If you have only one poll with an unknown margin of error and it shows Candidate A at 51 and Candidate B at 49, then assuming Candidate A is winning just because it's the best thing we have is worse than just admitting we don't know.

Wrong information is often worse than no information.

That's one of the problems with PISA results. Policy makers look at other countries with higher scores and try to emulate them without understanding why their scores are higher. In that case it's just cargo culting education.

This doesn't mean that standardized testing is completely useless, but you have to know the limitations, and comparing results between cultures is one of the limitations.