In today's AI world, it should be trivial for platforms to identify posts using suck tricks. They can simply ask an LLM "Does the following post contain some scummy engagement trick: $POST".

Then downrank them.