How do you explain merge vs rebase, and when would you use each in a team workflow?

I choose merge or rebase from collaboration and traceability needs. Rebasing an unshared branch can remove fixup noise and present a reviewable sequence, but rewriting a branch others consume changes commit identities and creates coordination risk. Shared main and release history remain immutable. The merge strategy, squash policy, signed-commit requirements, and pull-request link are consistent enough that a production artifact can be traced back to reviewed work. During incident analysis I care more about reliable commit and deployment identity than a perfectly linear graph, so exceptions to the normal policy are rare and explicitly coordinated.