INTRODUCTION - Are Bots Evil?

Each blockchain address is associated with separate public and private keys. The private key is used to sign a transaction and the public key is used to verify the signature. A user using a wallet will sign the transaction and

Each transaction on a blockchain needs to be signed by a private key and verified against a public key before it is validated by the network. Users should be familiar with this process if they have ever used a wallet to swap tokens in a decentralized exchange (DEX), buy an NFT or provide liquidity to an LP, but this can also be done by a computer program. Whenever there is a script automating any kind of transaction in the blockchain, we identify the address signing the transaction as a bot. Although the term can imply malicious intentions, the truth is that some blockchain use cases rely on automated transactions to work.

Here are some examples of blockchain bots, both useful and harmful:

• Automated Market Makers (AMM) - AMM-bots use an algorithm to enable asset trades by adjusting the prices of assets based on supply and demand. They have become a key component of Decentralized Finance (DeFi).
• Liquidity Providers (LP) - also related to AMM-based DEXes, liquidity pools are created to capture liquidity for an asset pair. Providing one or more assets to a liquidity pool yields some percentage of the fees and can also be automated.
• Trading bots - a trading strategy can be precisely executed as intended when automated, eliminating the psicological influence that a human trader could have. This can be applied to assets or NFTs, for example.
• Arbitrage bots - some bots are programmed to take advantage of pricing differences of some assets across different markets. Imagine buying in one DEX and selling the same asset a little bit more expensive in another. Note that some arbitrage bots can also lose money!
• Maximum Extractable Value (MEV) bots - a subgroup of arbitrage bots, MEV bots use different strategies to profit from how transactions are ordered in a block.
• Smart contract bots - some smart contracts are programmed to execute certain actions if conditions are met. These can include asset transfers, which are coded in the smart contract.
• Spam bots - intend to clog or take down a network by sending a high number of transactions without a clear use.

Some of the clear harmful effects described above that have appeared throught blockchain history have been dealt with by different protocols, some others are a key part of how DeFi works today and can be further developed. So not all bots are evil!

This dashboard will analyse the number of addresses and transactions that can be associated with bots based on how many transactions can an address sent in a period of time for Ethereum and some of its L2 blockchains - Polygon (zero-knowledge rollup) and Arbitrum and Optimism (optimistic rollups) in 2022.

ANALYTICS

ETHEREUM

POLYGON

ARBITRUM

OPTIMISM

KEY INSIGHTS

• Bot addresses represent less than 0.04% of each blockchain’s addresses but make between 10% and 40% of the transactions.
• Polygon is the chain with the most bot addresses (7.05k) and bot transactions (401M). This is almost 10x more bots (749) and around 5x more bot transactions (77M) than in Ethereum.
• Arbitrum has the highest percentage of bot transactions of all 4 chains (41.7%) while Optimism has the lowest (9.7%).
• In Ethereum, Polygon and Arbitrum, bot failure rate is between 2% and 4% lower than human failure rate. Ethereum bots are the most reliable with 0.2% failure rate. Optimism bot failure rate is the highest at 9%, and it is 4% worse that for humans.

METHODOLOGY

Scope of this dashboard is all transactions between January 1st and December 31st 2022 that can be found in the core.fact_transactionstables from ethereum, polygon, arbitrum and optimismdatabases from Flipside Crypto.

This analysis was done using @BlumbergKellen definition of bot as published in this tweet:

• more than 3 different minutes with more than 40 transactions or more than 3 different hours with more than 240 transactions

For comparison purpouses, all 4 chains were evaluated with the same parameters, but given the difference in total transactions, these could be tweaked. This dashboard is intended as an overview, for deeper analysis the nature of the transactions should be investigated to come up with more precise parameters.

Analyst @Kaskz