In the past, reaction theory was used to describe phenomena. In the last few years new promising developments may transform this field and make it fully predictive.
Nature has given us QCD (Quantum Chromodynamics) and whatever nuclear property we may be interested in, it is a result of the complex action of the fundamental strong force. Because it is truly impossible to calculate the structure of nuclei using QCD, theorists have found a nice way (using effective field theories) to make the connection between QCD and the force that binds nucleons (neutrons and protons) into nuclei. This "ab-initio" trend has proliferated in nuclear theory since the early 2000s. And, it is now catching on in reaction theory.
Just as in other physics problems, in nuclear reactions, there is fine tuning of particle thresholds, adding to the pulls and tugs that can strongly affect the dynamics.
For most reactions of interest, in addition to the effective theory that connects QCD to the nucleon-nucleon force, theory needs to build on another layer of approximations, one that takes the nucleons many-body problem, into a few-body problem. This introduce another type of effective forces, between nuclei. They are called optical potentials and typically carry uncontrolled uncertainties.
A new wave of "ab-initio" optical potentials
Several groups are now invested in developing method to extract optical potentials between a nucleon and target nucleus from first principles. Some are focused on generating these potentials at low energies (see Rotureau et al., and Idini et al.) while others focus on an expansion appropriate for high energies (see Burrows et al.). These methods to extract the ab-initio optical potential are only as good as the input (which consists of information coming from many-nucleon calculations). At present there are serious limitations in these many-nucleon calculations. But as those get resolved, there is a promising path ahead to transform the field of reaction theory into a predictive one.