Gas-schedule divergence dashboard
Historical Ethereum mainnet transactions replayed under a candidate gas schedule and compared with the baseline. Pick a schedule, then explore the divergence overview and transaction failures.
Impact at a glance
Group composition (totals)
How all replayed transactions are distributed across the six mutually exclusive groups.
Loading chart…
How the six transaction groups are defined
Every mainnet transaction in the pinned block range is replayed under the candidate gas schedule and categorized into exactly one of these six groups. The three changed groups — Succeeds with changes, Fixable, and Potentially broken — require a working baseline (the transaction succeeded before the repricing); a transaction whose baseline replay already failed is Already failing, so it can never be counted as newly broken.
No change g1
No recorded difference in outcome or execution trace under the repricing.
Succeeds with changes g2
Still succeeds at its original gas limit, but something changed: gas used, event logs, or output/trace. Needs a working baseline.
Fixable with gas-limit increase g3
Worked before; now fails at the original limit, but completes in the replay
with a higher limit of up to 10×. This does not establish that the required
limit is practical under block or protocol constraints.
Potentially broken g4
Worked before; now fails and is not rescued at the tested 10×
gas ceiling. A small subset is still out of gas at that ceiling, so fixability beyond
10× remains unknown.
Already failing af
The baseline replay itself failed, so a schedule-side failure is not newly broken. Includes permanent fail→fail rows and baseline-fail→success rescues.
Unknown g5
No retained per-transaction detail, usually because the producer’s per-block diagnostic cap was reached. This is a coverage gap, not an execution outcome.
Method and limitations
Each historical transaction is replayed independently against its canonical pre-transaction state under the baseline and the selected candidate schedule. The comparison records changes in success, gas use, logs, output, and execution trace.
The producer tests the transaction's original gas limit (1×) and a
10× ceiling. When the ceiling replay completes, gas used divided by the
original gas limit gives the measured multiplier used for Fixable transactions;
this is not a discrete 1/2/4/8 sweep.
This is a counterfactual replay of historical activity, not a forecast. It does not model how users, wallets, contracts, block builders, or fee markets would adapt after a repricing. “Potentially broken” therefore identifies replay risk rather than proving future production failure.
Terms: baseline = the original gas schedule; gas delta = candidate gas used minus baseline gas used; OOG = out of gas; call depth 1 = the transaction's entry contract; drill-in = retained per-transaction diagnostic detail; execution trace = the recorded execution path; p50 = median; p90 = the value below which 90% of observations fall. EIP-2718 type refers to the transaction envelope; an EIP-7702 authorization transaction carries an authorization list.