Distinguishing Features of Tellor's Oracle Network

  • Commentary
  • June 4, 2021

Recently, new entrants to the decentralized oracle sector, such as Tellor, have diversified users’ and applications’ options for incorporating external data on-chain. While Chainlink pioneered the original decentralized oracle model and remains a prolific data provider, there has remained the need for an application-specific approach to oracle design, highlighting the potential benefits to certain use cases of oracle centralization paired with pseudonymity, for example. In that same vein, this dispatch reviews the unique characteristics of Tellor, which adopts a more community-centric approach to decentralized oracle reporting, relative to Chainlink.

Tellor, like Chainlink, operates using Ethereum smart contracts and is not itself a Layer One blockchain. It is perhaps most differentiated from Chainlink by its socialized query selection process. That is, rather than simply matching data requesters with data providers, Tellor services only the queries that have garnered the most community support. At a high level, the process begins when a data requester submits a query to the network. Any community member may then signal their support for this query by contributing ‘tips,’ paid in the network’s native TRB token, to it. Every five minutes, the five queries to which the largest amount of tips have been contributed are selected by the protocol for data provision. Thus, node compensation is not specified by the requester but rather is determined by the active user base as a whole. In contrast, Chainlink facilitates a theoretically unbounded number of data queries, constrained only by the supply of nodes and limitations of the Ethereum network, but at the cost limited transparency of and third-party input to the oracle selection process. Chainlink allows for any user to submit a data request along with payment specifications to the protocol, for which a series of query-specific contracts is generated; any request for which there are an adequate number of qualified and willing nodes will be serviced.

Tellor’s process and mechanisms for selecting data providers is similarly collectivized. Rather than creating a marketplace that assigns data providers based on reputation and parameters set by the requesters, as Chainlink does, Tellor selects its data providers using a Proof-of-Work-esque mechanism. Data providers compete to respond to queries by performing synthetic ‘mining’ and, for each of the five selected queries, the first five miners to find the correct nonce are selected as data providers. In theory, this, combined with the stake miners are required to post, raises the cost of corruption such that it is less profitable to submit data maliciously. Interestingly, this provides an example of the true purpose of Proof-of-Work algorithms: when not paired with a Nakamoto-style consensus mechanism, PoW does not serve to achieve consensus but rather to economically disincentivize malicious and/or negligent behavior. In contrast, Chainlink utilizes only staking as an economic disincentive, while also employing a reputation system as a deterrent for such behavior.

Alongside this node selection mechanism, Tellor implements restrictions on data providers not present in the Chainlink ecosystem. In the latter, once data providers are paired with data requesters, the former may continue servicing a given query indefinitely provided it does so to the satisfaction of the requester and the network. Tellor, however, mandates a cooling off period, allowing nodes to service a specific node for only one out of every three five-minute periods. Further, while Chainlink utilizes query-specific aggregation contracts to algorithmically determine incorrect data provision, Tellor relies on its users to manually dispute unsatisfactory submissions. Reminiscent of the original Augur dispute process, Tellor’s dispute resolution mechanism allows any user to dispute data submissions, and this process repeats itself until no disputes are initiated.

Applied cryptoeconomic design emphasizes the need for nuanced consideration of how a given mechanism functions within a larger cryptosystem. On this basis, it is perhaps inappropriate to consider Tellor and Chainlink direct competitors. Tellor consolidates resources to serve only the purposes that are democratically determined to be most important. In doing so, it both theoretically enhances security and decreases data accessibility in a manner that a more traditionally marketplace-driven protocol such as Chainlink would not. While the relative immaturity of the Tellor network limits this discussion to theoretical considerations of merit only, it is reasonable to assume that Tellor is better suited for certain use cases while inappropriate for others. For example, it may be superior for use in prediction markets, which require maximal data reliability and wherein inherent subjectivity may require a user-centric, time-unbounded dispute resolution process. Conversely, individualized services such as Chainlink may be better suited for more niche use cases or certain decentralized financial applications wherein absolute assurance that data will be provided is a necessity. Ultimately, Tellor has the potential to meet certain demand that Chainlink current may not, though the relative scope thereof is currently unknown.