Okay so say your testing a brand new rocket engine idea. It uses a fuel nobody has tried to use before. So what you do is you figure out how much energy this fuel has and do some math to figure out how much you'll need to take with you for the typical rocket. You design an engine for this spec or better and thoroughly test it to make sure it's behaving like expected. You eventually mount it to a rocket and make sure in practice it behaves as you expect. Next you put a payload in the rocket and test it again. If at any point things don't behave as expected you have to fix your whole model.

SpaceX struggles to go a launch without their engines destroying themselves. Perhaps they should go back a few steps?

A rocket is not fundamentally new and hasn’t been for almost 100 years.

Rockets perform correctly when they deliver their payload to the correct orbit.

You can calculate the energy density of fuels, the efficiency of your engines at various atmospheric pressures, and determine the payload size you can deliver with your engines and fuel. Blowing up rockets for “tests” is so 1950s. We have whole college programs on rocket design. We have desktop computers more powerful than anything available in the 1960s, and NASA managed to design the Saturn V, a rocket of similar size to starship, with the computers of the time and fucking slide rules. The Saturn V had its problem, but each rocket managed to deliver its payload and perform its part of the mission without blowing up.

Your comment is classic tech bro. No understanding of real engineering principles and only a desire to shove some shit out of the door as fast as possible.