“Did 1559 cut overspending on gasoline costs? I used BigQuery to try and find out. First, we need to define overspending; my working definition is the difference between gas price paid by the user and the lowest non-zero gas price in that block (or 1 GWEI, whichever is greater).
Multiply that by the gas used and add it up per day and you get the total miners overpayment. Express it as a total gas ratio and you have a factor – so 2.0 would imply people paying twice as much as needed to have their transactions included.
So, did the EIP-1559 help reduce the overrun factor? Be the judge. See if you can choose where on the graph it has been introduced.
find it particularly interesting that it fell sharply when the fork occurred even though most transactions were still legacy transactions – probably because the base fee offers a good oracle price even if you send an old TX! “