This repository is a collection of public domain Ethereum MEV-Boost winning bid data.
Block data is extracted using cryo leveraging the Infura RPC, and merged with data from the following relays:
Data coverage begins on October 10, 2023 at block 18,320,000. We may backfill more data in time.
The data is delivered in Parquet chunks of 10,000 blocks, allowing for incremental bandwidth for users who choose to keep their datasets updated.
# Validated with python 3.11.6
import os
import pandas as pd # pandas==2.1.2
base_path = './data/'
file_paths = os.listdir(base_path)
# this assumes that the data directory is in the working directory
dfs = []
for file in file_paths:
df_tmp = pd.read_parquet(os.path.join(base_path, file))
dfs.append(df_tmp)
dfs.append(os.path.join(base_path, 'minority-relay-backfill/backfill__minority__relays__blocks__18320000_to_19530000.parquet'))
# add in the backfill data from other relays.
df = pd.concat(dfs)
df = df[df['payload_delivered'] == True]
# drop undelivered payloads
df.sort_values(by=['block_number', 'bid_timestamp_ms'], ascending=True, inplace=True)
# double sorting by block_number and bid_timestamp_ms allows the data to stay
# ordered inclusive of non-mev blocks missing bid_timestamp_ms data.
df.reset_index(inplace=True, drop=True)
block_datetime
: datetime64[ns]
slot
: int64
block_number
: int64
block_hash
: str
value
: str
base_fee_per_gas
: uint64
gas_limit
: float64
gas_used
: int64
num_tx
: int64
builder_label
: str
builder_pubkey
: str
relay
: str
bid_timestamp_ms
: float64
slot_time_ms
: float64
proposer_pubkey
: str
proposer_mev_recipient
: str
optimistic_submission
: bool
builder_fee_recipient
: str
block_timestamp
: uint32
extra_data
: str
payload_delivered
: bool