If you want to just see how the mobile games business is in flux, you only need to consider the third quarter earnings figures that the publicly owned companies have been turning out over the past couple of weeks to see the winners and the losers.

Leading the charge, as it has been for the whole of 2010 is Japanese mobile platform holder DeNA.

Now famous for its up to $403 million purchase of ngmoco, the publisher of Moba-ge-Town posted Q3 sales of $336 million and operating income of $169 million.

It expects to post annual sales of around $1.25 billion in 2010, compared to $500 million for 2009.

Maximising micro transactions

The massive growth hasn't come from the virtual goods it sells in its online mobile social games. Instead, it's come from opening up Moba-ge-Town as a publishing platform to thirdparty developers.

This now hosts over 150 developers and over 350 externally developed games - clearly far more than it DeNA could produced internally. And DeNA takes a cut of all sales made in these games; hence the massive growth in its financials throughout 2010.

On a different scale, but virtual goods have also been boosting revenues at Korean publisher Gamevil for the past years. Its game series such as RPG Zenonia and baseball titles such as Baseball Superstars have always made the majority of their profit on virtual items sales. Despite being slow to integrate such methods into its iPhone games, Gamevil has now done so, boosting revenues.

Its Q3 saw sales up 14 percent to $5.9 million, and net income up 11 percent to $2.8 million: a healthy margin.

A thousand cuts

However, the virtual goods games isn't as easy as that.

As was demonstrated by ngmoco prior to its DeNA takeover, it had lost at least $13.4 million experimenting with business models, building out its online platform and releasing highend freemium games.

It's a similar story over at Glu Mobile, which has been managing the transition from a fairly successful Java and Brew carrier-focused publishing model to a smartphone freemium model.

It's been posting losses for over two years and Q3 2010 was no different, with a loss of $1.6 million on revenue of $15.5 million. Indeed, it's projecting an annual loss for 2010 of over $15 million.

Glu's problem is that with smartphone games sales in Q3 only consisting of 15 percent of overall revenue, and micro transaction and in-game ads only being a quarter of that (around $600,000), it will need a massive expansion of this business in order to generate decent headline returns.

Paint it black

Yet this isn't to say that the traditional paid for business model is in decline. If you have sufficient scale or sufficient brands, it can be strongly profitable.

For example, Gameloft saw its Q3 sales rise 15 percent thanks to strong iOS and emerging markets' business to $49 million. Indeed, it says its Apple business has risen 80 percent during 2010; obviously from a small base, but significant nevertheless.

But though it's likely to post its best annual figures to date for 2010, even Gameloft will have to think seriously about its future operations following the shock news that arch rival EA Mobile saw its Q2 (same calendar period, different nomenclature) sales fall 4 percent.

It's the first time in recent memory that EA - which is the west's biggest mobile publisher - has posted a year on year decline in sales, something that points to a combination of a lack of releases and a lack of innovative business models.

During 2010, EA has relied on its console brands and largescale promotion by means of App Store price cuts rather than releasing exciting content; something its up to $29 million purchase of hyperactive App Store publisher Chillingo is expected to reverse.