> For the complete documentation index, see [llms.txt](https://knoshua.gitbook.io/analysis-of-odao-duties/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://knoshua.gitbook.io/analysis-of-odao-duties/guardrails-mev-penalties.md).

# Guardrails - MEV Penalties

The current version of [MEV Penalties](/analysis-of-odao-duties/overview-of-duties/mev-penalties.md)allows any number of penalties to applied to any number of minipools. But there is no need for unlimited penalties. There are 50,400 blocks per week, so that is a guaranteed upper bound. If we assume that [RP will not represent more than 33% of the network](https://rpips.rocketpool.net/RPIPs/RPIP-17) then (with 99.9% probability) RP will not get more than 16,900 proposals per week. Using the current market share of \~2.45% gives less than 1300 proposals per week (with 99.9% probability).

So it should be possible to enforce a maximum number of penalties in a given interval to reduce impact of malicious behavior while still allowing for legitimate use of the penalty system. This [proof of concept](https://gist.github.com/knoshua/0795746fe4171349eaed2349f4563f5e/revisions) limits penalties to 2500 per week. Again a (timelocked) pDAO setting might make sense here.

Completely penalizing all minipools would currently take \~156,000 penalties (\~13k minipools, 2 strikes and 10 1.6 ETH penalties). Using the limit of 2500 penalties/week would reduce value at risk to less than 3.5% of node operator ETH.

{% embed url="<https://gist.github.com/knoshua/0795746fe4171349eaed2349f4563f5e>" %}