How Do You Evaluate Theories and Physicists in Theoretical Physics?

[Silviu]

Hello! I have one more year in undergraduate and I would like to go for a grad school in USA and do theoretical physics. I am a bit confused about particle physics. From what I saw most of the theoretical models today aim to extend the standard model and yet most of them are proved to be wrong at LHC (at least at the available energy). My main question is how are certain physicists considered to be better than others when most of the theories that are not disproved, can't be proved either (in string theory for example). So as long as you can't test 2 theories, how can you say one is better than the other? So what it takes to be a good theoretical physicist, when your theories can't be tested? (and I include here most of the main topic nowadays: supersymmetry, string theory, dark matter, dark energy etc.). Thank you!

 

[mfb]

I'll only consider particle physics here - keep in mind that theoretical particle physics is only a small part of theoretical physics.
The question "which physicist is better" and "which theory is better" are completely independent things.

What is a good theory?
The SM, obviously. You cannot invent the SM any more - but that does not mean everything there would be done. To search for deviations from the SM (experimentally), we have to know what the SM predicts in terms of observable quantities - cross sections, branching fractions, particle masses, angular distributions and so on. These SM predictions are a lot of work. The SM has some free parameters that have to be determined experimentally - someone has to interpret the experimental results to improve our knowledge of these free parameters.
There are also things where purely theoretical descriptions would be possible in principle, but we cannot do them today. Parton density functions are an example. New approaches to calculate them from first principles with better precision would be amazing. For now, experimental inputs are more precise - but again, theorists are needed to convert the experimental results to improved PDFs.

BSM theories are rarely invented in a vacuum - they are made to solve problems of existing theories. A good theory solves as many problems as possible, with a minimal number of assumptions or new parameters, while making as many clear and testable predictions as possible.
The various flavors of supersymmetry, technicolor, extra dimensions, various GUT ideas, ... are all approaches in this category. While some people came up with the first ideas of the approaches decades ago, most work goes into understanding these theories. How can we possibly test a theory with more than 100 degrees of freedom? How can we constrain these parameters to make testable models? What does that predict for LHC measurements? Given recent LHC measurements, which region of the parameter space is excluded? Which measurements will give the most stringent limits on parameters?

Approaches to unify QFT and gravity (string theory, loop quantum gravity, ...) typically do not directly lead to testable predictions. They are more exploring the landscape, developing mathematical tools and working on proofs that the resulting theories are internally consistent. Some of the mathematical tools developed for these theoriey found applications in SM predictions.

What is a good theoretical physicist?
Someone working on one of these topics, producing accurate predictions or accurate mathematics, developing new approaches to tackle open problems in these fields, or even finding completely new problems, interesting for others.