In its fourth year, the GameHorizon conference in Gateshead, UK, is something of a boutique gathering.

Mixing up strong input from the north east English development community with a good selection of international speakers and a broad UK audience, it doesn't take itself too seriously.

But with a single - and fairly interactive - track (encouraged by the irreverent style of MC Square Enix's Ian Livingstone, it's a great place to chew the fat.

So what did we learn?

1. Free to go

Kicking off the conference, Zynga's Louis Castle demonstrated the power of free-to-play games, highlighting his own Damascus Road experience when moving from EA to the Facebook publishing giant.

"Games want to be free," he opinated, pointed out that giving away content means you can rely on your audience to do much of your marketing activity for free.

After all, you don't need to worry about discovery or targeted ad spend when you're gearing people up to encourage their friends to join in with the fun.

Yet even for companies nothing like the scale of Zynga, the ability for people to play a game for free is only becoming more powerful. It undercuts any issues of piracy, enables frictionless access and, if nothing else, provides you with a committed audience that can be used for other purposes.

Backing this up, during the conference, small US developer Nimblebit revealed that in five days, its free-to-play game Tiny Tower had been downloaded over one million times, generating 18 million sessions and 319 years of playtime, with zero marketing budget.

2. Power of paid

But where there's a ying, that's a yang, and often it's Epic Game's Mark Rein.

Despite the announcement of the free-to-play Infinity Blade X in Japan, Rein said if the $5.99 Infinity Blade had been pure freemium, instead of premium plus IAP, it would have required 30 times the number of downloads to generate the $11.2 million of net revenue it's generated to-date - something he didn't think it would have done.

From a totally different market, similar thoughts were voiced by CCP CEO Hilmar Veigar Petursson. Talking about the success of PC subscription game Eve Online, he said, "Free is just a funnel to get a very large audience and then sees who sticks around and will pay."

The point in these cases is that a certain audience is happy to pay for very high quality content, and you need a very large audience (at least ranging from 20 to 100 times on a 5 to 1 percent IAP conversion rate) to provide equivalent revenue.

Equally, what's significant is a large minority of players who are happy to spend money buying or subscribing to a game, are also happy to buy in-game items; in the case of Infinity Blade its currency, while for Eve Online the $65 monocle, which offers no character benefits beyond the aesthetic, has proved popular.

3. Which side is your bread buttered?

The conclusion to draw from these two points is only certain games will work as free-to-play. Certainly, at the moment, the types of games originally released as triple-A or freemium are very different in terms of their gameplay.

As Louis Castle pointed out, the games that Zynga releases on Facebook are updated at least weekly, with junior team members given the responsibility to make key decisions quickly. Helped by real-time metrics, Facebook games are constantly being tweaked in terms of their compulsion loops, not to mention fixing bugs and adding content.

"If your Facebook game feels complete, you should have released it sooner," Castle warned, underlining the minimum viable product mantra that’s become gospel.

It’s the exact opposite of the 'It's done when it's done' model of triple-A console developers however. Indeed, the 'By Gamers, For Gamers' mentality seems to be creeping back into development: fine if your audience is wannabe game developers, we reckon.

4. In a mirror, darkly

This difference in approach was further highlighted by the talk given by game theorist Tom Chatfield, who suggested a formal language to define the experience of playing games (and with toys) is required, or at least would widen and improve our current interaction with them.

Yet, in another way, the metrics-driven gameplay iteration loop of casual games already provides this, just not in the manner or for the types of games envisaged by purists such as Chatfield.

Moderated by revenue as a base outcome, there's little argument that people who play Facebook game aren't enjoying them. And as Louis Castle - again - pointed out, the lack of friction in that environment means that if you abuse your customers, large numbers of them will quickly stop playing, and paying for, your games.

And, after all, cashflow is a formal language. It’s called accounting.

5. Google doesn't care or can't care

It's become a truism that Android is a "terrible, dismal platform for paid content" - quote from Exent's Rick Marazzani. Despite its huge install base, it's an ecosystem that's still impossible to monetise.

But, everyone hopes and expects Google will eventually force though enough changes in terms of widespread carrier billing, raised download limits, options to deal with fragmentation etc that everyone will stop moaning and start making some money.

Conversely, the more people agree about the parlous current state of affairs, and the longer they wait for substantive positive changes, the more we're starting to think Google either doesn't care or can't implement what are fairly obvious improvements to its mobile ecosystem.

More than two years on from Android's launch, it's becoming harder to judge which of these states is more worrying for the platform's longterm success in terms of content availability.

Still, if nothing else, it's good news for Nokia, and it's not had much of that recently.