated sampled population could have much larger error.

There is no reason to believe that you could likely arrive at a completely different conclusion.

Depends on the conclusion. That iOS6 adoption rate is great, especially compared to Android ICS/JB? I agree.

That there is no fragmentation at all in iOS land, I disagree. There is too much statistical noise in these couple of charts to draw any valid conclusion.

Note that none of their blog post nor the Osnews article conclude anything, they only presented facts. You brought your own conclusion.

And the point of the article is to tell us that Google is more fragmented and slower to be updated than iOS. But you won't accept this because you are living in denial. Everyone knows this and innumerable data sets show the same thing as this data: iOS is the most rapidly updated platform ever and completely crushes Android with regard to uptake and fragmentation.

I've always agreed with that the iOS6 adoption rate is great (or rather, I've not even talked about it). My posts focused and are still focusing on the sampling method of these two charts, poking fun at the inherent sampling bias.

You've however assumed from the start that I disagreed with the... err... *your* conclusion, and that therefore everything I presented was false.

Edited 2012-10-19 22:02 UTC

Reply Parent Score: 2