A fascinating aspect of cryptoeconomics: critically-important code that ~never executes.
- Slashing
- Coin voting vetos
- Rollup fraud proofs
In each case, the mechanism ensures an attack is -EV > no attack > countermeasure not exercised. Hard problem to ensure it's still live & working.

Real World Ethereum 🌎 
building /daimo ✳️ 

For something like L1 validator slashing, there's enough background noise of people accidentally running broken setups to ensure we have working slashers.

For eg a fraud proof on a less-popular optimistic rollup... IDK

My wife had this q about Optimistic L2s. 

If L2s are optimistic and people assume they will work 99% of time, will people get lazy and just assume it’s correct with no one actually submitting fraud proofs.