An all-too-common ridiculous assessment of election outcome

From this post by Scott Gallloway:
It's likely [at November election time] we'll have a GOP nominee who's been convicted by a jury of his peers for crimes against the United States. The evidence for that last prediction will be submitted at three separate trials, and the odds of beating all three are terrible: Defendants accused of federal felonies have only a 30% chance of avoiding prison time in just one trial -- beating the rap in three trials comes in at a 2.7% probability (30% * 30% * 30%).
Of course, the Trump cases are hardly typical of federal felonies, and the three events are far from independent.