Proof-of-Stake consensus protocols give rise to complex modeling challenges. We analyze the recently-updated Tezos Proof-of-Stake protocol and demonstrate that how sometimes rational participants are incentivized to behave dishonestly.