Yield Guild Games (YGG) is the canonical gaming-guild DAO: a decentralized organization that pools capital to buy the NFTs a blockchain game requires, then lends those assets to players who cannot afford them and shares the in-game yield. It is the first ecosystem-directory entry whose members coordinate around playing games together rather than around a protocol, a treasury, a collectible, or a research portfolio — the gaming/guild archetype that social, venture, and collector DAOs do not cover.
YGG grew out of the Philippines during the 2020–21 pandemic, when co-founder Gabby Dizon began lending his Axie Infinity game assets to unemployed players so they could earn income they otherwise had no capital to unlock. Formalized as a DAO in 2020 (Dizon, Beryl Li, and "Owl of Moistness"), YGG became the defining institution of the play-to-earn era — and, as that era boomed and then collapsed, a live case study in what happens to a guild when the game economies it depends on stop paying out.
The scholarship model
YGG's original mechanism is the scholarship. Play-to-earn games like Axie Infinity required a player to own NFTs (a team of Axies) before they could earn — a capital barrier that priced out exactly the players in emerging markets who most wanted the income. YGG's DAO treasury bought those NFTs at scale and lent them to scholars, who played and returned a share of what they earned.
The split is three-way: the scholar keeps the majority of what they earn (commonly ~70%), and the remainder is divided between the guild and a community manager — a mentor responsible for recruiting, training, and supporting a cohort of scholars. That human layer, not the token, was the guild's real product: a recruitment-and-training pipeline that turned idle treasury NFTs into productive, income-generating players. By August 2021 YGG had over 4,700 scholars who had collectively earned more than $8.6 million.
The model's fragility is the same as its strength: scholar income is only as durable as the underlying game's token economy. When Axie's rewards token depreciated through 2022, scholar earnings fell with it, and the guild's asset-lending thesis had to broaden beyond any single game — the pressure that later pushed YGG toward being infrastructure rather than a single guild.
Token, Vaults, and SubDAOs
The YGG token is an Ethereum ERC-20 with a maximum supply of 1,000,000,000, allocated roughly 45% to community programs, 22.4% to investors, 15% to founders, 13.3% to the treasury, 2.5% to the public sale, and 1.75% to advisors. That cap is fully minted: totalSupply() on 0x25f8087ead173b73d6e8b84329989a8eea16cf73 returns exactly 1,000,000,000 × 1018, read at Ethereum block 25,730,773. Its July 2021 IDO on SushiSwap’s MISO sold out in 31 seconds, raising roughly $12.5 million. The token confers governance rights — holders can submit and vote on proposals over strategy, investments, and treasury use — on a one-token-one-vote basis, meaning voting power tracks the freely transferable, purchasable YGG balance.
YGG Vaults are the staking primitive. Each vault represents the reward stream from one or all guild revenue sources; stakers receive rewards in proportion to the tokens they stake and the revenue the assigned source generates. It is a way to route real, activity-derived yield back to token holders rather than emitting fresh inflation.
SubDAOs compartmentalize the guild. A SubDAO is a specialized branch organized around either a specific game or a specific region — game SubDAOs concentrate the operational know-how of one title's economy, while regional SubDAOs such as YGG SEA (Southeast Asia), YGG Japan, and IndiGG (India) localize recruitment, language, and community. Each can carry its own leadership and, in some cases, its own sub-token, letting local operators who understand a game or market run semi-autonomously under the main DAO.
From asset guild to Guild Protocol
YGG's most consequential turn is the move from owning the assets to issuing the reputation. With its Guild Protocol Concept Paper (September 2024), YGG reframed itself as infrastructure any guild can build on — "Onchain Guilds," which YGG calls "a web3 primitive, serving as the building blocks for guilds across the ecosystem."
An Onchain Guild has three parts: a member list verified onchain and an SBT given to members upon joining; a multi-sig treasury wallet whose earnings are "distributed evenly among guild members permissionlessly and trustlessly through a smart contract"; and activities (quests) that earn reputation. The load-bearing element is the soulbound token: YGG-issued, non-transferable badges that signal guild membership and record achievements. Since 2022, YGG has issued these achievement SBTs through its questing programs — the Guild Advancement Program (GAP) and Superquests — "to recognize and reward in-game excellence as well as community contributions."
This is the same intuition behind Vitalik Buterin’s "soulbound" proposal: some things — reputation, membership, contribution history — should not be transferable, because their whole value is that they were earned by a specific participant. YGG applies it to guild reputation. Notably, though, it stops short of applying it to governance: YGG's own votes still ride the buyable YGG token, so a member's soulbound reputation and their voting power live in two separate places.
The publishing detour and the AI-data pivot
As play-to-earn cooled, YGG tried to become a game publisher. It launched YGG Play and, on October 15, 2025, opened the YGG Play Launchpad with $LOL, the token of its first in-house game LOL Land — a browser board game on the Abstract chain in what YGG branded the "casual degen" genre (short-session games with real token stakes). LOL Land drew over 630,000 monthly users and multi-million-dollar revenue at its peak, and by YGG's own accounting the publishing arm crossed $9 million in lifetime revenue by the end of Q1 2026 across nine signed games.
It did not last. On July 6, 2026, YGG announced it was sunsetting the YGG Play business unit, citing the "10/10 crash" — the October 10, 2025 macro shock that triggered "the largest liquidation in crypto history, wiping out over $19 billion" in leveraged positions inside 24 hours — and concluding that "YGG Play simply cannot be commercially sustainable in this climate." The human cost was stated plainly in the same post: 35 people across the unit's functions, with YGG publishing a talent contact address to place them.
Unusually for a crypto wind-down, the sunset was published as a dated operational schedule and then executed on it. The sunset guide of July 8 set out the sequence: purchases and quests disabled 9 July at 3:00 PM SGT; LOL Land and Waifu Sweeper playable but unpurchasable until 31 July; the YGG Play platform, the Launchpad (app.yggplay.fun) and Community Questing (community.yggplay.fun) shut down 31 July; two third-party titles, GIGACHADBAT and Ragnarok Breaker, handed to their studios (Delabs Games and Planetarium Labs) on 1 August; and the 2026 YGG Play Summit cancelled outright. Holders were told to unstake $YGG and redeem in-game point balances before the 31 July snapshot, after which balances would be converted to $YGG and pushed to the associated wallet with no support available. The deadline held: as of this revision yggplay.fun and www.yggplay.fun both return a Cloudflare error 1000 ("DNS points to prohibited IP") — the domain's records are gone, not merely parked — and YGG's own announcement now lives on the parent domain, yieldguild.io/news.
The next chapter is the AI data economy: a B2B pipeline built on gaming datasets, aimed at the $3.9 billion AI training-dataset market YGG cites from Grand View Research, on the argument that models trained on sterile logic need records of human irrationality and that games are where people make complex split-second decisions at volume. The existing "YGG Alerts" channel was rebranded AI Alerts and pointed at remote AI-training work in the Philippines, which YGG says drew 27,000 applications in its first five days. The same post disclosed the balance sheet behind the decision: a treasury of $20.6 million at the end of Q1 2026, of which $6.2 million sat in stablecoins, T-bills and large-cap tokens, and a restructuring that "extended our operating runway to four years."
The arc is the honest lesson of the whole play-to-earn era compressed into one organization: a guild built on games that paid, that had to keep reinventing what it was — asset lender, then reputation infrastructure, then publisher, now data business — each time the ground under the previous model gave way. The durable residue is the guild network and the soulbound reputation layer; the yield-bearing thesis proved cyclical.
One proposal in five years
Everything in the section above — retiring two of the organization's own games, shutting three of its domains, releasing 35 people, cancelling its flagship event, and redirecting a $20.6 million treasury into a different industry — was announced by the company. None of it was voted on by the token holders the whitepaper describes as governing YGG.
That is checkable rather than rhetorical. YGG's official Snapshot space is yieldguild.eth — self-described "Official YGG snapshot," created 6 October 2021, with a single erc20-balance-of strategy over the YGG ERC-20 and 224 followers. Queried through Snapshot's public GraphQL hub, its proposalsCount is 1. The one proposal is "Create YGGSPL subDAO?", opened 8 November 2021 and closed 16 November 2021: 74 voters, 621,186.75 YGG in favour, zero against. That winning side is 0.062% of the token's 1,000,000,000 supply. The only other space carrying the name, yggdao.eth, has never held a proposal at all.
So a guild that issues soulbound reputation badges to tens of thousands of members has, in four and a half years, put exactly one question to a token vote — and the largest decisions in its history were not among them. This is the failure mode a proposal count never shows on a dashboard: not a contested vote lost, not a delegate cartel, and not deliberate governance minimization either, but a governance venue that simply never convened. The DAO label survived; the practice of using it did not. Read alongside ApeCoin DAO's dissolution vote and the Aragon Association winding itself up, it is a third way a DAO stops being one — and the hardest to see, because nothing ever happens.
How Caper approaches this
YGG and Caper converge on one idea and diverge on where they put it. Both treat a soulbound, non-transferable token as the honest record of participation — YGG issues them as guild badges and achievement SBTs, and Caper mints a soulbound stake token that drives exit redemption (verified against contracts/core/src/caper_dao.rs: DIVISIBILITY_MAXIMUM, non-transferable via a depositor rule locked to the DAO itself, minted one per ballot and 0.01 per XRD of gross trade value).
The difference is what that soulbound token controls. In YGG the SBT is reputation; governance power still rides the buyable YGG balance, so a large enough purchase buys votes. In Caper the earned token is decisive in the weight itself: voting power and exit share are both compute_vote_weight = (t · v) / (V · T) (in contracts/common/src/lib.rs) — where t is a member's caper-token amount, read live from their wallet when they vote — a ballot escrows nothing — and handed in when they exit, v their earned soulbound votes, V the total soulbound-vote supply, and T tokens in circulation — and exit() reuses the exact same helper before burning the vote tokens. That means the part of influence that cannot be handed from one wallet to another is folded straight into both the vote and the payout.
This is not "your bag doesn't count" — t is a multiplier, so stake still matters. It is that v is a second factor that has to be accrued — on either surface, one per ballot and 0.01 per XRD traded — and, once accrued, can never be transferred, so control cannot be assembled by buying it off the members who already hold it. YGG proved soulbound reputation is worth issuing; Caper puts it on the critical path of who actually decides.
The section above adds the other half of the lesson. A governance venue that is never convened is indistinguishable, from the outside, from one that does not exist — so Caper does not keep a separate venue at all: a caper's proposals and its treasury are the same component, and the exit right is available on any block without anyone scheduling a vote. Where YGG's members had one proposal in five years to express a view, a Caper member who disagrees does not need a proposal to act.