The best 20% — George Lockett

The Pareto Principle, or the 80:20 rule, states that (approximately) 80% of the outputs come from 20% of the inputs.
I don't think it technically qualifies as a decision razor, but it's certainly a useful organising/prioritisting/clarifying heuristic to apply in lots of contexts.
I've used this in the past when thinking about clients. I read somewhere that applied this as: 80% of your good-value returns probably come from 20% of your clients; 80% of your headaches and problems probably come from 20% of your clients. They may overlap, but they probably don't, sigificantly. So, focus on that 'good' 20% and try to find more clients like them; try to fire, eliminate, or otherwise mitigate the 'bad' 20%.
In practice in the sort of work I do, there's not the same commoditisable approach, volume, or breadth to clients BUT it was still illumunating to undertake the exercise of figuring out where those 80% of desirable outputs came from, and what linked those clients together, and turn them into more generalised 'rules' for evaluating new contracts.
Other end of the value spectrum, but: I recently did the same thing to the email newsletters I subscribed to. I added a bunch, wasn't reading them as much, etc. etc.
Generally: I had an abundance of them and too little time to actually read.
So I worked through which ~20% were most valuable/interesting to me, and unsubscribed from basically anything else.
Some others stuck around that wouldn't have made that cut, but they usually had some specific, seperate value, or otherwise were so infrequent or irregular to not be worth counting.
I now have fewer things to read and enjoy them more.
I've now done the same thing with my backlog of side projects.
I can actually quantify roughly how much time I'll have to invest in side projects for the rest of the year and hold that up against my rough estimates for all the things I want to pursue that are non-day-job work.
The Pareto Principle provided a useful and clear lens to think about what actually mattered most to me in terms of what is exciting to work on and which best supports my goals.
There's a trap in thinking you can do it all (you can't! there will always be more new things generated vs your time to actually invest in doing them).
(My reading list is a good model for this. I eliminate things from there with abandon when it gets too full. Because it's always going to fill up faster than I read things. I either have to reconcile myself to a long, unusuable reading list, or find another way things get cleared off the list without guilt.)
But there's a further trap in thinking you have to fill all the time you have. Cutting everything but that ~20% technically leaves me with a time surplus for the year. BUT obviously the answer isn't to back-fill it with other backlogged projects. Beyond the usual 'things will take longer than I have estimated' angle, it's better instead to leave room for more projects that are like that 20% in terms of what they offer.
Just because something matches your desires and goals a bit, that doesn't mean it's inherently worth doing. There's an opportunity cost to pursuing it at the expense of something better or more diverse.
The main catch I can see to this approach (which certainly doesn't apply to everything) is the risk of things starting to look the same. If I'm eliminating everything outside of that initial 20% (be that clients, projects, or newsletters) and using that to inform my selection criteria for new things, it risks losing the sense of serendipity and novelty from something that doesn't 'look right'.
(c.f. Taleb's 'Look the Part' test. When choosing between two things of seemingly equal appropriateness, choose the one that least looks 'the part'.)
But in practice I'm not overly worried by this. Life isn't perfect, not everything has to be optimised, and my standards for excluding things that don't match just aren't that good to exclude all possibility of new things. They're going to leak in, and that's good. The leaks can let the interesting in. (And the annoying. Almost certainly they'll let the annoying in, too.)