Inspiration

The inspiration for this project came from a variety of places. The idea of lost NFT assets is not a new one. Cryptopunks, a 4 year old NFT project, is perhaps most famous for having many lost assets. Being one of the oldest NFT projects certainly means that some of these assets would be lost throughout the years, whether it be through lost private keys, death of the owner, or any variety of reasons. Punks were also famously a free project that had little to no value for years after their release. Predictably, many owners simply forgot about them because they were nearly worthless.

I had seen a few Twitter posts from noteworthy accounts detailing their lost apes already. On June 17th, 2021, @TAYLORWTF sent Ape 8860 to the 0xdead wallet address, effectively “burning” the ape. Why would someone do this? Well, I had to ask for myself:

@farokh has also revealed he lost 3 Bored Apes due to a lost seed phrase:

And perhaps my biggest inspiration was from @0xMunger's research on Lost Punks. They created a script in Python where they sought out to determine how many Cryptopunks were lost. The criteria they used was simple: if a wallet containing a Punk has had no interactions in 4 years, it is considered lost:

I wanted to do something similar to @0xMunger's project, but put my own spin on it. Since I do not know Python, I would find a way to use Javascript to find the amount of lost apes and display the data on a website with a clean UI. This way, anybody could easily understand the data points and it could automatically update as more apes are lost.

Styling

I decided to use Ant Design to style my website. This was my first time using a styling library.

I was incredibly impressed with the different components and the extensive documentation offered by Ant Design. It made my job much easier than doing CSS by hand, especially in terms of making my website responsive and mobile friendly.

Home

The homepage features statistics relating to how many apes are counted for each data point. I also decided to display 12 lost apes at random to get visitors curious about the rest of the project.

Data Page

Each data page consists of a grid of each ape and their ID number that applies to the selected criteria. Every time a Data Page is refreshed, it will randomize the order of apes. This creates a more dynamic page where you can truly see the wide assortment of apes for each criteria. I included a basic search box to search for apes by ID. I also included a few basic sorting functions to sort the grid of apes accordingly.

Clicking on an ape will take you to the Ape Details page. Here, you can see the metadata for that particular ape as well as links to Etherscan and Opensea.

Overall, it was important for me to showcase the apes in an interesting but simple way. I love the BAYC artwork and seeing them in a grid showcases the variety of different apes out there.

Lost Apes

The Lost Apes page has all of the same features as the other pages with one added component. Toggling the “Show Lost Apes Wallets” will make a table appear showcasing the wallet address and its corresponding ape count. The interesting data point here is that there is a single wallet with 27 apes in it that satisfies the Lost Apes criteria.

Vision

My vision for this project was to showcase all of the apes that are considered “lost” according to certain criteria. This made me think about certain data points I could use to narrow down this amount.

I decided to use 5 different data points that would help me determine if an ape was lost:

1. Unclaimed $Ape

The $APE coin claim took place in March 2022 and was open for 3 months until June 2022. This was an obvious data point to include in my research. The $APE coin claim was one of the most lucrative claims in the history of Ethereum. Each ape received 10,950 $APE coins, currently valued at nearly $40,000 (and valued at over $200k at its peak).

Due to my previous experience, I knew I could find this data using the Etherscan API. I would need to find the Apecoin Deployer address and extract the Logs from the Ethereum transactions. From there, I could find the ID of every ape that successfully claimed $APE coin and create an array of those values. I then created an array of numbers 0-9,999 to represent all ape IDs. Then I could compare the two arrays and the IDs that were not found in both arrays were the ape IDs that did not claim $APE coin. In total, I found that 96 apes did not claim $APE coin.

2. Unclaimed Dog

The Bored Ape Kennel Club claim was the first claim for ape holders in June 2021. This claim was only open for a week, so I expected there would be many more apes that did not claim their corresponding Dog. I did not expect to use this data as criteria for Lost Apes because this claim happened so long ago and was very short. But I knew finding this data would be quite easy since I had already written similar functions to find Unclaimed $APE.

Finding this data was identical to finding the Unclaimed $APE coin apes. In total, I found that 398 apes did not claim their Bored Ape Kennel Club dog.

3. Unclaimed Otherside

The Otherside land claim occurred in April 2022 and was open for 3 weeks. This was another valuable data point, as it was a lucrative claim after the $APE coin airdrop. Being the most recent ape claim would give me a lot of insight on the current status of lost apes.

Although this was another claim, I could not use my previous methods from finding the $APE coin and Dog unclaimed apes. There were a few ways this claim was different than prior methods:

Because of these differences I had to come up with a different strategy. First, I had to use the Yuga Labs ethereum address to find the Otherside Lands that they claimed after the public drop. However, Mutant lands were also included in this data. Luckily, ape lands were numbered 0-10,000, so I could simply filter by lands that were claimed by Yuga Labs and had an ID below 10,000. In total, I found that 266 apes did not claim their Otherside land.

4. Burned Apes

Burned Apes are apes that have been transferred to a “burn” address. The odds of a burn address being claimed are nearly 0, so it is safe to assume Burned Apes are lost forever.

This was the most straightforward data to find. Although any unused address can be classified as a burn address, the most used burned addresses are 0x0000 and 0xdead. Doing some research online, I decided to use the Alchemy API to find this data. I was able to filter the burn address by NFTs in the Bored Ape Contract and return the token ID hex that was present in the address. Then, using web3.js, I could translate that token ID hex to a token ID human readable number. I was able to follow this process for both burn addresses and combine the arrays together. In total, I found that 2 apes have been burned.

5. No Transfers

The premise of this data point was simple enough. I wanted to find out how many apes have never been transferred. However, this proved extremely difficult. The fool proof way to do this was to check each ape individually and get their transaction count from the Alchemy API. However, I cannot call an API 10,000 times and loading this data would take too long for a Javascript project. So I had to improvise.

Instead, I decided to take a snapshot of ape holders at sellout and now, at the current time. I was able to use the Alchemy API to find all ethereum addresses holding an ape and their corresponding IDs at a certain block number. I choose to use the block number for the last minted ape to find the ape holders at sellout, and I would use the latest block to dynamically find the up to date list of ape holders. Then, I could compare the two arrays and if the address and ape ID matched, I could reasonably assume that that ape has never been “transferred”. Although this isn’t a perfect way to find this data, it was the best I could come up with. Some limitations include:

Although alone this data would not help me find the total number of Lost Apes, I thought it was an interesting data point nonetheless and I would be able to use some parts of this component to find Lost Apes. In total, I found that 222 apes are in the same wallet that minted them.

Lost Apes

Now to finally find an answer to my question: How many Bored Apes are lost forever? I decided to use 3 criteria to determine if an ape is lost:

The first two criteria were easy to determine, as I had already extracted this data previously. I compared arrays from the apes that did not claim $APE coin to the apes that did not claim Otherside land. The apes that appeared in both arrays satisfied the first two criteria. In total, 75 apes satisfied both criteria.

However, filtering by the third criteria proved to be extremely difficult. There were no easily accessible APIs that I could find to return this data, so I had to get creative. First, I wanted to extract the wallet address that corresponded to each ape. By doing so, I could display a table that would return the ethereum address with the amount of apes in the address. I used a combination of the Alchemy API and web3.js to return the current holders of the 75 apes. From there, I used web3.js to get the transaction count of each address to detect inactivity, similar to the strategy I used to find the No Transfers apes. However, I could not simply find the transaction count for each address, but the transaction count since the Otherside mint. To do this, I received the total transaction count for each address, and the total transaction count up to the block of the last Otherside mint. If there was no difference between the two counts, then the address has had no activity since the Otherside mint. I was then able to filter out the active ethereum addresses from the 75 apes. This eliminated 7 apes. In total, 68 apes satisfied all 3 criteria.

I finally had the answer to my question. 68 apes are lost forever.

Testimonials

Finally, after a few weeks of heads down coding, I was ready to share my project with the world! I was extremely proud of this project in particular because I had to learn so many new technologies like React Router, React Redux, and Ant Design. I also had to find creative solutions to hard problems as well. Extracting data from the blockchain is not very fast, so I worked hard to minimize the amount of APIs calls and promises. Narrowing down my criteria for lost apes also took a lot of thinking and experimenting. But that's really what coding is all about, finding creative solutions to problems that no one has encountered before!

My project got a lot of engagement, especially from the BAYC community! With over 250 likes, 40 retweets, and 50 comments! Here are some of my favorite responses on the project:

There were a few people who actually knew the backstory behind a few of the lost apes too. That was super interesting to learn and read about!

Kevin, the CEO of Dust Labs / DeGods has also said some encouraging words about the project.

And to take it a step further, I've started researching the indiviudal Lost Apes as an attempt to find the owners and understand how they became lost. While most investigations eventually run dry, there have been a couple successful outcomes.

Sharing my projects with the world and gathering feedback has been instrumental to my growth as a web developer. I'm so happy to contribute to a community I feel so strongly about in a fun and creative way. It really gives me the confidence to keep learning and building.

I'm so proud of what I have created and look forward to tracking more Apes that become lost over time!

Update (1/20/2023)

I've added tabs for Unclaimed Sewer apes and Unclaimed Sewer (BAKC) dogs. These tabs represent apes that have not claimed their sewer pass and dogs that have not been used to claim a sewer pass respectively. It was important to update this project in real time in order to help prospecting buyers make informed decisions when purchasing an ape or dog.

I plan to continue updating Forgotten Apes as more claims become available for holders!