This is the first of a three part series detailing core ideas JPG has been exploring and building towards this year.

Every 6 months or so, a tweet enters the void to the tune of 20 or so likes declaring that “20XX is the year of the TCR.” I hope those 20 people see this, because I think there’s finally a compelling case for it.

If you were in the crypto ecosystem around 2017 and 2018, you probably remember TCRs (token-curated registries) as a pervasive meme on par with Harberger taxes and bonding curves. TCRs are a subset of curation markets, systems aimed at incentivizing the reduction of information asymmetry. In their most simple original construction, TCRs are lists with each entry moderated by a form of economic governance. As an example, Alice stakes 100 $Token to submit an entry to the registry, Bob disagrees and challenges the proposal with 110 $Token, and it is sent to arbitration by the rest of the community, with voters and the winning party receiving the loser’s stake.

There were several interesting experiments in that era around this idea including AdChain (registry for high value digital ad space), FOAM (decentralized and curated maps app), and TCRParty (registry of top CT influencers for retweeting). These were the days in which security tokens were receiving more attention than DeFi, and so this was viewed as an extremely promising application segment. After all, token-curated registries feel extremely blockchain-native: economically safeguarded information by a consensus of stakeholders.

In true crypto fashion, many game theory blog posts were published about proper incentive structure, the need for reputation vs. pure economic structures, the level of objectivity vs. subjectivity involved in various registries and its implications. Reading this in 2022, it is obvious that crypto participants can be (optimistically) expected to make only one game theoretically optimal decision per month, so it is not entirely surprising TCRs did not take off in a compelling way.

And yet, the promise of TCRs remains alluring. So much revenue and value on the internet (and IRL) is based upon curation, from newsletters and instagram feeds to entire apps based on product recommendations and reviews, not to mention endeavors like Wikipedia. It’s clear there is a nut to crack here for web3, one distinct from decentralized social apps we hear much more about.

• Most of the information being curated was off-chain, and there were better off-chain sources than what was created on-chain. If you are going to add friction to an existing type of application, you better make the outcome correspondingly valuable to that friction. Most of the initial TCR applications (built or theorized) often revolved around curated off-chain information, which was then limited in its composability. Additionally, most of these information segments already had highly usable indexes, leading to massive effort necessary to even reach parity.

• The protocols were built with economically rational actors in mind, and the economic assumptions were often weak. One of the problems with creating a TCR based upon economic assumptions is that the value of participation (productive or malicious) is difficult to ascertain. Users are, however, forced to make such judgements as the basis of participation. The nature of TCRs is such that the control of information is not necessarily connected to that information’s usage and the value reliant upon it. If a TCR serves as a list of verified oracles for a lending protocol for instance, the value of manipulating the TCR is related to funds in the lending protocol, not the value of the TCR governance token.

• Governance apathy hits hard. This is obviously a well known issue within crypto, but it is highly exacerbated in historical TCRs largely due to the factors above. Curating a list of already existing information can feel futile, and there’s little personal or global value in it. On top of that fact, TCRs bring in economic calculations and the risk of loss. There’s the potential for reward, of course, but you must believe in the project’s long term viability. In the end, governing a TCR feels a lot like going to work.

• It’s difficult to monetize publicly available data, until the protocol is viewed as the source of truth for such data. Finally, there’s the question of what’s it all for. How does one even generate value with a TCR? The data is public, so you can’t effectively charge for access to it. Instead you must charge for inclusion into it. This is only viable once the registry has become an accepted source of truth, allowing for inclusion to be a necessary step in marketing and discovery.

All of these problems are interrelated. Choosing the wrong data set to curate aggravates issues around economic governance, which leads to greater governance apathy and difficulty in traction necessary to monetize. This makes economic governance even more difficult and an anti-flywheel is created, which brings data set completion forever out of reach.

• Curate information that currently lives on-chain, and which has not been properly curated. By filling an existing void with the potential to be utilized by other builders, the base of an economically valuable TCR is laid. Each new entry is new value created, as opposed to a system in which new value is created only once parity with off-chain sources is reached.

• Base the system upon reputational reward in a high status and high passion arena. Rather than playing money games, play status games. Rather than making each decision a rational P/L calculation, create an incentive system around being knowledgeable in an arena people are desperate to prove they are knowledgeable.

• Gamify governance and make it rewarding in and of itself, as opposed to a spreadsheet maintenance task. Shift what governance feels like. It’s not going to work, receiving your maintenance payout, it’s playing a knowledge and reputation game. In other words, leverage intrinsic rewards rather than extrinsic. Once again, because ideally the information is in some way new, it also becomes beneficial to participate simply for the exposure to freshly generated information.

• Simply become the source of truth for such data. This one is kind of hard to get around. A registry needs to be the source of truth for its data set. One caveat is that a registry can be a source of truth rather than the source of truth and still generate interesting value if the crowd that is sourcing and viewing the information is large and influential.

By shifting the dataset and reward systems above, we see a more traditional flywheel emerge. Each new entry to the dataset feels valuable, so people are eager to participate and have their contribution recorded. The dataset then becomes more valuable, leading more users to want to contribute to its creation and their own reputation, until eventually the data set is near-complete and viewed as a primary source of information. Anytime new data is created, the creator of the data then needs to be included on the list to ensure discovery.

There are, of course, obstacles to  such an approach, most notably including optimal reputation system design and susceptibility to adversarial and sybil environments, but these are not insurmountable, simply challenging. We believe starting with the right building blocks will allow for the emergence of viable token-curated registries, upon which many iterations and optimizations may emerge.

If this type of work is of interest to you, please reach out in the Discord with questions, feedback, and thoughts.

Additional TCR Resources:

• Megadump of Info: https://github.com/miguelmota/awesome-token-curated-registries

• Basic Design Patterns: https://hackernoon.com/token-curated-registry-tcr-design-patterns-4de6d18efa15?source=placement_card_footer_grid---------1-43

• Some Use Cases: https://medium.com/paratii/token-curated-playlists-2-throughput-fitness-use-cases-ux-6421733b0063

• Need for Reputation: https://blog.relevant.community/tcr-design-flaws-why-blockchain-needs-reputation-c5771d97b210

• Subjective vs Objective: https://medium.com/coinmonks/subjective-vs-objective-tcrs-a21f5d848553

• FOAM as Case Study: https://blog.foam.space/foam-token-curated-registries-for-geographic-points-of-interest-60d3c043f183

• Simple Grading Mechanism: https://medium.com/coinmonks/graded-token-curated-decisions-with-up-downvoting-designing-cryptoeconomic-ranking-and-2ce7c000bb51