A recent survey of 10 million mobile gamers claimed only 2.2 per cent of the free-to-play audience spent any money at all. That's worrying - does it mean developers are deliberately designing games to cater for the minuscule minority rather than the vast majority?
"That is a flawed and misleading assumption rooted in pre-digital thinking," answered Nicholas Lovell, owner of Gamesbrief, a F2P-focused games business site. "It is based on the thinking that a person who does not pay is barely better than a pirate, an evil freeloader, who must be made to pay or else.
"F2P design assumes that the majority of people will never pay and that's OK. They will be happy to wait rather than accelerate. They will choose not to spend money on wearables or status items. They will play the core experience, not every last facet of the game.
"These freeloaders are nevertheless vital to the success of your F2P game. They provide context for the spending amongst the most committed fans. Some of them dip in and dip out, while others are massively committed in time terms, but not in money terms, and that's OK."
Think of an energy mechanic in such a game, a resource that depletes as you play, and you can either pay to refill it or wait a day for it to refill slowly for free.
"It is a retention mechanism akin to a weekly cliffhanger at the end of a broadcast TV show," reckons Lovell. "It provides a natural drop-off point for many players. Some choose to spend money to play more right now (a bit like a box set in television). Others choose to wait.
"It is natural that players will leave a game. This 'sessioning' is an important part of a game that aims to keep players playing for days, for weeks and for months.
"So no," he added, "I don't agree that it means that developers have to design games to be frustrating or obstructive for almost the entire audience. Some games do, I'm sure, but that is not a fundamental premise of F2P design."
It's also worth noting that the low 2.2 per cent figure could be to do with the 30 games Swrve's sample were playing. The people Lovell advises have higher conversion rates, perhaps because their games appeal to a smaller but more committed audience.
Also, while the sample size of 10 million people sounds like a lot, in free-to-play terms it's not, really. Divide 10 million by 30 to average 330,000 players per game - that's roughly 0.1 per cent of Candy Crush's monthly active user figure (250 million).
"So while it is a very big sample," Lovell said, "it may not be a representative sample. And broadly, this is something we expect. We expect the majority of players to spend zero. We have reduced the barriers to entry. People who don't think of themselves as gamers are now playing. It costs us very little to service each additional player (unlike in physical games) so we can accept a low conversion rate."
A separate piece of research showed a leaderboard of online games with the highest average revenue-per-user (ARPU). World of Tanks was the winner with an average spend of $4.51, reported SuperData Research, followed very closely by Team Fortress 2. Then it's Guild Wars 2 (a paid-for game), War Thunder and Planetside 2. Places five-to-10 run Combat Arms, Crossfire, Dota 2, Heroes of Newerth and League of Legends respectively.
"Wherever you are reading this, look around and estimate the wealth of the people around you"
So these, then, are the games doing it right?
Such averages are "unhelpfully misleading", Lovell explained. "Wherever you are reading this, look around and estimate the wealth of the people around you. Pretty hard to do, I reckon. And if Bill Gates suddenly walked into your vicinity, everyone around you would suddenly become a multimillionaire. On average.
"The same is true for ARPU in F2P games with a variable pricing model. With a game-as-a-product, if someone says the average price was $20, you can imagine that some people paid $25 and some people paid $15, but most people paid $20. With a F2P game, if you have a million players and an ARPU of $1 you have literally no idea if a million people paid a dollar, or one person paid one million dollars.
"The truth is," he added, "F2P is not a single way of doing business. Some games are very niche but with committed players, a large proportion of whom pay a lot of money each month. Others are massive but most people spend nothing, and those who do spend do little. Both businesses could have very similar ARPU, but very different audience sizes, conversion rates and distribution of how much the low and high spenders spend."
Whatever the business model, F2P business itself is booming, however hard it can be sometimes to swallow. And there are examples of games doing F2P in an engaging and non-offensive way - games such as Hearthstone and Team Fortress 2, for example.