one human running a fleet of trading bots. real numbers. no theatre.

flashbots relay sample, 200 most-recent payloads (≈0.7h window, june 9 2026). total eth paid to proposers: 6.84 eth. mean per block: 0.0342 eth. median: 0.0088 eth. extrapolated daily mev to proposers: 246 eth. caveat: relay-only subset, not total mev.

top builder by value paid: 0xb7c5e609f8a0… at 3.584 eth. second: 0x850b00e05538… at 0.795 eth. third: 0x83d6a6abb623… at 0.621 eth. fourth: 0x88510a78794b… at 0.614 eth. fifth: 0x99cba5050480… at 0.266 eth. top-3 builder dominance: 73.1%.

falsifiable line: if any single builder captures >50% of relay payloads over a 24h rolling window, i’d flag relay-level censorship risk. not network-level, just the relay’s ability to filter or delay blocks.

the median block pays 0.0088 eth. the mean is 3.9x higher. that spread is where the bots live.

more at falsifylab.com #OnchainAlpha #PerpFunding #FundingArb $ETH

— research and educational content. not investment, legal, or tax advice. do your own research. positions and views may change without notice.

Originally published on FalsifyLab Substack.