Sometimes assumptions have to be made especially in this case.

The AEC policy is to cast aside your vote so it doesn't count towards the election - e.g. if you write your name or signature on the ballot paper, it will tossed out when they count the vote. If you take a selfie with your ballot paper showing how you voted, they are also required to discard your vote.

Reliable data on exactly how people voted is not available and cannot ever be available (under current laws)... so, yes, pretty much anything relating to the election is based on assumptions. That doesn't make them useless. I'm assuming the sun will rise tomorrow morning... I don't have any proof, but a lack of proof doesn't make my assumption wrong.

The data they provided is pretty compelling - for example the Wadeye electorate where 92% of the vote was Yes. And by the way, about 90% of that electorate were are indigenous in the 2016 Census (pretty old now, but it likely hasn't changed much).

What's also interesting is those communities bucked other major trends in this election:

• Wadeye is about as far as you can possibly get from a major city and still be on the mainland... Adelaide is the closest proper city — 3,230km from Wadey. In general the further from a city people are the more likely they are to have voted No.
• I'm making an assumption here, but based on the fact they are so far from the nearest university there probably aren't many university educated people in Wadeye... and people with no university education were also more likely to vote No.

If indigenous and white people shared the same opinion on the vote, then you'd expect Wadeye to have one of the strongest No votes in the country. But nope - they voted 92% yes.

get them to retract it if you're so sure.