Play-to-earn

What the phrase actually promised

Games have paid players in currency since the first dungeon dropped gold. Play-to-earn added one thing: an exchange rate. Once the reward currency listed, an hour of play had a dollar value that anyone could look up, and the game stopped competing with other games for attention and started competing with jobs for hours.

Naavik saw the consequence before the category did. Their November 2021 deep dive opened by stating that Sky Mavis' mission was more about spurring economic opportunity through play than about building a great game, and that players engaged with it more as a job than for fun, often as scholars from developing countries who borrowed in-game assets from wealthier managers in exchange for a share of earnings.¹ That is a description of a labour market, and labour markets need an employer with revenue.

Who was actually paying the wage

In the Axie structure the earning currency, SLP, was minted by play and consumed only by breeding. The dollars that scholars withdrew came from whoever bought that SLP on the open market, and the largest source of that demand was people entering the game and needing currency to breed their way in. Naavik's 2021 conclusion followed directly: the value of new Axies and SLP was propped up by new players putting fresh money into the game, and if new player growth diminished it could send the economy into a recession.¹

That report also flagged the second-order version, which teams miss more often. Even if daily active users keep growing, if the pace of new monthly additions declines, total monthly revenue can decline with it.¹ A play-to-earn economy is sensitive not to the level of growth but to its derivative, and almost nobody models it that way.

The dated record of what happened

DappRadar and the Blockchain Game Alliance published the on-chain numbers in April 2022. Axie's activity peaked in January 2022 at over 55,000 daily unique active wallets on average and had fallen to almost 22,000 by March. Their report attributes the start of the decline to mid-February, when Sky Mavis cut the SLP reward ratio to make the in-game currency sustainable over the longer term.²

That sequence matters more than the headline. The Ronin bridge exploit happened on 29 March 2022, roughly six weeks after the on-chain decline began, and the same report states plainly that the plunge in activity was not wholly a result of the attack because it was already in decline.² The hack is the story people remember. The emissions cut is the story that explains the design.

Read the cut on its own terms and it was correct. The team reduced a faucet that was oversupplying its currency. The problem is that in a play-to-earn economy the faucet rate is the wage, so responsible economic management and a pay cut are the same action, and the players respond to the pay cut.

What the players did next

The academic record complicates the usual telling. A 2025 study by Jordan Ali and Gili Vidan of Cornell, published in Big Data and Society and based on analysis of Axie community posts, found that most players who stayed through the downturn were not passive victims but active strategists managing their position through it.⁴

That is worth holding onto when designing. The population was reading emissions changes, adjusting behaviour and reallocating effort, which means a design team that treats its earning cohort as unsophisticated will be modelled against and beaten. Any reward change you make will be anticipated, front run and arbitraged by people whose income depends on it.

What survived the term

The label is largely retired and the underlying mechanic is not. Naavik and Delphi Digital's January 2023 review described a widespread focus on player earnings, expressed in the term play to earn, taking precedence over fun and sustainability, and led to a financial bubble, with the reset that followed pushing builders back toward games and long-term economies.³ Ownership of items, tradeable assets and revenue sharing all persisted. Earning as the reason to play did not.

The design rule we apply is a single sentence: token rewards are a retention mechanism layered on a game that is worth playing, not an acquisition channel. Acquisition economics work right up until arrivals slow, and then they invert, because the same mechanism that paid people to join now requires people to join in order to pay.

Practically that means capping what a player can extract per hour, funding the earn pool from external revenue rather than from issuance wherever possible, and testing whether retention holds when the payout is set to zero. If it does not, the token was the product. This page is design reference and not investment advice.