The fun disaster of blockchain gaming data
For great reasons, I promise
What’s a game?
Data standard needs in web3 are changing daily. That's the challenging part.
The fun part? It's because web3 gaming allows for things that have never before been easy.
Standardization and consistency are the two most important things in data analytics. Even accuracy doesn't best those factors: if something is 15% off, but the definition is understood and it remains consistent and directionally correct, that's better than analyses that bounce between 0% and 10% margins of error. One I can learn from and action on, the other less so.
An example:
In a 10% margin of error, let’s say I see that Game X sees revenue rise 5% in a week, while Game Y saw a 2% decline. If I’m Game X, I’ll feel like I’m besting my competitor, and if I’m Game Y, I may go into a pricey UA frenzy or make costly design changes. But actually, Game X went down 3% and Game Y up 7%. You get it.
In web2, the centralized nature of gaming, and the limited types of transactions (direct purchase, subscription, microtransaction) make standardization easier, and if it lives on Xbox, Playstation, or Steam, it's a fair to assume it’s a game.
We don't have that in web3. If an item lives on Polygon, Immutable, Avalanche, etc., it could be a game item, a PFP, a coffee loyalty program membership. And because these ecosystems are permissionless, anyone can push any smart contract without approvals or tagging; the chains themselves don't know what is what. Many more games are living on separate launchers, or simply as HTML games, than on traditional game platforms for now.
This makes simple counts extremely hard—How many web3 games are there? How do we understand market and genre aggregates? But here's the real complexity: we've officially entered the era of mainstream interoperability.
What’s a game item?
Nike's .SWOOSH x EA announcement last week was a release that blew my mind given the people I know making calls at EA. Pretty soon, Nike NFTs can be brought into EA Sports titles like Madden, so you may see your friend's Madden player wearing a sweet new pair of kicks that you haven't seen in-game before.
Nike hadn't revealed much about the .SWOOSH world, but one thing it definitely didn't mention was games. Assets that two weeks ago looked like fashion NFTs and part of a burgeoning Nike loyalty program overnight became a new class of gaming assets. But are they?
What's an analyst to do? Here are some considerations:
Are all .SWOOSH and EA compatible items now counted as game assets? Or only those that are actively brought into the game?
If the latter, this would be at odds for the standards of other game collections like PFPs that may have no in-game utility so aren't "brought into" the game, but are part of a gaming ecosystem.
If an EA-compatible .SWOOSH item is sold, should that count as transaction volume for a now web3-enabled game, or does it belong solely to the fashion label?
To put this in perspective in the web2 sense, it's as if the Room & Board couch I bought for my apartment could've been mapped to and unlocked an identical couch in the Sims. Should that couch and its purchase price get rolled up as a Sims microtransaction? When I sell the physical couch, does that count as a change to the volume of game assets?
Analysis by behavior
In this world of multi-use NFTs, we may be building a Venn diagram set of analyses based on behavior and usage, not origin. That's relatively new. Never before, for example, have in-game transactions been counted only if an item is used. It's a nice-to-know for understanding LTV, but doesn't typically change earnings totals. (However, other currency and in-game item accounting practices have led to chaos in the web2 past, like when Zynga changed its recognition of digital goods based on a new, theoretical user lifespan.)
In web3, behavior matters, utility matters. Who's to say that a Bored Ape isn't a game asset rather than a PFP if its primary usage ends up being for games? For Dookey Dash, Apes certainly were gaming NFTs.
But the inverse could be true, too. If an Axie Infinity item never touches an in-game use case, but serves only as a trading chip on the secondary market (which, I argue, is a metagame in itself), is it not a game asset?
The implications of this uncertainty can be disastrous for developers who are building business plans and UA funnels based on analytics (which, duh, they should be). During my SuperData days, we helped companies greenlight games into new genres based on aggregated accounting of those genres and their trends. In web3, the potential to dramatically under—or perhaps worse—over-value a sector is significant. Mislabeling could also leave the major wins of interoperability opaque to developers and brands.
We're only seeing the tip of the iceberg. The complex web of overlapping and interconnected ecosystems and assets has just barely begun, and as an industry we're still far from grasping it in a sophisticated—and apples-to-apples—way. But I certainly don't bemoan a challenge arising from all the awesome new things this tech will let us do.