top of page

Metaverse Data - Don't trust, verify!

The Gauntlet Thrown

It came to our attention that what we assume to be a smear campaign is being run against our metaverse home – Decentraland – purporting that there are only 38 daily active users.

Atlas CORP is not affiliated with the Decentraland Foundation – we are not the platform infrastructure providers of Decentraland – yet we feel an obligation to stand up for our home and refute this to show that there are roughly 8,000-10,000 daily active users. We are Decentraland DAO grant recipients to make this data available to the community, which came after a contentious vote about the use case (we hope the DCL Foundation now appreciates the importance of having a third party corroborate and provide this data).

What happened?

Recently, Coindesk author @camlearnscrypto published an article broadcasting to the world that Decentraland only has 38 daily active users, or “DAUs”. This immediately caught our attention as the value was so egregiously off from what we know to be the true number [1] that it’s almost impossible to think this was done out of incompetence, but out of malice.

It's a bit frustrating that an article lacking this much journalistic integrity could be released without any attempt to fact check or validate the data. Coindesk claims that they are

…are building the most influential, trusted information platform for a global community engaged in the transformation of the financial system and the emerging crypto economy.

As a "crypto-economy” publication, they should understand that the core ethos behind the crypto ecosystem is that you do not need to trust anyone; you can instead run open source software to verify things for oneself. The author could have hopped into Decentraland and seen more than 38 users prior to publication, but instead believed a third part source - dAppRadar - that only shows smart contract interactions for a subset of Decentraland’s smart contracts.

To understand why this type of crypto-reporting is problematic, let’s first discuss what it means to think about web3 data and why it’s a departure from web2 data.

The difference between web2 and web3 analytics

In the web2 world of yesteryear, determining user data was straightforward for those managing infrastructure. Websites need to serve data to users directly and details of the requests can be logged – such as using Google Analytics as a third party helper tool to do so [2]. Gaming platforms require centralized servers and therefore can log connections and time spent connected similarly.

The key point here is that it’s the same people running the infrastructure that report on the data. You have to trust them that they’re telling you the truth. You, a user, almost never have access to their user logs and raw data used to compile reported results. This is a conflict of interest, as increased users may lead to increase funding and support which provides an incentive state a higher user count, and if you can’t verify their statements through your own computation, there is no good way to validate their statements. You just have to take them at face value.

The crypto ecosystem provides an alternative – don’t trust, verify. We were drawn to Decentraland as it envisioned a metaverse platform built on cryptocurrency principles. It goes beyond ownership of assets on the blockchain (e.g. via NFTs) to a platform run on truly distributed infrastructure. Anyone can run a Decentraland node which allows users to connect and play [3]. As of today, the node deployment comes with open APIs that can be used to retrieve information on current users – as identified by their ETH addresses – such that anyone can see this data. It separates the incentives of those running the infrastructure from those collecting and reporting on the data to keep the node operators honest about their uptime and the number of users running on their specific hardware.

Further, most transactional data is not run through closed web2 / tradFi platforms that do not expose the data but instead on open blockchains and through smart contracts. This removes the ability for platform purveyors to be able to misstate any earnings or transactional volume as it can be looked up on public blockchains. You can just look at the blockchain to see how much has been transacted!

Tl;dr – If you’re running a truly decentralized platform like Decentraland, the analytics can be observed and computed by separate parties from those running core infrastructure, removing any moral hazard that might exist to misstate true numbers for gain. It’s a lot more likely to be accurate and if you don’t trust the numbers, you can go get them yourself.

How to think about Metaverse data?

Metaverse data doesn’t come in the same shape as web2 analytics. Another reason this article was problematic was that it only focused on one narrow definition of data, and misconstrued it with another metric’s definition. The article claimed that the number of smart contract interactions was indicative of daily active users, which is akin to looking at a brick-and-mortar store’s total receipts printed by the end of day and calling that foot traffic. The metaverse by definition contains more information in its data than the web2 space and therefore can provide a much richer set of analytics.

In one sense, active users is fairly similar in web2 vs. web3: the number of unique visitors to your platform. Metaverse active users has additional dimensionality that you might not find on a flat, web2 site such as geographic location within the metaverse, or realm (the specific server the user is playing on). In fact one of the first things we discovered in the data were users playing on multiple realms simultaneously – effectively showing up in the metaverse twice, with a time on site double the observable window having spent an hour in a scene over the course of 30 minutes – which doesn’t have as much of a real-life analog.

In web2, transaction data is rarely public. Aside from self-reported figures (e.g. McDonalds promoting billions served) and reporting disclosures required of publicly traded companies, individual transaction data is almost never accessible to someone outside of the centralized organization.

In web3, transactions occur on-chain. If organized properly, this data is readily readable from the blockchain. Users can peruse smart contracts deployed to sell items – wearables, land, club membership, subscriptions – directly. This allows any user with an interest to retrieve and aggregate the information at their leisure.

dAppRadar and Coindesk reported on one smart contract from Decentraland, omitting the vast majority of the on-chain ecosystem of smart contracts. Decentraland in particular spans many smart contracts across multiple blockchains (e.g. Ethereum Mainnet, Polygon Matic), and services like The Graph help index this data to obtain answers more easily. It was all there for the taking, if you know where to look.

Why Atlas supports data initiatives

It is imperative to the creation of a truly Decentralized metaverse that data for platforms like Decentraland remain open and available such that independent users can validate use cases to protect against slanderous cases such as the recent Coindesk article.

The whole premise of crypto is trustlessness, that one should not have to trust another but instead rely on the protocol to function properly removing trusting individuals from the equation. Coindesk failed to independently validate their data claims, demonstrating a fundamental misunderstanding of the very ethos behind the ecosystem they claim to support.


[1] Even we can't report on the "true" number of users, as it is possible to play Decentraland on private servers not listed by the DAO smart contracts, so there could easily be more hidden private servers. Isn't Decentralization neat?

[2] Once centralized, Google can in theory perform crazy derived-analytics on top of your data without your explicit permission.

[3] As mentioned in [1], anyone can run their own private node but to be listed for the public to play you need to be approved on the Catalyst Smart contract.

bottom of page