Analyse sandwich attacks on v3

[viniabussafi]

first scope: dex_sandwiched table

[viniabussafi]

Aggregated view here

[viniabussafi]

To view all transactions: https://dune.com/queries/4532901/7566569

[viniabussafi]

V3 had a high percentage of sandwich attacks on 25/12, but on average it has been below V1 and V2

[viniabussafi]

trend has been of growth over the past 2 days for v3

[viniabussafi]

switched to dex.sandwiches

[mendesfabio]

• look at USD volume instead of # txns
• find dex.sandwiches without a dex.sandwiched - or the opposite

[viniabussafi]

• look at USD volume instead of # txns

Looking at the codes dex.sandwiches (dbt_subprojects/dex/macros/models/dex_sandwiches.sql) identifies the attack transactions, while dex.sandwiched (dbt_subprojects/dex/macros/models/dex_sandwiched.sql) identifies the victims so, for the purpose of this analysis I believe it makes more sense to look at dex.sandwiches (i.e. the percentage of the volume of a pool that is due to sandwich attacks, representing a very toxic flow)

[viniabussafi]

Here is a query for the sandwiched trades on V3 considering both dex.sandwiches and dex.sandwiched: https://dune.com/queries/4665624. Checking if results actually make sense. WIll leave the issue open so we discuss on call