The problem is analysis paralysis usually strikes because you have to make a decision today which have profound consequences a long way down the road.
For example, if you're deciding between Ruby and Go, it's a choice between things like a more extensive gem library vs not having to rewrite the code to achieve scale. Those are things you can't test in an experiment, but they're choices your future self will have to live with.
That's why I believe it's good to just have a rigid flowchart when making technical decisions of this nature. The time to make or modify the flow chart is not the same time as when you need to make a decision. Later, once the product is off the ground, play around the edges with new technologies and use those learnings to modify the flowchart. This, to me, was the best way to avoid analysis paralysis.
For example, if you're deciding between Ruby and Go, it's a choice between things like a more extensive gem library vs not having to rewrite the code to achieve scale. Those are things you can't test in an experiment, but they're choices your future self will have to live with.
That's why I believe it's good to just have a rigid flowchart when making technical decisions of this nature. The time to make or modify the flow chart is not the same time as when you need to make a decision. Later, once the product is off the ground, play around the edges with new technologies and use those learnings to modify the flowchart. This, to me, was the best way to avoid analysis paralysis.