That would be a really shit poll though, your methodology would be “I found people who I knew would answer the question this way” and everyone would see. If there is a methodological issue with this poll that isn’t hand-waving I’d be interested to hear it.
E.g. “but who answers the phone anymore?” They know this issue and use a bunch of complicated statistical shit to adjust for it or just improve the methodology. The people making these polls actually tend to know what they’re doing.
Most moral army in the world