A tactic is meant to be tried, measured, and — if it does not work — discarded, without anyone altering the strategy above it. From this follows the single most useful rule about tactics: a tactic should be easy to abandon. The moment you find yourself defending a particular action because giving it up would feel like a loss, a confusion has occurred — you are protecting the method rather than the goal it was meant to serve.
Skill in tactics is therefore not what it looks like from the outside. It is not mastery of any particular move but fluency in producing and discarding moves quickly — keeping the loop turning rather than becoming attached to one turn of it. The same test sorts levels in disguise: an action no one will give up has quietly been promoted to a matter of principle, and the honest response is to ask whether it deserves the promotion — not what is it, but what is it in service of.
See also: Policy, Strategy, and Tactics, Mistaking Activity for Direction.