how i made $15,500 at 16 through magiceden

summer 2024. i had just recieved the tensor airdrop, and was feeling elated. i already had some experience farming nft airdrops via tensor, and magic eden had announced a reward programme back in january. at this point, me and a friend started looking into it. magic eden was the number 1 marketplace for BTC ordinals, was winning solana marketshare, and was expanding onto eth. it had raised 160m at a 1.5b valuation, and despite this, it was incredibly underfarmed. i could barely see anyone talking about it, let alone farming it. i started farming it on may 22. the rewards programme worked as such: for selling/buying nfts (volume) you recieved diamonds (points) proportional to the volume, and for providing liquidity (bids, asks) you recieved volume. i decided to farm via bids and asks, as farming via volume meant paying collection royalties and a magiceden marketplace fee.

i purchased a bot to farm, but despite this, it was a very competitive environment. me and my friend quickly figured that for bids, distance to floor did not matter, which was awkward, as the bot was configured by distance to floor. instead, it was about the rank of your bid. we worked out that diamond distribution followed an exponential curve relative to bid rank. this meant a LOT of refining, we had to manually enter delta from floor to get to the acquired rank. the magiceden ui only showed 2 decimal places, ie 10.52 solana. to compete for rank, we had to use incredibly tiny delta, like 0.000001 solana.

there was also a lot of pvp. people would bid up above the floor price (before royalties), and other bots would follow, dragging bids higher and higher above floor, at which point the same people dragging bids up would dump the nfts on bots who chased them above floor. this meant that even with a bot, you needed to actively monitor positions, so screentime was essential.

diamonds per collection for bids depended on a few different factors, none of which were public. after some trial and error, we determined that the floor price and volume of the collection were the key things, but some collections even with a high floor price and volume earned relatively few diamonds, so we decided to focus on just the ones we knew provided a good amount.

after a week or two of farming, me and my friend realised that your diamond balance updated at 6 minutes past the hour, and 36 minutes past exactly, every hour. at first, we thought nothing of this, as we assumed this was just a distribution every thirty minutes based on localised random snapshots of the orderbook. my friend wanted to test further, and after a day or two, we realised that the diamond update time correlated with the snapshot time, rather than many snapshots being taken randomly between diamond update times as we previously thought, with time in the book being weighted. this meant that we could place a bid at xx:04 and xx:34, and remove at xx:07 and xx:37 respectively, thanks to the static snapshot times. this meant that per hour, we only were exposed for 6 minutes. this also allowed us to bid higher, as the time of exposure was lower.

this was gamechanging for a few reasons. firstly, we could avoid the normal pvp of bids being dragged above floor, and people's bots constantly fighting for ranks. secondly, it minimised the risk of being hit. this was important because each time your bid was filled, you were essentially guaranteed to lose money, thanks to marketplace fees.

before realising the snapshot times, i had to maintain a high rank for the full time i was farming, which was a lot more risky, and made being hit a near guarantee, as these collections would have sudden dumps all in one go. it is important to understand that most of this capital was mercenary, they did not want exposure to a down only class of assets, they wanted to minimise money lost, and minimise time holding these toxic nfts. people would dump 5-10 nfts in a single clip, clearing out the top part of the orderbook, and people below would pull bids as they didn't want to get dumped on, so now you had an nft, no bids nearby to dump into (and you didn't really want to do this, as you would lose money due to the spread + fees) and you just have to list it at floor and hope someone buys near the price you got filled at. however, you likely weren't the only one hit, so it's you and a few other guys all pvping and undercutting each other to be the lowest listing to get bought first. i remember being hit on degods, and the floor collapsed under me, and i lost $300.

it was pretty stressful as you had to make a split second decision whether to list the nft or dump into bids if any remained, because the person who dumped on you could have more nfts to dump, or other people who got filled alongside you might dump into any remaining lower bids. bids acted as a minimum price for the floor, so if people who got filled alongside you dumped into the bids, then when people who chose to list start to undercut each other, there was more space for the floor to drop. overall, this meant that being hit was a good way to guarantee you lost money, and this dragged up your average cost per diamond.

this is from tensor, not magic eden, but fundamentally the microstructure was the same

as this snapshot method allowed me to reduce the active time of my bids by 90% whilst also allowing me to bid higher to earn more diamonds, it massively increased the risk:reward ratio and ev of farming magiceden and allowed me to rapidly accumulate diamonds. i eventually hit 100k after a few weeks of this method. in my best week, i earned 35,000 diamonds.

i additionally gave this method to another friend, who automated this strategy for himself, setting a bot to autobid at the exact times (he also gave me some of his profit).

at this point, i decided to stop farming. i had started with around ~4k, and burned $1500-$2000. these were partially subsidised by my job (~$750, and a few hundred from trading memecoins). i was still down a significant amount relative to my portfolio at the time, so i sunset my farming after roughly a month. at this point, there was no end date to farming in sight, despite the team teasing q3, i didn't want to take on excess risk, and i moved on to other opportunities.

my final result was around 115,000 diamonds, after about 5 weeks of farming.

then, a few months later, magic eden announced their airdrop checker. at this point, i was pretty worried, as magic eden announced tokens would also be given to OGs, as an OG boost, users of their wallet, and for other small criteria. i worried that this meant i would be heavily diluted, and spent the days leading up to the checker incredibly worried and stressed, unable to focus on other things. i remember in the hours before the checker released sitting around and waiting, too nervous to do anything else.

i qualified for 2551 $ME. this was an insane amount for me based on premarket prices, so i started selling early on premarket. i remember sitting in class during a vectors test, and i was so giddy with excitement that i couldn't focus or answer any question. i remember i got 5/35 in that test. at that point, i knew my airdrop would be worth 5 figures, and i was euphoric. to 16 year old me, that was an obscene amount of money. i couldn't believe it. i didn't know what to do so i went to some friends for advice, who recommended hedging at least some on premarket, which i did.

hedging on premarket reduced my total. i could have sold all at $5.5 on open market at launch, but it was so much money to me then that i had to protect my downside. i remember for the claim, we had to download their shitty mobile wallet which didn't work, spent ages trying to claim while price went down. it opened at $10, and one of my friends managed to claim instantly and sell at $10, whilst me and all the other poors were stuck trying for like 30 mins.

ended up selling all for 12664, and my friend gave me a $2870 tip (this includes what he initially paid me), so in total i made $15500. this was done in a span of 5 weeks, using around $4000. my friend who i farmed it with made about 75k, but he was farming with 5-6x my size — using our method the airdrop was pretty linear regarding size:airdrop amount. who knows where i would be if i had more money at the time to capitalise on this, or if i had farmed for a little longer.

regardless, a fun memory, and i am happy to say that i have underperformed relative to that benchmark since!