I have absolutely no idea how you’re supposed to write good poems.  

But if I were to guess, you should start out by practicing.  Practice writing bad, and, if you can manage, mediocre poems.  Then when you are ready, bring out the Scientific Method.  

“Whaaaat?”  (I can hear you.)  

Trust me.  

1) Ask a Question:  This seems easy enough, but questions are pretty hard to come by.  The easiest way to do it is to pretend you are about 3, and put “Why?” at the end of any explanation anyone gives you until you’ve found something interesting and answerless.  

2) Do Background Research:  Read lots of poems.  About the subject, not about the subject.  Read and read and read and ponder your question.  And read some more.  

3) Construct Hypothesis:  Do not do this on purpose.  Prediction is against the rules.  You do not squish tea bags to make your tea brew faster.  That’s gross, and a surefire way to acquire burnt fingers.  This is where “finding empirical evidence for natural phenomena” and “writing poems” differ.  Your current idea of the universe IS your hypothesis.  All of the thoughts in your head ARE the hypothesis.  Do not write another.  

4) Test with an Experiment:  Here is where scientists write down observations.  I would just take them in.  Breathe deeply.  Open your ears— see what people are saying.  Open your eyes— watch what the world does when someone gives it respect and attention.  Smell things.  Touch things.  Make yourself a good peanut butter sandwich and ask yourself how it actually tastes.  

5) Wait.  

6) No really, wait.  

7) Hang on.  

8) Oh no, you just couldn’t hold the thought in, could you?  Man, now you have to…

Just kidding.  Nice poem.