That's a fair point. IMO though, Windows Mobile is superseded by Windows Phone but this is all semantics, the point still stands because, as I have said, the actual change or dichotomy of the product is irrelevant to the market performance statistics.


The point is that a 0 to x number is a very poor way of trying to show a trend, where as basing it on a company's past performance or over an actual period in that sector (whether you choose to compound Phone/Mobile or not) is far more sound, far more reliable.

You don't even need to look at windows mobile, because if you look at windows phone over an actual period rather than a 0 to x you will still see the actual trend (which is still decreasing). Compounding windows mobile and windows phone will inflate Microsofts marketshare by shifting the graph up as a whole but will not change the trend you see for WP into a negative unless it was actually negative. Splitting it can show a negative as an increase if all you do is look at a 0 to x number instead because 0 to x will always show an increase, you cannot have a negative market share. if Windows Phone was performing well it should have shown a quarterly increase in the compounded graph, not still show a quarterly decrease over an actual period. I will reiterate the point, what you cannot do is look at a 0 to x number and say that it is now more popular or better performing as a product, a different product or otherwise, a separate entity or not, if you could then you can even change the trend of the exact same product by deciding to split it (arbitrarily or not). Some companies actually do this and some people actually fall for it.