Toni Wahrstätter ⟠
@nero_eth
Followers
5,560
Following
530
Media
199
Statuses
1,219
Research & Data at Ethereum Foundation |
Vienna, Austria
Joined December 2021
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
İran
• 1562367 Tweets
Netanyahu
• 780771 Tweets
The ICC
• 345651 Tweets
International Criminal Court
• 171705 Tweets
Cauca
• 93071 Tweets
Merchan
• 79829 Tweets
Bibi
• 40702 Tweets
Costello
• 36951 Tweets
Red Lobster
• 34127 Tweets
Yildiz
• 30314 Tweets
#KızılGoncalar
• 28392 Tweets
Montero
• 23864 Tweets
Amal Clooney
• 21844 Tweets
Braves
• 21251 Tweets
Kid Rock
• 20937 Tweets
Milli Yas
• 16200 Tweets
ETH ETF
• 14592 Tweets
Zeynep
• 14460 Tweets
BIA GIL
• 14068 Tweets
Memorial Day
• 13952 Tweets
NICOLA EN HOY
• 12962 Tweets
Henrique
• 10711 Tweets
Pinned Tweet
Happy to share that our paper on Stealth Addresses got published!🎉
We introduce a standardized Stealth Address protocol for Ethereum (ERC-5564), enhancing privacy with unlinkable transactions.
9
22
128
The seamless journey of blocks, from creation to validation, through the mev-boost ecosystem:
21
66
314
With EIP-1559, zero gas transactions are no longer possible.
Together with
@lightclients
and
@gballet
, we reintroduced them by implementing a PoC for a gas ticketing service.
It uses blind signatures to fund unfunded accounts in a privacy-preserving way.
19
52
306
New dashboard online - 🙌
May it shed light onto the extent of censorship and guide us towards implementing practical solutions for enshrining a strong form of censorship resistance.
More info 👇
16
106
260
I'm excited to share our latest research on adjusting the block gas limit.
Together with
@VitalikButerin
, we're exploring paths to reduce the maximum block size, balance calldata pricing, and ultimately raise the block gas limit.
5
78
272
Data shows that the Flashbots MEV-Boost Relay (same for the Flashbots builders) is censoring
@TornadoCash
transactions.
Since the Merge, not a single transaction interacting with the Tornado Cash contracts has been included in their MEV-boosted blocks. /1
21
60
253
Should we increase Ethereum's block gas limit?
To guide this decision, I've just published an analysis focusing on block sizes, gas limits, and scalability:
24
47
246
Excited to share our latest research on blockchain privacy, specifically focusing on stealth addresses!
Special thanks to
@msolomon44
,
@BenDiFrancesco
and
@VitalikButerin
for the great collaboration.
Available at:
Looking for a TL;DR?
There you go👇
11
67
244
Recently, I've been looking into the data Ethereum is producing post-merge, and it's great to see all the new opportunities for data science research.
I've summarized my experience with analyzing MEV in a blog post in hopes of helping others get started /1
5
58
237
P2P org announces that it will call getHeader later in the slot, trading off chain stability against more MEV profits.
This is exactly the centralizing behavior that we shouldn't tolerate, as long as we, as users, have the choice about where to stake our assets.
1/2
16
56
234
Excited to launch !🎉
Recently, there's been a growing discussion within the community about timing games. These games can lead to a centralized playing field, and being unaware of them can be equally restrictive.
I hope the site guides us well!
16
56
219
Quick Fact. Although commonly called flowcharts, their actual name is Sankey diagrams.
Introduced in 1869 to show the movements of the Napoleonic army during the French invasion of Russia (1812).
Now we can use them to "illuminate" the MEV-Boost Builder -> Relay block flow.
9
29
181
In the last 30 days, Arbitrum, Optimism and Scroll together posted over 10 GB of data to Ethereum L1.
That's more calldata than all non-rollup users together.
Dencun is coming: March 13, 2024 at 13:55 UTC.
5
40
186
It was a pleasure to present ERC-5564, a Stealth Address standard, to the wider community at
@EDCON2023
.
What began as a discussion about various stealth address protocols turned into an effort to create a unified base that works for many different stealth address protocols.
14
33
177
Updated and beautified the charts to MEV-Boost!
Note: The MEV Boost Relay API was used for the data; No data of manifold yet.
8
32
178
Today, 4 relays hold 18-30% market share each.
Rsync builders' share climbed from ~10% in early March to stable 15% now.🌟
No builder has more than 25% market share;
4 (out of ~35) are at 15-25%.
Find chart at:
4
31
177
Excited to publish a new dashboard: .
The aim is to add more transparency to Ethereum's mempool. The site shows you all kinds of stats around private order flows and the actors involved.
Private order flow has grown more and more into a highly important
13
43
166
70% of the total calldata comes from rollups publishing data. Damn ready for blobs.
7
22
141
Happy to launch 🎉
The site provides live data on calldata and blob usage. Furthermore, it comes with visualizations on rollup data, block sizes, blob inclusion, and more.
This is still v1 and I plan to add more graphs in the near future.
10
43
143
For those interested in Ethereum data analysis, I've open-sourced a data set with validator pubkeys mapped to labels (Lido, Kraken, Rocketpool, etc. including the individual Lido Node Operators).
You can access it at (updated on a daily basis).
The
11
32
138
The Flashbots relay dropped to <80% market share while
@bloXrouteLabs
max-profit is slowly catching up.
In the meantime, we saw two new relayers (of
@GnosisDAO
and
@ultrasoundmoney
) launching.
Relay diversity improving, check.
9
24
133
Solo stakers miss relatively more slots than professional NOs.
But it's not performance that a blockchain such as Ethereum demands from solo stakers.
It's decentralization, censorship resistance, and credible neutrality.
Instead of relying on pools, we must find ways to support
18
22
131
Latency is an important factor in the MEV game.
In order to unlock a latency edge for builders, the
@ultrasoundmoney
relay introduced optimistic relaying.
Instead of simulating the blocks received from builders, to verify that they adhere to the rules, optimistic relaying
9
26
130
With the latest hard fork, beacon chain block roots became accessible in the EVM, enabling proofs against consensus layer data.
I hacked on this proof of concept for verifying a validator got slashed.
More details and the code in this post:
5
28
133
That's what Lido's 26% block market share entails:
A healthy ecosystem within the ecosystem itself.
9
16
116
The data behind is now Open Source and available almost in realtime.🎉
The first published table contains info on every slot since the merge, MEV-Boost and the mapped block builders, validators, and relays.
More tables to come.
10
26
113
We are excited to share the recent publication of our research paper, "Time to Bribe: Measuring Block Construction Market", a comprehensive analysis of the block builder and proposer landscape on Ethereum.
Working collaboratively with
@HatforceSec
,
@lzhou1110
, and
@KaihuaQIN
, we
14
27
107
Block building is becoming increasingly competitive:
@builder0x69
has overtaken all Flashbots builders in the number of blocks built in the last 14 days.
It also overtook Flashbots' "0x..beef" builder in total number of blocks built, moving into the top 3 block builders. /1
9
18
101
Private Order Flow Sankey.
Private OF txs bypass the public mempool, going directly from the searcher/user to the builder.
JaredFromSubway produces the highest number of private OF, sending to every major builder.
Except from that, we see a lot of vertical integration.
12
18
103
Blobs now constitute a significant portion of the data that validators are handling.
As the usage of blobs began to increase, the trend in calldata usage started to decline.
9
18
102
Most rollups post regularly over the slots of an epoch.
Arbitrum prefers to post around slots shortly after the 2/3 threshold.
This is advantageous because blocks will get finalized in less time.
7
14
98
A censoring builder at 40% market share🚨
Right now, we're essentially relying on a single builder with 26% market share (Titan builder) to include OFAC sanctioned content.
Inclusion Lists may help: let's face this problem and ship them in Electra.
8
13
97
Reorged blocks present some challenges.
To provide a clearer overview of reorgs and the entities involved – such as CL Clients, Validators, Relays, and Builders – I've created a quick dashboard.
You can check it out at .
Reorgs not only cause the total
1
11
89
now displaying the avg. MEV-Boost payment per relay/builder/validator over different time frames in boxplots:
12
20
87
What do Ethereum users do?
* ~20k transactions sending ETH every hour
* ~10k token transfers
* execute is the second most executed Solidity method, after transfer and before approve.
3
20
90
The RANDAO plays an integral role in Ethereum, providing the necessary randomness for the selection of proposers and committees.
I looked into the possibility of RANDAO manipulation and carried out a variety of simulations using historical data.
Full post:
2
14
86
New ethresearch post online.🔍
I summarize some findings from analyzing a simulated post MEV-Burn world.
6
19
84
Relay diversity improving:
Flashbots' relay share reduced to 50%. Ultrasound & Agnostic relays saw big gains. Major builders such as builder0x69, beaverbuild and Flashbots have contributed to the improved diversity by submitting blocks to those new relays.
5
18
85
It's not the validators. Many validators using mev-boost would propose an uncensored block as long as it outbids the censored block.
This is important to not mix up.
10
8
80
Why don't we have finality within a few seconds already?
Reading through
@benjaminion_xyz
's Eth2Book, you'd find out.
If you'd like to have a quick TL;DR, let me try:
Ethereum is a monolithic blockchain, meaning that all validators process all information.
This means, every
4
21
79
In light of the ongoing concern about censorship, I've summarized some findings about censorship of builders in a blog post:
2
23
81
EIP-7623: Increase calldata cost:
* reduce max. possible block size & variance without impacting regular users
* reduce inefficiencies between average & max block sizes
* make room for block gas limit & blob count scaling
4
20
76
The very top of a block (ToB) is the most competitive place for transaction inclusion. Searchers compete to get their bundles in as early as possible.
The chart below shows a distribution of indices for private order flow transactions.
It's easy to spot the searchers.
3
12
77
Improving Ethereum's decentralization through anti-correlation penalties.
In this post, I analyse correlated attestation penalties and their impact on today's staking landscape.
Are validators in the same cluster (eg same exchange, same user) more likely than unrelated validators to miss attestations at the same time? If so, can we tweak rewards to favor decentralized staking?
Possibly yes.
705
450
3K
2
24
79
Just deployed a bunch of new stuff to .
You'll now find...
* a general metrics tab,
* separate tabs for relays and builders (validators to come),
* sankey diagram with builders, relays and validators
* and many new charts (incl. racing bar charts)
4
21
74
Probably the best diagram of the inner mechanics of PBS that I've created so far.
6
9
76
Wacky Races 2.0:
Successful payloads delivered since November;
Builders are aggregated based on block extra data.
3
11
76
A user lost 32 ETH by making a deposit to the ETH2 Deposit Contract with an invalid signature.
Stick to the official staking deposit cli to prevent such issues.
10
10
73
A subset of the largest validators, their relative size, their block building preferences and the relative number of OFAC incompliant blocks proposed.
4
17
66
Chart shows slot share of builders within MEV-Boost payment intervals (in ETH).
While builders such as eth-builder land many blocks with small MEV-Boost payments, , beaverbuild seems to leverage a competitive edge as MEV opportunities increase.
14
11
66
As part of the Advent of Crypto Data, I'm happy to announce the DotPics Data API 🎁.
The API will provide access to all the data I've been collecting over the past years, which I've already been using to build various .pics dashboards.
The goal is to empower others to harness
10
19
66
As of today, more than 150k ETH (~ $ 300m.) in MEV has been captured by validators since the merge.
12
6
63
Solo stakers are missing their slots mainly because they don't show up (=are offline) and are not reorged significantly more often than other (arguably bad) validators. Same applies to Rocketpool.
The actual surprise is Coinbase, though.
14
12
66
Considering the relevance of timing strategies, I've expanded on the number of validators displayed on .
This update aims to enhance our ability to track which validators are engaging in zero-sum timing games and which ones aren't.
11
8
63
300k ETH (~480m USD) distributed through MEV-Boost so far.
Money that would have gone to the searchers and builders pre-PBS, now flows to validators securing the chain.
7
16
61
Analyzing the % of private order flow transactions for the past week among the 15 largest block builders, sorted by block count (desc).
The "Flashbots Impos..." (Imposter) is a builder that tries to impersonate Flashbots.
Mempool data is courtesy of
@blocknative
.
8
13
58
I've recently added new charts and features to :
- builder racing bar chart - tracking successful payloads over time
- avg. mevboost payment per builder over time
- builder aggregation now based on block extra data + known pubkeys.
2
14
58
We've heard a lot about protocol-level censorship in recent weeks.
To shift the focus from censored blocks to uncesored ones, I put out .
It may help validators identify the best (non-censoring) relays. Stay safe out there.
6
13
59
Yesterday,
@mikeneuder
and
@gakonst
published this great article on PBS and how it impacts the fork choice rule.
Here are a few notes (and a new diagram) that may be helpful to better understand the situation:
👇
it was a pleasure to work with
@gakonst
to put together this piece and present some of the nitty-gritty details of Ethereum's fork-choice rule! if you are curious about how latency can impact consensus, check it out 📖⬇️
5
15
113
1
10
57
Privacy is normal and blockchain users have a right to privacy.
As a community, we shouldn't accept services censoring innocent users.
68
180
752
0
12
60
Validators now displayed on .🎉
You can get the stats of the individual
@LidoFinance
node operators by clicking the pop-up icon in the table.
6
9
57
Quick reminder of why client diversity is critically important:
This bug would have caused Geth nodes to reorg out valid blocks produced by other clients, resulting in honest validators losing their rewards.
On Feb. 6th
@nero_eth
and
@parithosh_j
sent a transaction on Sepolia which uncovered a bug that could've happened on mainnet. The bug was quickly fixed in the clients and now, after all nodes updated to Dencun, we can finally disclose the issue:
9
41
203
2
14
57
The OFAC-imposed sanctions have been a widely-debated topic in 2022 and the discussions are still ongoing.
Together with peers, we've worked on a paper titled "Blockchain Censorship", in which we analyze state-imposed sanctions and their impacts on Bitcoin and Ethereum.
↓↓↓
7
15
55
Executable Specs for ERC-5564, Stealth Addresses:
2
10
56
Charts featuring validators now available on .
Special thanks to
@hildobby_
for open sourcing the heuristics to map validator pubkeys to entities.
3
14
52
200k ETH of MEV distributed to validators since the launch of MEV-Boost.
7
16
52
New diagram on builder efficiency deployed to .
Efficiency is measured by comparing the avg. number of bids with the total number of blocks that make it on-chain.
The closer to the top left, the more spam.
The closer to the lower right, the more efficient.
4
9
52
The number of non-censoring relayers has increased recently. 🎉
So grew the map on !🌪️
Now also displaying the relayers of:
- Relayooor
-
@ultrasoundmoney
-
@GnosisDAO
-
@AestusRelay
5
15
50
It's cool to see an update being rolled out and achieving instant adoption.
4
9
52
Not only proposers profit from PBS.
For instance,
@beaverbuild
recently netted a juicy 36.8 ETH profit from block
#17
,497,511, after paying 1.8 ETH to the proposer.
Made possible by the searcher 0xA69babe..., who swapped millions into WBTC and WETH within 34 txs in that block.
9
9
45
now displaying the block building market over time.
Vanillia Builders includes multiple individual stakers not using mev-boost; aggregation is based on block extra data and known pubkeys.
2
5
49
It's great that we're talking about gas limits, block sizes, and the blob count.
Today, the avg. block size is around 125KB.
The maximum possible block size can be 2300% larger, namely 2.8 MB.
This provides much room for calldata, which is relatively cheap. Therefore, one can
Today,
@nanexcool
and I are launching an effort to help raise the Ethereum gas limit from 30mn to 40mn
This can result in a 15-33% reduction in L1 tx fees
We are calling on solo stakers, client teams, pools and community members to help
#pumpthegas
110
187
852
3
10
51
The number of MEV boosted blocks has increased to ~50% over the last few days.
We saw
@foldfinance
and
@bloXrouteLabs
continously increasing their share in relayed blocks.
Also
@blocknative
is coming very strong.
Flashbots still clearly in the lead.
Data: Relay/Data API
2
9
48
At the end of the day, almost 8k ETH were distributed to validators through mevboost.
Almost double the amout we saw during the FTX breakdown. 1/2
3
4
50
Typically, validator rewards remain quite consistent and predictable in the long term.
However, the introduction of MEV-Boost has added an unforeseeable element to this equation.
For instance, in the past 14 days, nearly 20% of validator revenue has been generated from
6
9
48
I have published some of the recent charts and will try to keep them up to date:
I am still looking for a place to host them.
Any suggestions or ideas?
6
4
44
Just deployed a new chart to 🙌
It shows the number of bids submitted per builder to relays over the most recent days.
The chart allows filtering for each individual relay, helping to uncover potential bugs.🐜
3
5
46
New post on "One-stop social recovery wallets" using zk-SNARKs:
TL;DR: By decoupling spending logic from asset management, handling multiple accounts becomes effortlessly streamlined and can be done without sacrificing privacy.
0
12
47
It's an ETHPrague hackathon project and consists of a Python CLI and a coordinator/ticketmaster, written in Go, that facilitates the ticketing service described in the post above.
Find the repo in the following:
2
3
44
Quick reminder: MEV is still real.
Chart shows the median proposer revenue per block over the last 30 days (15 Nov - 15 Dec).
4
7
45
We've reached the plateau (=target) of 3 blobs per slot while the EL payload part is continuing trending down.
10
9
47