The progress in particle physics crucially depends on a few large experiments, releasing new data to the entire community of theorists several times a year.
Such a format of theory-experiment interaction leads to what is called the "ambulance chasing" phenomenon. Namely, every time after an announcement of a preliminary experimental result (i.e. one with less than 6 sigma deviation), the arxiv.org preprint server explodes from the amount of theory submissions explaining it. Widely-known examples would be the detection of superluminal neutrinos five years ago, or the recent ATLAS observation of a two-photon peak at 750 GeV.
Quite often, it turns out that the result happened to be within an experimental error bar and is not confirmed by future measurements. However, whether right or wrong, the theory papers receive a fair amount of citations - the more the earlier the preprint appeared.
The race for priority is so intense that dozens of theory preprints are submitted within hours (!) after the official announcement, which means that they were written in advance based on some insider information from the experimentalists.
In any case, in a recent preprint Mihailo Backović (Catholic University of Louvain, Belgium) developed a theory describing the ambulance chasing phenomenon both qualitatively and quantitatively.
The analysis is based on the assumptions that the number of papers on a particular topic can be described using Poisson statistics, and that the interest in the topic as well as the number of available ideas decrease in time (he considers power-law and exponential decays). This resulted in a two-parameter model, which provided a perfect fit to 9 cases of ambulance chasing, considered by the author.
Thus, even if all these theory papers were wrong in explaining the measurements, they at least can be used to study universality in complex systems. :-)