Speaking as a professor, I have frequently been a judge (one of five for my field) for our annual poster contest (all sciences) in which students produce a poster describing their research, suitable for an academic conference (often actually accepted for display in such a conference). These typically take students about a month to produce, they are not done lightly, typically contain illustrations, charts and text with references.
Drawing from that experience, I'd say the winning criterion is surprise.
Most research produces nothing very surprising. It isn't that I knew their results in advance, but their results are in line with what I would expect myself if I undertook their research.
The students may make a valid discovery or provide statistical evidence something is true, something I would guess is true but did not know for certain was true -- but the key element for winners always seems to be a surprising result that, once you learn it, makes sense. The results stick with you, get you thinking, make you wonder about something else or further implications, or get you talking with the presenter (the student) about where this goes next, what are the next experiments, where do they think it is going.
Translating that experience to a fiction contest; I think a surprise that in retrospect makes sense could go a long way toward winning. A twist that resolves the conflict in an unexpected manner, whether it is humorous or not.
I realize that is often how a joke is described; the punchline is an unexpected statement or action that, in retrospect, works.
But in fiction the twist can be like The Sixth Sense, it isn't really funny in any way, but it is fun because it suddenly recasts the story in an entirely new light, and for some reason that makes us humans laugh, we really like it.
It is a prescription that might be difficult to fill, twist endings and surprises are not that easy to devise, and may fall flat. But hopefully a little more specific advice than "write better."