An operator running a 40,732-borrower monitoring network across Aave v3, Spark, and Morpho just posted a live ops dashboard showing 0 competitive wins in a 209-slot sample, and the reason cuts to the core of BuilderNet's orderflow attribution math.
Val from AdValorem didn't publish projections. They published live numbers: 1,888,510 submissions to relays, 952,705 accepted (50.45%), p50 submit latency 192.7 ms, simulation pass rate 95.81%. Their own rbuilder rail sits on the same box, 5 days 11 hours uptime, 5,201,817 blocks built locally. Yet in the competitive window where AdValorem bid against every major builder—buildernet-flashbots, quasar, titan, buildernet-nethermind, builder-plus—they landed exactly 0 slots and collected 0.0 ETH in subsidies.
What AdValorem Actually Operates
The network is a wedge upstream of bundle construction. AdValorem continuously monitors borrower positions across Aave v3 Ethereum (19,731 addresses), Aave v3 Base (18,769), Morpho (1,479 across 251 markets), and Spark (753). Coverage isn't theoretical: 24-hour re-check ratios hit 95.23% on Aave v3 Ethereum and 35.85% on Morpho. They admit the Morpho coverage is uneven and flag which markets are weak. Scheduled audits run 4x per day, with a blocklist enforced before any candidate reaches bundle construction.
On the builder side, rbuilder v93 runs with reth + lighthouse, fanning to four relays in full mode with max_bid_eth=5.0: flashbots, titan, aestus, agnostic. Ultrasound and manifold are disabled due to past degradation. Eden was dropped after it started redirecting to titan. Bloxroute is commented out pending re-eval. The extra_data tag on landed blocks reads "MEV SearcherNet". A public dashboard at https://mev.advalorem.io streams all this in real time.
BuilderNet's Attribution Blind Spot
AdValorem's first question to BuilderNet is the most important for anyone building orderflow infrastructure: How does BuilderNet's redistribution math attribute value when the incoming transaction isn't a pre-packed bundle but a raw candidate from upstream monitoring? The value of a liquidation-adjacent transaction lives partly in its existence—the fact that someone watched the position and generated the candidate at all—not just the execution price. BuilderNet's current model treats origination, inclusion, and backrun-of-inclusion as a single event, which means the originator gets no credit if they don't also win the bid.
Two operational issues surfaced during the TEE-attested integration attempt: port 7936 refuses connections from AdValorem's box across all three published BuilderNet IPs, and the attested-get v0.1.9 client fails parsing the measurements payload with ".cannot unmarshal array into Go value of type multimeasurements.LegacyMultiMeasurements.". The measurements endpoint returns 35 entries spanning v1.0.0 to v1.6.0. Val is offering raw logs and traces.
DOWG and MEV-Share: Missing Standards
The Data on Orderflow Working Group has focused on builder-share-of-winning-blocks metrics (xof.pics, Barter). No standard exists for origination-side data: which borrowers are watched, by how many independent operators, with what staleness budget. AdValorem can publish anonymized snapshots of their 40,732-borrower coverage stream, but they want to know if DOWG wants that schema before publishing into the void.
MEV-Share was designed as a backrun substrate for swap orderflow. AdValorem's third question hits the practical constraint: Has anyone used MEV-Share hints to backrun liquidation transactions? The hint-to-bundle latency budget against oracle updates is tight. The hint schema may not expose enough position-level context to identify a candidate as liquidation-relevant before constructing a backrun. Hint volume versus usable hint volume for non-swap categories is totally opaque.
AdValorem isn't asking for a partnership, capital, or program inclusion. Three technical questions, in public, because the answers determine whether liquidation-originated orderflow gets properly rewarded or stays as a subsidized input to someone else's block. If BuilderNet can't answer these attribution questions, the liquidation orderflow network—and others like it—will remain on the sidelines, and the block-building game stays a bid-value arms race with no credit for origination.
This article is based on a post by Val / AdValorem on the Flashbots collective forum. All numbers and URLs are quoted directly from the source.
Source: Liquidation-originated orderflow as a builder input - attribution questions for BuilderNet, datasets for DOWG, throughput questions for MEV-Share
Domain: collective.flashbots.net
Comments load interactively on the live page.