Nobody explains calibration to first-time managers before their first one, so it lands as a surprise: a room of managers comparing ratings across teams, sometimes moving someone's review score up or down before it ever reaches the employee. If you walk in without understanding what's actually happening, you'll either get talked out of a rating you had good reason to give, or you'll over-defend one you shouldn't.
The purpose of calibration is consistency, not to relitigate your judgment. A "high performer" on one team and a "high performer" on another team should mean roughly the same thing across the org, and calibration is where that gets checked. That's a legitimate goal. Where it goes wrong is when a rating shifts because you're a less forceful advocate in the room than another manager, not because the case was actually weaker.
When someone challenges a rating you believe in, respond with the specific evidence, not with volume or repetition. "Here's what this looked like in practice: [example]" holds up better than "I really think this person deserves it." If the pushback reveals something you genuinely hadn't weighed, say so, changing your mind with new information is different from being talked out of a well-supported call.
"I hear the concern, here's specifically why I rated it this way: [concrete example, with impact]. I'm open to hearing what I might be missing, but based on what I've seen this quarter, I still think this is the right call."
Calibration is a comparison exercise. Your job is to make sure the comparison is based on the work, not on who argued harder in the room.
One thing worth doing after every calibration, win or lose: write down what actually moved a rating, yours or someone else's. Over a few cycles, that record teaches you what the room actually values, which is usually more useful than anything written in the official rubric.