But even when someone doesn't vote, that doesn't mean they aren't Maga?

So if half of voters are Maga and 1/3 of eligible voters actually vote, it's almost impossible to say something about the population as a whole. We can say about 16% would be the lower limit, but it could be a lot higher. If we take the voters as an unbiased large sample, we could extrapolate and say 50% of the population is actually Maga. But since voters are by definition a biased sample, it's hard to say what the actual number would be. Especially with humans, that have complex interactions, like a certain persuasion could actually mean someone is less likely to vote. Or the other way around, where wanting to vote makes one persuasion more likely. This makes the whole thing pretty hard.