1. They didn't see the recession coming.

2. They thought HDTVs would pickup faster.

3. The fanboys claimed blu-ray would explode once HD-DVD was out of the way, and Sony actually thought those people represented the mainstream (although the inverse was just as unlikely).

To say that using revenue as a measure of marketshare is stupid is very ignorant. Marketshare is used to try and determine a demand for a certain product in an industry. If you understand economics you would know that there is a demand curve. The higher the price the less the demand along a demand curve. However it is very possible that a product may be selling less at a higher price at a given point in the products life but has a demand curve that is greater(more shifted to the right) than the demand curve of a product that is selling more at a lower price point. The only way to try and determine who's demand curve is greater is by the measurement of revenue produced by those products. Raw sales numbers would tell you very little if the two prices are dramatically different(10% or higher difference).

The reason raw numbers are used in the video game industry is because the number of consoles sold is an indication of how many people can buy a video game on a particular console. And that is the type of information that is important to a publisher of video games. Measuring revenue may be important for the actual console makers but to outside companies, raw numbers are more important because that indicates how many people they can sale their products to.

As for the market share of those actual product, software sales, raw numbers are used because in general in the video game industry most video games have been retailed at the same price and would usually see a similar price decrease over time thus revenue and raw numbers would be at pretty equal ratios. However this may not be true anymore as wii games are cheaper than HD games. It would make sense now that video games should be tracked in terms of revenue to get and understanding of the actual demand curves which is the best way of determining demand for a product not just how many total sales it's had while ignoring price. The only problem with this is it may be too hard to determine which games are sold at which price because of all the different sales and coupons and discounts given out on software sales from retailers. And therefore raw numbers will continue to be used for trying to measuring software demand though using revenue would be a much more accurate measurement.

Demand curve? That seems like some marketing term to spin numbers.

If you sold 10 units last year and then sold 1000 units this year. You've increased 100 folds of sales. In percentage, 1000% increase.

If you sold 1,000,000 units last year and then sold 1,500,000 this year. You've increased sales by 50%. Crappy sales, aren't it?

@lord It's actually an economics term. Though my degree is in marketing...

And it's actually a more wholistic view of the numbers since it encompasses both price and total units sold. 

Okay, but concept of trying to measure a demand by curve? That seems really hard to gauge.

The demand curve is measured by the sales at each price point. Obviously you can't assume all things are equal like you do in theory as there are many external forces that can't be controlled, but it can be plotted. You can find the demand curve for the 360 with the sales of each SKU as it gets discounted. You can even attempt to look at the sales of the different SKU's and compare them but that would not be as accurate as the different SKU's offer different value. Many times products are put on sale just to see how well they sell at a certain point to get an understanding of the demand curve. Then the company can be better equipped to decide at what price point would they maximize profit.

Okay, but it's still not as good a curve as all those companies predicted.

The prediction you are talking about was for Japan and it was more or less correct.

I think it's misleading when only the top 20 titles are measured. There are tens of thousands of titles that are not in the top 20.