[Jensson]

> While they have utility as tools to make running trials on interventions easier or more straightforward, they’re too often used to run different trials altogether than what’s promised in the abstract.

Yes, this makes it a lot easier to find something to publish. If they stop doing this they hurt their career prospects, it might be good for science but it isn't good for the individual.

In order to improve science you need to see it from a capitalist perspective where the researchers are desperately fighting over resources. As long as that fight is won by performing bad science we will mostly get more bad science. The most critical part to change is peer review, today we treat peer review as the main cornerstone of the scientific method, even though it is very unrelated, it is only to help filter out the worst of spam, it doesn't say much at all about the validity of the paper.