Count how many times you stepped outside the frame last year

This is the last one in the F2P Mobile and Premium PC/Console series and the subject is the creative side, the part product people do not carry. People who come out of creative processes and people who come out of iterative processes should not be mixed together, and there are two big examples that show this is not a problem specific to mobile F2P.

The word creative has long since been turned into something else in mobile F2P. Putting out fifty creatives a week or running A/B tests counts as creative work, when that is iteration itself.

Amazon Game Studios was run for eight years by an executive who had never made a game before and came from the company’s books side. They tried to make games the way Amazon does business and what the player wanted to play came second. Their first AAA game, Crucible, came out on 20 May 2020, was pulled back into closed beta a month and a half later, and was cancelled in October.

The real issue is the games they already had in hand. New World opened on Steam with 913k CCU, its peak last month was 722. They published Lost Ark in the West, the game climbed to 1.32M CCU, last month it was at 10k. One of them shook the MMO world, the other was already a very strong game, and both melted under the same hand.

The end came anyway. In the wave of cuts in October 2025 Amazon largely stopped first party AAA development, especially on the MMO side. In January 2026 the person running the studio left as well and the focus shifted to Luna. New World’s servers shut down completely at the end of January 2027.

At Sony it ran the other way. A target of 12 live service games was loaded on top of studios that make premium games, and Bungie was brought in to transfer the know-how of the iterative side. That was how the creative side was pressed down. The Last of Us Online was cancelled after 4 years and a team of hundreds, and Naughty Dog gave as its reason that it did not want to turn into a fully live service studio.

Most of that list either never came out or shut down as soon as it did. Concord sold 25k copies at $40, shut down two weeks later, and its studio shut down too. The same thing is coming for the Horizon games soon.

So what was the know-how Sony bought? Not making games, but getting more out of the same game. In 2020 Bungie brought in the Destiny Content Vault and removed campaigns, modes and zones that players had paid for out of the game, and sunset a lot of fan favourite weapons. They could take content back the same way they gave it.

In this machine the game doing well does not protect anyone either. Less than two months after The Final Shape, one of the most successful expansions in the series, came out, 220 people were let go, 17% of the team. The game had beaten every internal projection, but this was the third round of cuts since the Sony acquisition. This was the model they wanted to teach creative studios.

The machine has no bottom either, it keeps squeezing and in the end it squeezes itself. This May Sony wrote off $765M for Bungie, in June Destiny 2’s last update shipped and content support ended, and two weeks after that about half the studio was gone, 292 people. The side that was supposed to teach the know-how finished itself off while it was trying to teach it. Destiny is a game I buried 6k hours in, and after 10 years it still has not realised its own potential, a wasted game. It had incredible IP value. Somebody please make Destiny 3.

Neither Amazon nor Sony is a mobile F2P company, in fact neither of them was even doing PC F2P. Both had the money, the team and the IP. What was missing was the side that decides what to make, and that side cannot be bought. There is no mobile F2P problem here, there is a problem with the iterative model, and the platform makes no difference at all.

The reverse of this exists too, and it is the premium side’s own disease. Someone who never learned the iterative side builds his game for years without showing it to a single player, never measures appeal even once, and says “if I make it good they will come.” Nobody can predict what will happen at launch either, because they have looked at nothing that would let them predict it. In Concord both sides’ diseases stacked on top of each other.

And creativity is not what mobile F2P defines it as, making the pig a king and making the UI purple. It is how often you go for new and wild ideas and how much thinking you do at the edges. When you are tied to data alone those edges stay inside the frame and the work becomes iterative. But if you go and try to combine a deckbuilder with a soulslike, you are creative, and if on top of that you collect data on it and think it through, you are set.

HC, hyper-casual, has to be separated out here as well. It was the place in mobile F2P where people stepped outside the frame the most, because the cost of testing an idea was close to zero, genuinely absurd things could be tried and were. The industry then tried to turn that muscle itself into iteration, but the muscle stayed with the person doing the work, and that is why teams coming out of HC can get attention on the premium side with their first move.

So the creative side is not something that cannot be learned, it is that it cannot be learned inside the loop of either mobile F2P or PC F2P. Every reward calls you back inside the frame, and every time you step outside it you take a penalty. Someone who stays in that loop for years ends up never using that side at all.

The AI side has to go in here too, because it is going to sharpen this split even further in the coming period. In F2P the real moat was never the game itself, it was the machine. Putting out fifty creatives a week, testing them, reading the number that comes back and setting up the next one. AI makes exactly those three steps cheap.

When everyone can iterate at zero cost, iterating fast stops being an advantage. So the side AI devalues is not the creative side, it is the iterative side. The product person becoming a game designer in 6 months will drop to 6 weeks from here on, but the title will lose its value at the same speed. The one decision AI cannot make in your place is the first one, what you are going to make.

What I carried over from mobile was being able to look at games as pure business, and it worked exactly the same way on Metal Thunder. In mobile too, whenever I went in the creative direction we started finding games again, and when I changed publisher I dragged every game people called dead into soft launch. People with a real game design pedigree blend the creative and the iterative easily anyway.

I came back to PC because the state of mobile had used me up and because I wanted to feel something again, and between people who do not know this job and a pile of personal problems landing on top of each other, I went through an unbelievable burnout. Some of my games still get 20-30k players a month on the store, completely random and organic, and I cannot even find the strength in me to update them.

Ask yourself one question, how many times in the last year did you try to step outside the frame. If the answer is zero you are on the iterative side, and that side is getting cheaper in the coming period. If there is an answer, collect data on top of those attempts, being creative does not mean being an enemy of data. The road to being a good developer runs through blending the two.

At the end of the day the player plays to have fun and it makes no difference at all what the platform’s dynamics are. Throw as much money as you like on any platform, nobody plays a game that gives them no pleasure and it fails.

The point is not choosing between iteration and starting from scratch, it is deciding which of them you build on. The iterative side can be learned and is going to be learned anyway, the creative side only grows as long as you feed it.