I didn't mean to bring focus on the randomness of jury selection. (In fact it's obvious that in this particular case, it was very biased, which was a big part of the problem.)

Rather, "juror" should be someone's job.

People pulled off the street are not going to have the critical thinking skills, and will be broadsided by pressure tactics and manipulation.

Also, some people who are inconvenienced by jury duty will be resentful and just focus on their personal exit strategy (get it over with ASAP).

Professional jurors wouldn't be an improvement. They will get cozy with, and biased towards, the people they see and interact with on a daily basis. It completely undermines the concept of being judged by a jury of your peers if there's a permanent juror class.

I went to school at the University of Virginia, which offered something like this for honor trials. (UVA Honor violations were specifically: lying, cheating, or stealing.)

For a trial, you could choose between a random student jury, a jury of Honor Society members, or a mix.

Statistics showed that non-random juries were way more likely to convict students, and I have no doubt permanent jurors would do the same over time.

I understand your criticisms for sure, but the alternatives are worse.

Your argumentation isn't very strong, because you're presenting random jury selection from the population as being the only viable alternative, due to issues in other alternatives, without considering ways in which those could be mitigated or fixed.

Elsewhere in the thread, there are comments about how some countries have eliminated random juries; so a good start would be to study how things work elsewhere.

> non-random juries were way more likely to convict students

All that matters is whether they were more or less likely to correctly convict or acquit.

If it so happens that all the accused liars, thieves or cheaters are actually guilty, then the higher the conviction rate, the more accurate it is.

Honor Society members are not professional jurors; they are just a set of people from which a jury is less randomized than from a broader set.

> All that matters is whether they were more or less likely to correctly convict or acquit.

Which you have no way of knowing, because if we had a process to determine that, we wouldn't need juries to begin with.

What you can tell, by comparing the outcomes between decisionmakers, is which way their bias leans relative to the alternative. And "biased towards convicting more innocent people" is bad.

> And "biased towards convicting more innocent people" is bad.

How can you write that following a sentence which admits we "have no way of knowing".

We can't even say that more convictions is no fewer innocent convictions, let alone more.

Say that 1% of the accused in some jurisdiction are innocent. In that situation, the ideal conviction rate is 99%. It's possible to have a 2% conviction rate where the convictions are wrong half the time (and the acquittals are all wrong), and a 1% conviction rate where the decisions are all backwards in relation to guilt.

You're assuming the full-time aristocrat will do a better job, which is the thing there is no evidence of. It could be completely the opposite -- the aristocrat makes their decisions based on political connectedness, finding everyone guilty except their cronies who are actually guilty.

In the absence of any evidence to the contrary, the assumption should not be that either of them is inherently better at the job, and then a higher total conviction rate implies a higher false conviction rate.