Did you read the entire thing first?

https://www.snopes.com/fact-check/mail-in-ballot-voter-fraud...

What's True

While no U.S. government agency officially compiles state-by-state data on voter fraud, and requirements for mail-in voting vary by state, analysis by elections experts shows that fraud is slightly more common with mail-in voting than in-person voting at polling places.

What's False All types of voter fraud in U.S. elections is minuscule in comparison to the number of ballots cast, according to elections experts. Taking that into consideration, it is problematic to make comparisons between types of ballot-casting systems and erroneous to claim mail-in voting "substantially" increases the risk of fraud.

snopes? Really?

> experts shows that fraud is slightly more common with mail-in voting than in-person voting at polling places

So where is the line between slightly and substantial?

Where it makes a difference in the vote at hand? Or, more likely, well before that

How would anyone know what number of votes constitutes that difference until Nov 3?

Interesting thing to consider..

If fraud is more common with mail in voting and some states (or everyone?) converts entirely to mail in voting, how much will fraud increase overall?

Will it increase enough to change the overall results? With Michigan and Wisconsin being decided in 2016 by less than 1% of the vote, there's not much margin for error, fraud, or mistakes.

another issue I don't see brought up in generalist areas is electronic voting machines. closed source / unaudited / unauditable software in voting machines - what % of fraud exists in those, and how would we even tell? lots of posturing about 'mail in' stuff right now, but compared to electronic machines used in many districts, I'd still prefer mail-in paper ballots.

Agreed. The entire system and the people involved must be open for audit and review.

Imagine what happens when $countryX realizes that bribing a few mailmen is even more cost effective than misinformation campaigns?

By your reasoning, what he said was not true in the sense that it cannot be verified.

Right, neither claim can be falsified until after the fact, so why add a "fact check" ? We won't know the implications of large scale mail in voting in the US during a particularly charged election until after its happened

His claim wasn't that there will be some amount of fraud... it was that they won't be "anything less than substantially fraudulent" (https://twitter.com/realDonaldTrump/status/12652558351245393...).

Claiming mail-in votes will be majority fraudulent, and by implication that the entire vote is invalid... is a much stronger claim, which IMO requires much stronger proof.

Given that mail-in ballots have been in used for a long time, there's a good history of data, so it's not predicting the future out of nothing, but based on past evidence.

The twitter fact-check link in fact goes into that precise thing.

How many fraudulent votes constitutes a substantial amount? What percentage?

This is such a bizarre and useless take. So now I can claim that gravity will turn off tomorrow, and because you don't know the future you just have to sit there quietly and let me spread obvious misinformation?

Trump is making an extraordinary claim. He must back up that claim, whether that's by revealing that there's a true plot against him; referencing historical data; or something else.

They set up a commission in 2016 and found nothing so they closed it quietly. But they are still making the same claims. To me this shows that they have no interest in establishing hard facts. Trump says whatever benefits him as long as he can get away with it.

It's extraordinary to claim there will be an uptick in fraud if we do large scale mail in voting in the US?

Even the above linked claim in snopes says fraud is more common with mail in ballots.