Monitoring historical arbitrage can be a tedious job, therefore we want to lay the foundations for a generalizable system by understanding trading activity across multiple blockchains. We start by collecting bot activity from Ethereum and Polygon, then we use Explorative Data Analysis of emitted events to define a heuristic. If you are curious you can try it out here:

MEV Explorer (Have I been arbed) showcase

How it's made

It was made with AssemblyScript, The Graph, Python for Data Science and React for the frontend.The idea is: A trade often involves the WNative Token with the bot as a sender and receiver in one trade. To keep it simple we decided to use the graph protocol. This allows us to deploy our heuristic for the following networks: mainnet, xdai, poa-core, matic, fantom, bsc, avalanche, fuji, celo, fuse, moonriver, arbitrum-one, optimism, aurora. The heuristic works best for Ethereum, Binance Smart Chain and Polygon. After deploying the subgraph, we can look at historic data, flag bots. This could be useful for blacklisting bots or adding specific tokens to your exchange. In our frontend you can search for your smart contract and check if it was involved in bot trading activity. What makes the technology special is that we used the novel TransactionReceipt that was just released a week ago and hasn't been documented.