Thread by @project_shutter

As we're seeing the earliest signs of #MEV on #L2 taking off, let's take a minute to consider some perceived design/architecture choices and trade-offs which L2 teams have to consider.🧵

https://twitter.com/project_shutter/status/1524702364543897606

1) Balance the needs of your stakeholders:
- End users don't want to get sandwiched, users need to be adamant with setting (max) slippage

- Arbitrageurs bring volume/activity to the rollup and can help with eliminating pricing/timely liquidations etc.

2) Trust assumptions:
- if you can trust the sequencer not to front-run, things are fine,
- if you don't they're likely not
- but: MEV protection mechanism may introduce additional trust assumptions

3) Impact on UX/DevEx (e.g., latency, EVM/tool chain compatibility)
Most future-proof MEV solutions require changes to the protocol and these in turn have impact on users and on developers. They might affect latency, modify data formats, change transaction flows.

4) MEV Distribution vs. Minimization

Dealing with MEV should have three goals: Minimize it as much as possible, maximize extraction efficiency as much as possible, and make sure profits are allocated effectively.

Good news: Turns out, not all of these are trade-offs: A shutterized Rollup effectively eliminates the headaches by minimising malicious MEV as much as possible, while leaving the distribution of the remaining ("benevolent") arbitrage to the rollups chosen distribution mechanism.

It does this all while exerting minimal impact in terms of technological complexity on the rollup side and in terms of UX for end users as well as DevEx for dapps building on top of the rollup.

Read more here:

https://twitter.com/project_shutter/status/1511036526943248388

And watch @AnotherJannik's presentation at MEV day here:

https://twitter.com/project_shutter/status/1527284808254038016