Thank God I have no one doing the Leaving Certificate this year. I don't envy anyone trying to devise a 'least-harm' solution, but I've got an idea worth considering that addresses a fundamental weakness of the proposed 'predicted grades' system.
In order to solve a problem, you have to define it correctly.
Let's go back to the Leaving Cert itself and understand what it attempts to achieve. It has been determinedly moulded over the years to perform one function only - to allocate places in higher education.
But the whole point of the Leaving Cert is not what it does but how it does it. Its 'how' is the reason we cling to this barbarous rite of passage.
So why do we have mass hysteria about the Leaving Cert every year? In the dark ages, when I did the Leaving Cert, everything didn't hang on it.
Students who wanted to go to one of the National University of Ireland colleges, like UCD or UCC, could sit its matriculation exam. Trinity did its own 'matric'.
The technical colleges conducted interviews for many courses and I know several people with poor Leaving Certs who got into courses on good interviews and went on to achieve great success.
But the Department of Education pushed for a single method to act as the gatekeeper to further education.
Gradually, all other entrance exams were wound down, interviews discouraged, and now the Leaving Cert is the educational equivalent of St Peter at the pearly gates. Fail it and you fail everything.
Almost everyone recognises the brutality of a system in which one bad day can precipitate disaster; when education is reduced to regurgitation; when class is so depressingly influential in the final outcome; when teachers are hamstrung by a rigid curriculum; and when it fails to measure so many important qualities in students.
But despite all this, in this small country, when favours and influence are so easily arranged, we recognise that whatever other inequities are present, at least on this one day, it is entirely anonymous.
Every student becomes a number. Daddy can't help you now.
It is limited in all other ways but this: for one day, it's fair. Brutal, but fair.
And this is why people are so concerned about predicted grades - it undoes that one fundamental advantage.
Slagging off teachers is a cheap shot, so I'm not going to accuse them of deliberately manipulating grades for or against particular pupils.
But with the best will in the world, they can get it wrong sometimes.
Everyone has a story about the teacher who dragged them down but who they proved wrong in the exam. Or all the students who did a terrible mock exam but pulled it out of the fire by the time the Leaving Cert came around.
My idea removes the power of teachers to influence who gets what course in college - random selection.
So, let's say it is Law in UCD and it has 500 places (that's a guess). Take everyone who put that first choice on their CAO form and do a raffle. For those who lose out, move on to the applicant's second choices and so on. Cracked?
Well, we already do this and always have done to some extent. Random selection is used when a bunch of applicants have the same points, but there is a limited number of places left on a course. If there are 100 students on the same points, but only 50 places, it's a lucky dip.
This was relatively common on the old A, B and C grades, when lots of people had similar points. It's less common since they narrowed the bands, as a H2 or H3 is separated by only five points. But it's still used.
So the principle is accepted. Funnily enough, in my time you knew when you got in on random selection and students would furtively admit to their good fortune. These days they don't tell applicants when they got in on the raffle, so people aren't as aware of its existence.
I floated the idea around some academics who were supportive of the idea.
One criticised the plan because he worried that students' top choices are often far beyond their capabilities and he'd have students who just weren't able for the course sitting in his classroom.
But I checked with the CAO and around 80pc of students applying for a Level 8 course, a degree, are offered one of their top three choices.
Of all applicants for a Level 6 or 7 course, 89pc get their first choice, and 98pc are offered one of their top-three choices.
The alignment between application and what's offered is extraordinary. It shows that students are actually fantastic at predicting their own grades.
My system only works by using choices people have already made. You couldn't do it every year because students might apply for fancy courses they can't manage.
Now, the trick to any solution is that students buy into it.
I expect predicted grades would lead to a huge wave of appeals, and increased applications to courses next year that will impact the class of 2021.
But if you give students what they want, the course they applied for anyway, without the influence of teachers and bell curves, they might be convinced.
It's certainly worth asking them.