Re-examining EIP-1559: 2 years after the proposal landed, is Ethereum safer?

By James Prestwich

Compilation: Luffy, Foresight News

Before EIP-1559

A long time ago, in the crypto world before the arrival of EIP-1559, miners decided block sizes through a simple voting mechanism. Each block, miners can move the block gas limit up or down by up to 1/1024. Therefore, when a miner produces a block, the gas limit will move towards the hash rate weighted average preferred by the miner. This mechanism has existed since the birth of Ethereum.

With the advent of EIP-1559 the gas limit was doubled, but the way to set the limit was gone. Between post-EIP-1559 and pre-merge, miners voted on the gas target, and the gas limit was defined as twice the target. For example, if the limit before 1559 is 10mm (note: mm refers to 1 million), and the target after 1559 is 10mm, then the limit after 1559 is 20mm. 1559 provided an effective block size increase, but with serious consequences (in the form of base fee adjustments). Miners in the 1559 system can vote on the gas target, which is no different from the previous adjustment system that moved 1/1024.

Now 1/1024 (0.09%) may seem small, but it means about 5% increase every 50 blocks, or about 5% increase every 12.5 minutes, or doubling every 3 hours. Suppose you take a nap and suddenly wake up to find that your Gas limit has been doubled (or halved)!

Brief review

EIP-1559 has 2 inputs and produces 2 outputs:

The block producer sets the target gas by adjusting the target by 1/1024 for each block. In this way, the Gas limit (2x the target) is set indirectly. Block producers set the base fee by combining a Gas target with actual Gas usage. When actual > target, the basic cost increases; when actual < target, basic cost decreases. There is a complex elasticity relationship here.

It's important to remember that target usage has no impact on actual usage. In fact, each block producer sets the target (through the adjustment mechanism) and the actual usage (by including transactions when building the block). Effectively, each block producer controls both inputs to EIP-1559 and therefore can control outputs (for a particular block). The target in any given block is the result of the producer's latest move in an infinitely iterative "set gas target game". Thus, over time, the behavior of the average converges to the equilibrium value in this game.

EIP-1559 and merge

Merging has two main impacts on the Gas limiting mechanism:

• Due to a slightly shorter block time, the gas limit increase period of 5% has been shortened to 10 minutes. This means that approximately every 2.4 hours, the gas limit has the potential to double (or halve).
• Miners are replaced by "stakers" or "validators" or others.

The first further speeds up the already fast compounding process, but perhaps that doesn't matter.

I think the second one is more important. But not in isolation, but as part of a larger systemic trend. We'll come back to this later.

What’s holding back block targets from rising?

You might be wondering, why isn't the block target going up? Wouldn't it make transactions cheaper for users and make more money for validators? If these things look good, why is the target 15mm and not 15,000,000mmm? Why is the 15mm target (30mm limit) the current balance?

The main reason is that it has always been set up based on suggestions from the core developers. The general consensus is that we set it to a level that we know users can handle in a reasonable amount of time. Huge high-Gas blocks will DoS nodes and prevent confirmations. Validators (formerly miners) must construct blocks and validate other people’s blocks. And if the block is too large to be reliably constructed or verified within the 12-second block period, then the validator will lose money when confirmation fails. This can also create network instability as nodes may not be able to handle gigabit blocks well.

EIP-1559 mentions this in a fairly brief security section, considering security related to block size:

This EIP will increase the maximum block size, which can be problematic if miners cannot process blocks fast enough, as it will force them to mine empty blocks. Over time, the average block size should remain about the same as without this EIP, so this is only a matter of short-term capacity bursts. One or more clients may not handle short-term capacity bursts and errors (such as out-of-memory or similar) well, and client implementations SHOULD be able to correctly handle the maximum capacity of a single chunk.

We tend to believe that validators have enough profit motivation to keep the block limit within a reasonable range and that the 15mm target is reasonable given the customer situation.

EIP-1559 Is it still safe to use PBS?

MEV-PBS changes these circumstances. With PBS, proposers no longer pay block construction costs directly. They outsource the entire process to builders. In fact, common MEV-PBS structures (such as Flashbots relays) hide a block from a validator until it commits to proposing the block. As part of the building process, validators communicate their preferred gas limits to relays.

The builder uses specialized, optimized software (not an off-the-shelf client) to construct a block, and then delivers the block (with gas limit adjustments) to the relay. The relay ensures that the builder's block meets the proposer's requirements.

This brings up an interesting situation. Proposers are now prohibited from paying block construction fees. Builders pay these fees exclusively and don't have to worry too much about larger gas limits. So why should gas targets be kept low in a MEV-PBS world?

It’s not safe yet

Proposers are always motivated to increase the gas target infinitely to eliminate base fees, but are limited by the cost and risk of doing so. Builders, on the other hand, want to confirm blocks and are discouraged from building blocks that are too large to be verified within a confirmable time period. In other words, builder profitability is walking on a tightrope. They need to carefully balance building larger actual blocks to get more fetches versus the effect of block size on finality. If builders generate blocks that are too large for validators to validate, they risk losing the entire block reward. This means that builders have a strong incentive to keep the actual block size at a reasonable verifiable size regardless of the gas target.

Proposers, on the other hand, do not need to consider such a balance. They can simply increase the gas target for protocol execution indefinitely and let builders find a profitable balance. Doing so eliminates the base fee while keeping the actual block space limited.

By raising the gas target (and the protocol-enforced limit) while keeping the actual block size within the normal range, not only were proposers able to extract more MEV with anomalous block sizes, they also eliminated base fees while artificially limiting supply. If builders keep blocks smaller, users will have to pay gas tips, returning to the pre-EIP-1559 fee market. In effect, they could patch EIP-1559 by "hypothetically" increasing the block space, and then deny anyone the ability to use it.

EIP-1559 appears to be a clear win for proposers and builders. They can either extract more MEV for themselves in edge cases (by occasionally building huge blocks) or convert all base fee burning into tips paid by artificially limiting the supply. Moreover, the cartel is incentive compatible because neither party appears to gain from the deviation.

This brings me to the central question of this article.

When will the proposers and builders increase the Gas target to 100mm+?

Given the constraints on confirmability, builders should keep the actual block size close to the current value regardless of the target size. Some reasonably small blocks can maximize profits. However, for proposers, some ridiculously high gas target can maximize their profits. Since proposers and builders have different positions, they can collude to increase the gas target while reducing the actual size to maximize profits.

As long as the actual gas of the block is kept around 15mm, isn't the 15000mm Gas target more incentive compatible than the 15mm Gas target?

Therefore, whether it is a block builder or a validator, pushing up the target Gas, but keeping the actual blocks generated smaller, will make the income higher.