Philip Daian posted a paper to arXiv at 15:10 UTC on April 10, 20191. He was a doctoral student at Cornell Tech. Seven co-authors signed it, with professor Ari Juels as the senior author2. Its title was "Flash Boys 2.0."
The paper named a profit that had gone unnamed: miner extractable value, or MEV1. MEV is the money a block producer can make by choosing which trades go first, last or out4. A bot that pays a higher fee to trade ahead of you collects some of it.
In 2026, that name organizes Ethereum's block market. On October 1, 2026, 92.0 percent of Ethereum slots carried a block built through MEV-Boost11. Flashbots, the group Daian co-founded, wrote that software56.
The paper that named it
Daian's team spent 18 months tracking trades on six decentralized exchanges2. Bots bid up fees to get their trades placed first, they found. The paper calls these contests priority gas auctions1.
Then it made a bigger claim. “We additionally show that high fees paid for priority transaction ordering poses a systemic risk to consensus-layer security,” Daian and his co-authors wrote1. In plain terms, a miner paid to reorder trades might go further. That miner could rewrite recent blocks to take funds, the team found2.
Daian explained the stakes in Cornell Tech's write-up on April 29, 2019. “The blockchain doesn't get rid of the middleman. It just turns one middleman into 100 middlemen, who you hope are not all being bribed or working against you for their own reasons,” he said2. The paper was presented at the Cornell Blockchain Conference on April 13.
Note the noun he chose. Middlemen are people with motives, and the paper set out to count their take1.
Both sides, in their own words
Did naming the problem make Ethereum's block market fairer?
“was originally introduced in the 2019 paper Flashboys 2.0 by Daian et al.”
“MEV-Boost was explicitly designed to foster decentralization, causing the side effect of enabling risk-free sandwich extraction from unsuspecting users, potentially raising concerns for regulators”
What the coverage missed
CoinDesk covered the study on May 7, 2019, in a two-minute story by Yogita Khatri3. Its headline called the bots "Predatory." The story credited a report in Homeland Security News Wire. A correction says it first credited the Department of Homeland Security by mistake.
The story quoted Daian on miners and bots. Its text leaves out the paper's new term, miner extractable value3.
That term was the part that lasted. Flashbots later wrote that MEV “was originally introduced in the 2019 paper Flashboys 2.0 by Daian et al.”5 By then the M stood for maximal.
From a warning to a market
Flashbots launched with an essay dated November 23, 20204. That essay warned about the very thing Flashbots would later face. “MEV Extraction could likely go in a direction where it is centralized to a few players,” Flashbots wrote. Its fix was open software, starting with MEV-Geth for miners.
The software spread fast. MEV-Boost lets a validator hand the job of filling its block to outside builders. Relays pass the best-paying block from builder to validator8. Before the Merge, more than 90 percent of Ethereum's hashrate ran MEV-Geth, Flashbots said in October 20225. Within two months of the Merge on September 15, 2022, MEV-Boost reached almost 90 percent of validators.
Flashbots called the result a problem of its own. “We have been very successful in our mission to democratize MEV extraction, almost too successful,” it wrote5. The same post made a forecast. “We believe the MEV-boost relay will no longer exist in two or three years,” it said.
Four years have passed. The dashboard at mevboost.pics shows MEV-Boost at 85.6 percent of slots on June 1, 202411. Its highest first-of-month reading came on May 1, 2026, at 94.2 percent. On October 1, 2026, the share stood at 92.0 percent.
Share of Ethereum slots whose block came through MEV-Boost, on each month's first day, in percent
- 85.6% low, June 2024; 94.2% high, May 2026
- Every other monthly reading
- Height: Share of slots, percent
Source: mevboost.pics [11].
Read the chart as a table
| Item | Value |
|---|---|
| Jun 2024 | 85.6% |
| Jul 2024 | 91.8% |
| Aug 2024 | 92.2% |
| Sep 2024 | 87.2% |
| Oct 2024 | 91.6% |
| Nov 2024 | 90.2% |
| Dec 2024 | 90.7% |
| Jan 2025 | 90.6% |
| Feb 2025 | 89.3% |
| Mar 2025 | 90.9% |
| Apr 2025 | 91.3% |
| May 2025 | 91.1% |
| Jun 2025 | 91.3% |
| Jul 2025 | 92.5% |
| Aug 2025 | 90.3% |
| Sep 2025 | 91.3% |
| Oct 2025 | 91.4% |
| Nov 2025 | 89.5% |
| Dec 2025 | 93.2% |
| Jan 2026 | 92.8% |
| Feb 2026 | 93.7% |
| Mar 2026 | 91.9% |
| Apr 2026 | 92.4% |
| May 2026 | 94.2% |
| Jun 2026 | 93.6% |
| Jul 2026 | 92.0% |
| Aug 2026 | 89.9% |
| Sep 2026 | 89.9% |
| Oct 2026 | 92.0% |
Who builds those blocks? On June 1, 2024, beaverbuild filled 49.2 percent of slots and Titan Builder filled 24.2 percent11. By October 1, 2026, Titan filled 47.9 percent. Quasar filled 19.1 percent, and beaverbuild filled 0.2 percent.
Flashbots itself stepped back from building. On December 5, 2024, it shut its own builders and moved its order flow to BuilderNet, a shared builder network9. In February 2026 it ranked block building in phases. Phase 0 is "Fully centralized, monolithic systems running on a company's cloud provider," Flashbots wrote10.
Share of Ethereum slots each builder filled, on the first day of six months, in percent
- Titan Builder
- beaverbuild.org
- Quasar
- Height: Share of slots, percent
Source: mevboost.pics [11].
Read the chart as a table
| Item | Titan Builder | beaverbuild.org | Quasar |
|---|---|---|---|
| Jun 2024 | 24.2% | 49.2% | 0.0% |
| Dec 2024 | 45.6% | 42.2% | 0.0% |
| Jun 2025 | 34.0% | 28.8% | 4.5% |
| Dec 2025 | 47.4% | 3.8% | 10.5% |
| Jun 2026 | 52.4% | 0.3% | 10.8% |
| Oct 2026 | 47.9% | 0.2% | 19.1% |
The case for Flashbots
Flashbots argues that daylight beats a hidden market. “Flashbots exists to fight the centralizing impacts of MEV by producing open access and open source tools and lead the field of MEV research,” it wrote in October 20225. Before Flashbots, bots fought in public fee wars, the 2019 paper found1. Flashbots let anyone send a bundle of trades straight to block producers through a sealed-bid auction4.
Its 2022 post framed MEV-Boost as a tool for newcomers. “MEV-Boost is a direct attack on any centralized, vertically integrated block building entity, to let new block builders and new specialized entities that wanted to participate in the MEV game to freely enter and openly compete,” Flashbots wrote5. Flashbots also said the design "reduced the barrier to entry for new block builders".
The group points to its open data. Insights from its MEV-Explore and Transparency dashboards “have been cited in most of the MEV research and publications in the past 2 years,” Flashbots wrote5. The mevboost.pics dashboard behind both charts here credits its data to Xatu Data11.
Competition shows up in the counts too. In its October 10, 2026 update, the dashboard listed 24 builders and 7 relays active over 14 days11.
The case against
Anton Wahrstätter of the Vienna University of Economics and Business studied the builder market with four co-authors in May 20237. Their count ran to 1,624,885 blocks through May 1, 2023, and outside builders made 1,312,852 of them. The team posted its raw data at mevboost.pics. “MEV-Boost was explicitly designed to foster decentralization, causing the side effect of enabling risk-free sandwich extraction from unsuspecting users, potentially raising concerns for regulators,” they wrote. A sandwich trade buys just before your trade and sells just after it.
Lioba Heimbach's team at ETH Zurich looked at the same setup in May 20238. “We study PBS adoption and show that the current landscape exhibits significant centralization amongst the builders and relays,” they wrote. PBS, or proposer-builder separation, is the split between the validator who signs a block and the firm that fills it.
The 2026 numbers fit that reading. Added together, Titan's and Quasar's shares come to 67.0 percent of slots on October 1, 202611. Flashbots' own launch essay named that outcome as the risk in 20204.
What to watch
Ethereum's next upgrade, Glamsterdam, lists EIP-7732, which writes proposer-builder separation into the protocol12. The EIP's table dates Glamsterdam's start on the Sepolia test network to 13:53 UTC on October 6, 2026. Its mainnet row is still blank.
Flashbots' case needs that upgrade to spread block building across many firms. Wahrstätter's case gains if the top two builders keep two thirds of slots after it ships11.
Sources
- Philip Daian, Steven Goldfeder, Tyler Kell, Yunqi Li, Xueyuan Zhao, Iddo Bentov, Lorenz Breidenbach and Ari Juels, "Flash Boys 2.0: Frontrunning, Transaction Reordering, and Consensus Instability in Decentralized Exchanges," arXiv, April 10, 2019, https://arxiv.org/abs/1904.05234
- Cornell Tech, "Bots Exploiting Blockchains for Profit," April 29, 2019, https://tech.cornell.edu/news/bots-exploiting-blockchains-for-profit
- Yogita Khatri, "'Predatory' Bots Exploiting Decentralized Crypto Exchanges: Report," CoinDesk, May 7, 2019, https://www.coindesk.com/markets/2019/05/07/predatory-bots-exploiting-decentralized-crypto-exchanges-report
- Flashbots, "Flashbots—Frontrunning the MEV Crisis," November 23, 2020, https://writings.flashbots.net/frontrunning-mev-crisis
- Flashbots, "MEV for the Next Trillion, It's Time to Get Serious…," October 11, 2022, https://writings.flashbots.net/mev-for-the-next-trillion
- Laura Shin, "The Chopping Block: Why the Once-Taboo MEV Is Now a Core Part of Ethereum," Unchained, January 28, 2023, https://unchainedcrypto.com/the-chopping-block-why-the-once-taboo-mev-is-now-a-core-part-of-ethereum-ep-449/
- Anton Wahrstätter, Liyi Zhou, Kaihua Qin, Davor Svetinovic and Arthur Gervais, "Time to Bribe: Measuring Block Construction Market," arXiv, May 25, 2023, https://arxiv.org/abs/2305.16468
- Lioba Heimbach, Lucianna Kiffer, Christof Ferreira Torres and Roger Wattenhofer, "Ethereum's Proposer-Builder Separation: Promises and Realities," arXiv, May 30, 2023, https://arxiv.org/abs/2305.19037
- Flashbots, "Migrating to BuilderNet," December 9, 2024, https://writings.flashbots.net/migrating-to-buildernet
- Flashbots, "decentralized building: wat do?," February 14, 2026, https://writings.flashbots.net/decentralized-building-wat-do
- Toni Wahrstätter, "MEV-Boost Dashboard," mevboost.pics, daily data from June 1, 2024, read October 13, 2026, https://mevboost.pics/
- Tim Beiko, Alex Stokes, Ansgar Dietrichs, Nixo and Parithosh Jayanthi, "EIP-7773: Hardfork Meta - Glamsterdam," Ethereum Improvement Proposals, read October 13, 2026, https://eips.ethereum.org/EIPS/eip-7773