Normally when you swap on Uniswap the transaction is publicly relayed through nodes until it reaches miners. Due to this anyone who wants to can see the transaction before it is mined.
Bots watch for these transactions before they are mined and try to extract money from them.
This is how "sandwich bots" find their victims.
Here's what a sandwiched transaction looks like:
🥪 gets ahead of it, buying up a token to raise the price for the victim
Victim executes their swap, now getting less tokens but raising the price more
🥪 sells their tokens at the elevated price, pocketing a profit
Alternatively in some cases it makes sense for bots to land right behind a pending transaction and arbitrage away any profit made from potential distortions.
Ethermine's backrunning bot landed an impressive 3 arbs this way, including behind Vitalik's tx
Vitalik had 4 transactions fail at an eye watering ~1000 gwei gas price and was probably mindful of the aforementioned sandwich bots that could making be making his transactions fail.
So what did he do about it?
Recall that bots watch for pending transactions while they are in flight to miners. In order to sandwich a trade it needs to be made public.
So what if transactions were sent directly to miners instead of through full nodes where anyone could see them?
That is the basic premise behind new products like Flashbots, Taichi, MistX, or ArcherSwap: provide a direct line to miners so users can't be frontrun.
Vitalik used ArcherSwap to sell some of his Shibu earlier, and his transaction then succeeded where they had previously failed.
Read more about ArcherSwap in @calebsheridan's tweets below
ArcherSwap is built using a fork of MEV-Geth and MEV-Relay - two codebases from Flashbots Research.
ArcherSwap routes trades to miners on the Archer Network, which in this case was mined 2miners. etherscan.io/block/12421138
At Flashbots we're super excited by innovation like this that provides better experiences (e.g. frontrunning protection and creative use cases) and outcomes (better prices) for users.
A brief thread on a novel MEV searching strategy, where we chase the trail of a mysterious bot backrunning private flow and reveal how they do it.
@blairmarshall pointed out a bot that appears to have private access to user orderflow that was landing bottom-of-the-block blocks on the Flashbots builder. That didn't make sense to me. We don't run backrunning bots! So we investigated.
MEV-Boost payments were at an alltime high yesterday, totaling 7691 ETH (!) which is nearly double the previous ATH of 3928 ETH during the FTX fiasco this fall.
A few statistics on MEV on Ethereum yesterday in this thread
You can't compare stats these 1:1, but the ATH for daily miner profit from mev-geth was 6397 ETH in June 2021. That's the *profit* of running mev-geth vs a vanilla mempool mining client.
A similar metric here would be the difference in payment for validators from running mev-boost or not. There's not a great up to date estimate of this out there I think
You could derive it by looking at the value of the mempool builder we submit (0xa1defa) and the winning block
Brutal & unfortunate. A bad aggregator led to a user getting rekt.
A reason that a competitive MEV market & orderflow auctions are deeply important is that cases like this could potentially be entirely mitigated. The user could be paid back the arb that their wreckage left.
The user's $2m ended up in a Uni v2 pool, which an MEV bot was quick to snatch up for basically nothing of course. They paid a premium to a miner for this $2m but it really wasn't anything (~20 ETH).
The MEV market was strangely uncompetitive in this case. 20 ETH paid to the builder for capturing $2m is an insanely good trade and you'd expect competition to push that 20 ETH much much closer to the value of $2m.
MEV-Share builds on MEV-Boost by further unbundling the transaction supply chain.
Whereas MEV-Boost enabled collaboration between validators and builders, MEV-Share does the same for searchers and users - empowering users to be paid for their transactions.
With privacy users can bargain for the MEV they create without permissioning searchers. But, programmable privacy allows users to selectively share information to enable optimization and collaboration.