The cards are a gimmick. The actual mechanism is forcing silence until everyone has already committed to a number, so no single voice anchors the room before the rest of the team has thought for themselves.
Anchoring is the real problem planning poker solves
In a normal estimation discussion, the first number spoken pulls every estimate after it toward itself — even from people who privately disagreed. This isn't a discipline failure; it's a well-documented cognitive bias that happens regardless of how experienced or senior the team is. Planning poker's actual function is structural: everyone reveals their estimate simultaneously, so the anchor never gets planted before independent thought happens.
Disagreement is the valuable output, not the inconvenience
When the revealed cards span a wide range — a 2 and an 8 for the same task — that's not a process failure to smooth over quickly. It's the most useful information in the room: someone sees complexity or risk the others don't, or someone else is missing context the others have. The conversation that follows a wide spread is worth more than the conversation that follows quick consensus, because it's where hidden assumptions actually surface.
Relative sizing beats absolute hours
Story points work better than hour estimates partly because "is this bigger or smaller than that other task we did" is a question humans are reliably good at, while "how many hours exactly" is a question humans are reliably bad at, even with experience. Planning poker's point scale leans into the comparison teams are actually capable of making accurately.
It breaks down the moment it becomes a formality
Planning poker loses its value the instant a team starts converging on the same number out of habit rather than independent judgment — which happens naturally once everyone's learned each other's typical estimates. If sessions stop producing spread, it's worth occasionally reshuffling who estimates what, or explicitly asking "what would make this an 8 instead of a 3" to re-introduce the independent thinking the format depends on.