A show is worth attending if the people who can buy from you will be in the building and you can get in front of enough of them. Everything else, including the size of the event and how good the stand looks, is secondary to that and frequently misleading.
Read the exhibitor list before the attendee list
The exhibitor list is published early and is a reliable proxy. If your competitors and the companies your buyers already work with are exhibiting, the buyers follow. If the list is full of adjacent industries, the audience is adjacent too.
It also tells you what the show costs to compete in. A hall of large custom builds means a modest stand will be invisible, which may be a reason to attend without exhibiting at all.
Ask the organiser the uncomfortable question
Ask for the seniority breakdown and the job title split, not the headline attendance figure. Twelve thousand attendees of whom most are vendors selling to each other is a different event from three thousand of whom a third hold a budget.
An organiser who will not answer that has told you something useful. The ones running a strong event are usually pleased to be asked.
The booking test
Six weeks out, try to book ten meetings with named companies on the attendee list. If you can get to ten, the audience is right and the show will work. If you cannot get to three, the stand will not rescue it.
This is also the cheapest possible pilot. Attending without exhibiting, with a diary of booked meetings, costs a fraction of a stand and answers the question for next year properly.
One show properly beats three thinly
Most teams exhibiting at several shows a year are under-resourcing all of them. The week before, the week itself and the fortnight after each need real attention, and three shows in a quarter means none of them get it.
Pick the one where the booking test succeeded, spend the budget of the other two on doing it properly, and judge it at ninety days.