Why Is Everyone Still Playing Planning Poker?

That's part of the problem: everyone has their idea of what Planning Poker, or any other practice in Agile, is supposed to be and whether it works. Some people will say you are doing it wrong because you use time as story points, and others will tell you that you shouldn't use days but hours.

There are too many assumptions: for example, we assume that the group estimate is more correct as if everyone had the same knowledge of the code base, everyone was at the same skill level, and technical debt will not slow us down.

The story was supposed to take two days, but it took the whole week in practice. Whose fault is it? Is the estimate updated, or does it stay the same? And if it needed to be more accurate, what use was it?