At the time many things are obscured by "the fog of war," however, in publicly traded businesses the wrong course is perfectly obvious after the fact by examining the result as measured by the return on investment in the stock.
If you wish to laud Microsoft's choices, carry on without me.
Microsoft's fundamental business problem is that they are on the wrong side of some disruptive innovations. And that is a very difficult business problem to face. No matter how competent the management, if a company is caught in that trap it is normal to see poor returns in the stock market. Read _The Innovator's Dilemma_ and _The Innovator's Solution_ for more on that.
Microsoft's problems are made more difficult by the fact that Bill Gates stepped down and left Steve Ballmer in charge. They went from having one of the most effective CEOs out there to someone who I think is a liability.
However, despite those factors, it is still possible to look at specific decisions and have opinions on how good or bad they were. And I think that Bill Gates' decision to start the Xbox was good.
I think it is possible and laudable to make up your own mind about specific business decisions. However, it isn't really possible for us to argue the point unless we can agree on a metric for measuring their efficacy.
From an empirical perspective, a theory ("XBox is a good decision") is only meaningful if it is falsifiable. I only have two metrics in my toolbox for starting a new line of business: The ROI of the specific business and the ROI on the stock overall.
However, not all discussions neatly map to empirical methods, and even then there are other metrics besides those two ROIs. If you are saying that there is some validity to this discussion that can't be measured quite so neatly, if there are these "strategic" considerations that require a leap of faith to appreciate, well I can't argue with you.
I'm not trying deflate your argument by suggesting it's faith-based. Most corporations make a lot of decisions that aren't falsifiable. Executives stand up all the time and say, "it looks bad but it could have been worse," and it's very difficult to naysay them because we can't run A/B testing and find out whether Microsoft-Xbox does better or worse than Microsoft+Xbox.
So... In the spirit of Christmas I will wish you well and thank you for sharing your opinions in a reasoned tone.
In the case of the Xbox, Microsoft did both. It spent money developing the Xbox. It spent money investing in Windows. It also took lessons from the Xbox to make Windows itself a better gaming platform. And finally the cost of the Xbox approach was not as much as you'd think, because the division stood to recoup much of its costs.
How much did this cost Microsoft in the end? Less than you'd think, given the article. While it is true that costs were measured in billions and profits in millions, those costs were fixed development costs in the low billions, and profits recently have been quarterly profits in the hundreds of millions. I believe that at this point, over all generations of the Xbox, Microsoft is in the black. (But if they had invested their money differently, they could be further in the black.)
So what was the result in the end? I think they sliced something like a quarter off of Sony's profit margin. Almost for free. And Sony failed to execute on their strategy of having TVs replace PCs in the home.
Whatever you think of the merits of the strategy, Microsoft clearly thinks it works. They have used it repeatedly. For example look at Bing. And I wonder whether Google TV is an attempt to return the favor.