Now that is an interesting theory, Alex! I have noticed that they seem to like simplicity. A very simple, linear structure with just one idea, moving forward square by square like a chess pawn. No knight's leaps or pirouettes. I had put that down to wanting to maximise reads - and hence user satisfaction - by not overburdening comprehension skills, going for that 8th grade reading level to keep people on board.

But 'staff cuts' + 'focus on easily and objectively identifiable criteria' + 'preference for low perplexity style and structure' could also add up to something else - you're damned right.

It would also be intriguing to know what info a Curator sees on their dashboard when they open a piece for review.

Do they have prior stats on that writer, pub and lead topic tag, marking them up or down based on past read ratio stats?

Medium's general lack of technical sophistication would suggest that would be beyond them. But it's certainly something you would expect to be factored in.

"We just boosted three of this writer's/nommer's pieces, and they all bombed with our readers [maybe because the algo sent them to the wrong people??] That's $135 down the drain. Spike the next few pieces."

We need a mole!