Someone said that history, through Marc Bloch's contribute, became a sort of "sociology sensible to diacronic changes".

There is no history without facts, and no facts without sources attesting said facts' existence. Some historians believe that if no written records of a facts are found, said fact simply does not exist.

If a study of several sources of various type (be them archive, literary, iconographical, archaeological etc.) points towards a reconstruction of an event, based on these informations, rarely there is an absolute, mathemical certainty that said event played out in the way it is reconstructed to have happened. We can recreate a plausible conclusion, an educated guess, a reconstructive hypothesis of the diacronic perspective that event places itself in.

As an example, let's suppose we find documents regarding the construction of a farmstead by a Frankish count. We also found parts of said church within an archaeological dig and a mention of said building in a notary act regarding a litigation some 200 years later.
We can reconstruct when it was founded, who was its beneficiary, who was appointed priest there, where it was located. We may never know, unless explicitly stated, why it was founded, why the first recipient was chosen, what happened between its buildig in 788 and the litigation it features in in 988. We must fill that "contingency gap" by studiying its context further, taking into account all the plausible circumstances (what Braudel would have called "long duration phenomena"), and conclude that, most likely, count Hincmar of Velbret founded the church of Quatruit in the year 788 and granted it to Roderic of Soisson.
Who was Roderic of Soisson? What was his relationship with the count? Was he a friend? Likely. His illegitimate son? Possibly. Was it a returned favour of sorts? Maybe.

We will always have an incomplete picture (or rather, we will never fully know exactley how an event occurred), but we can elaborate on it by looking at the general context where it took place in.

If we had to study, a thousand years from now, the attack of the Capitol Hill, we will never fully know if the attack was staged, improvised, coordinated. We might see that it didn't exactly fit a tendency of the period, so was it a peculiar instance? Was it a moment of change within the of a country's (or ideology's) perception of political theory and political life? Or was it a coup, a symptom of civil unrest to be placed in a greater picture of instability and stress worldwide and not only limited to one country? And if so, why that country in detail?

Ultimately, history is the study of human interactions and human "things". Proteiform and very varied as they are, I strongly disagree that maths can be applied to see a pattern or prevent them.