Thread by @larsiusprime

After months of incredible growth, the daily earnings of the typical player of Axie Infinity (a "scholar" in the Philippines) have fallen below the Philippines' minimum wage line for all but the high ranking players, and even they have seen earnings decline since August

This is just one of many findings of a highly detailed research project that I worked on in partnership with @naavik_co

Find our podcast summarizing the findings here:
share.transistor.fm/s/7c06129e

Our read the free report here:
naavik.co/business-break…

@naavik_co "Scholars" are players, typically in developing countries, who cannot afford to purchase 3 Axies (NFT-based virtual pokemon-like creatures) which are necessary to play the game. They are loaned these assets by "sponsors" who take a cut of their earnings.

@naavik_co Most of the players of Axie Infinity are scholars. Current estimates from Sky Mavis themselves suggest about 60-65%, based on data directly from the blockchain.

Here's what earnings look like for a non-scholar, who owns their own axies, and has a higher daily "energy" cap

@naavik_co Daily energy is a mechanic borrowed from F2P which regulates the amount of daily activity. 20 is the minimum, you can gain up to 60 by owning many Axies (either by breeding them, or purchasing more).

From what we can tell, most scholars are limited to 20 energy per day.

- The majority of the players of Axie Infinity are scholars
- 55% of players are in the Philippines
- 65-70% are in Southeast Asia generally
- 10% of players are in the USA

(Source: Sky mavis)

Axie Infinity has extremely impressive retention, but is this retention for a game, or for what is effectively a job?

High retention in games typically signals loyalty. But workers only retain while you pay them. We must closely watch retention as earnings fall.

We do see some signs that search interest in the game is declining, both worldwide, and in the two greatest areas of search traffic -- the Philippines and Venezuela

Here is a graph of SLP, the chief in game reward currency, over time.

"the supply of SLP ... continues to rise, which naturally pushes its market price down. This points to a SLP imbalance in the game, and if Axie Infinity does nothing, it’s only going to get worse."

SLP is consumed to breed Axies, which means the price of SLP is tied to the growth in new players, which are a chief source of demand for new Axies.

When players who are playing the game as a job cash out SLP for real-world money, SLP is not burned, it merely changes hands.

Here is a projection that the @naavik_co team put together for the value of AXS, Axie Infinity's governance token and the chief point of interest for investors

@naavik_co "What the above projection essentially shows is that Axie Infinity’s current revenue model is less driven by total player numbers and more driven by the pace of new player additions."

@naavik_co "This is a natural result of both breeding and marketplace activity fundamentally serving to onboard new players with their own Axies, and those events are front-loaded."

@naavik_co "Therefore, if the pace of new user additions slows — and Sky Mavis wants to maintain a consistent ~3.50 Axie Count / DAU ratio — the game’s revenue output will shrink. "

@naavik_co So... is Axie Infinity doomed?

Not necessarily.

But they face challenges they should take extremely seriously.

Here's a link to our assessment of "What's Next" for Axie Infinity:
naavik.co/business-break…

It's important to point out that in countless interviews and statements, Sky Mavis team knows and acknowledges their current business model is unsustainable.

They must transition to something that can sustain itself, and they have several plans to do that.

1. Axie: Origin / Battles v2
2. Land gameplay / Project K
3. Building out the "Axie Nation" / "Axie Universe" (ie, Metaverse stuff)
4. Ambitions to become a "vertically-integrated" platform for all of crypto-gaming

Play-to-Earn is not a sustainable business model if you have more money leaving the economy than is being put into it.

A successful carnival must have many attendees who pay $5 to ride the Ferris wheel and eat popcorn and have fun, and not expect $6 when they leave.

Therefore, all of the hopes that are pinned on the 4 future plans MUST manifest as something more than simply finding new avenues for growth.

Our intuitions about what KPI's (Key Performance Indicators) mean are upended in the Play to Earn model

High retention might not signal loyalty but mercenariness as long as daily earning is high.

And if the playerbase is expecting money out, is that an asset or a liability?

With that said, let's consider what has been planned.

First is Axie: Origin, the entirely F2P, non-crypto version of the game, sometimes called "Battles V2"

The original plan was to use this to "go mainstream," but distribution has certainly been cut off for Apple app store, Google Play, and now Steam. EGS remains open, potentially, but console launch seems unlikely as well (though no policies have been confirmed there).

It may be the case that a perfectly crypto-free version of Axie might be acceptable to some of these distributors.

In that case the question becomes what to do with that audience. Monetize them in traditional F2P fashion, or use them to drive traffic to Axie: Infinity itself.

The stated intent so far seems to be the latter. Whatever one thinks of traditional F2P monetization mechanics (they are often exploitative), using Axie: Origin as a traffic generating machine for the crypto game does not fundamentally address core contradictions in the P2E model

F2P, from a purely bloodless perspective, is capable of bringing in more money to the system than goes out of it, and is therefore at the very least self-sustaining. Growth that levels off leaves a paying customer base, but this does not seem to be the case with P2E so far

But this is just one of four planned things. Let's consider the next one, land gameplay and the Lunacia SDK

Axie: Infinity has a set of ~90,000's plots of land and is selling them off in batches, each backed by a tradeable NFT in their ecosystem. Details are sketchy about what "land" gameplay will look like, we mostly have to go off screenshots and some terse hints from statements

As best as we can tell, this will be something like a very simplified version of Animal Crossing / Habbo Hotel / Stardew Valley where land generates resources and you can decorate it.

But crucially it's been said that land will be the gateway to the Lunacia SDK, a metaverse.

The seems to be that there will be some kind of simplified Roblox-like environment, but you must own (or rent) land to deploy your experiences.

This is a great way to raise money, but it is not a great way to allocate scarce digital real estate to ensure the most creative usage

The observed 30-year history of land and housing crises in MMO's, as well as observed economic history from the real world, famously noted by David Ricardo, Adam Smith, and Henry George, tell us what will happen:

Speculators buy the land, and now the Rent is too Damn High

For more on that particular topic, see this article:

gamedeveloper.com/business/digit…

This is not a purely theoretical statement either, it's born out by countless real world economic treatises, and we have seen actual housing crises in Ultima Online, Final Fantasy XIV, and EVE Online. And the stakes were much, much lower there.

The issue comes from "land-like assets."

Land-like assets are any assets that are:

1) Strictly scarce in supply
2) Necessary for production
3) Obtain value based on their location and/or proximity to economic activity/population centers

EVE Online had land-like assets when it launched. They were called "factories" but they were more like land because they were scarce and created by the devs themselves. These were necessary to produce ships.

Speculators bought all the factories and held them out of use.

This is much the same motivation as scalping, but at least in the real world people can eventually make more PS5's, and you don't need a PS5 to engage in productive activity.

When assets are sufficiently land-like some extremely perverse economic incentives set in:

1) Creators cannot create without accessing land
2) You know there will not be any more supply
3) You can profit from your neighbor's activity while doing nothing

But this world is digital. Why not just make more land?

Well, the land has already been sold on the promise that it will *remain scarce*.

This creates a bought-in coalition resistant to change, the "transitional gains trap" as Gordon Tullock calls it

In an interview Sky Mavis hints they are aware of the problem, and to be fair we really don't know what land gameplay will look like. But they have already sold land to speculators who expect to profit from it in the typical way.

Sky Mavis in many interviews say they are interested in "removing middlemen" that are endemic to "web2" and the traditional games industry, but there are many challenges for a system where creators must pay virtual landlords for the privilege of creating value for Sky Mavis

Hopefully Sky Mavis takes these concerns to heart and reforms their plans for the upcoming land gameplay. It will likely be extremely unpopular with the faction that has bought individual land plots for as much as $1.5 million dollars

That said, Sky Mavis has the power to do this as their current design is fairly centralized. Assets live on a private blockchain that is operated by a set of hand picked validators that they control, and further must connect to a closed source app fully within their power

They have made many promises to gradually decentralize and to their credit they seem to be continuing along that roadmap; at the moment however they still retain significant power, so they have a lot of flexibility in what they can practically do with their ecosystem.

For instance, they can unilaterally ban any account and any of their axies "for any reason or for no reason" according to their legal terms of service.

They can also perform traditional game balance updates like buffs and nerfs, as they just did with season 19

AXS tokens will eventually be used to establish community governance, but it is not clear yet exactly how this will take shape. These features are still in their infancy, with the Katana DEX launching only a very short time ago.

Sky Mavis has said that their ambitions are to turn both Axie itself and Sky Mavis into platforms -- services upon which others build value.

Axie through the Lunacia SDK and other features, and Sky Mavis itself through Ronin and Sky Mavis Hub, for 3rd party crypto games.

The ban of crypto games on mainstream platforms might signal an opportunity, and all eyes are on Sky Mavis given their enormous player counts.

However, the one thing they don't have in common with Steam and Roblox (the two most cited antecedents) is time.

Steam launched in the midst of an era when everyone said "PC Gaming is Dead." They had many years to slowly grow without serious competition.

Same for Roblox, which originally launched in the mid 2000's and which many of my colleagues only learned about this year.

Meanwhile, investors are pouring literal millions into every competitor under the sun who has ambitions of becoming "The Platform" for crypto games. The ocean is no longer blue, it has gone from purple to red in short order.

Axie Infinity is not necessarily doomed. They do face serious concerns.

Play to Earn does not seem sustainable by the numbers and their leadership acknowledges this.

Virtual land actively disincentivizes creators to create.

Axie Infinity must evolve to survive, and quickly.

This has been a very critical report so far but I want to underscore that neither I nor anyone at Naavik want to see anyone fail. The path forward is through hard questions

Please kick the tires of this report and point out any mistakes, misapprehensions, or dubious assumptions.

The Sky Mavis team are trailblazers in this space and assembling the plane in mid-flight, and it precisely for this reason that we want to give them all the information we can see from where we're standing at ground control

Best of luck, genuinely, sincerely, to all

Thanks to @naavik_co and all of the collaborators on this paper -- that's @k_abhimanyu11, @aaron_bush_ , @jimmyrstone, @Ant_Pecorella -- give them all a follow!