The EPA doesn’t test the EPA range, the manufacturer does. So it’s entirely possible to cheat, and it’s likely you’ve heard of some famous cases of cheating. Maybe a case involving VW and emissions?

Tesla’s EPA estimate is the most out of line with real world tests whenever any independent organization tries to recreate the claimed range, and that shouldn’t be a surprise to anyone given Teslas long history of bending the truth. They also have the most egregious “correction factor” for their range test that I’ve seen any manufacturer adopt.