Using the tools of science to evaluate the past is fundamentally different from using them to evaluate the future. If a corpse is found in a room with a pistol whose odor is similar to that of firearms that have been fired recently, and the corpse contains a bullet with a funny little nick that happens to coincide with a defect in the pistol's barrel, and there is a shell casing on the floor which is of the same manufacture as the bullet and markings in some grease on the breach face mirror those on the casing, it's pretty likely that that firearm was recently used to fire the bullet that ended up in the corpse, and that the corpse was shot at that location. On the other hand, it would be possible that the corpse was shot somewhere else with that firearm, and that someone moved both the corpse and the shell casing.
It's even possible (though unlikely) that another firearm was used, and the barrel/slide transplanted onto the firearm the police found, or that someone deliberately produced identical nicks on two different barrels, or any number of other possibilities.
If everything fit the simplest theory (person shot where gun was found), there would probably be no particular reason not to believe it. But if something was discovered that didn't fit, there'd be no reason to believe with certainty the next alternate theory that was formulated.
An honest appraisal of the historical sciences would recognize that very little is really 100% certain. Archaeologists need to consider that some of the writings that have been used as the basis for our knowledge of ancient civilizations may have been in fact entertaining fictions at the time they were written. Cosmologists need to consider that objects may have passed through what we now consider the "known universe" at various times, disrupting its state but leaving no direct evidence of their existence.
The claim that objects in a vacuum within a uniform gravity field will accellerate uniformly, within a certain degree of precision, if not operated upon by other forces is a fact. You can perform thousands of experiments on the subject and get the same result. The claim that Egypt was once ruled by Phaoroahs is a very strong conjecture--either someone ruled over the land and had enough power and influence to get things like the pyramids built, or else some totally bizarre phenomenon unknown to western civilization caused them to come unto existence.
The notion that evolution could cause a horse and a donkey to be descended from a common ancestor via no mechanism other than random mutation and selective breeding would be a plausible hypothesis. The notion that a mosquito and an orangutan could have come from a common ancestor via no mechanism other than random mutation and selective breeding, however, seems a bit less plausible. Even if genetics were to suggest that two species are likely to share a common ancestor, it would not imply that there was not some, as yet totally understood, phenomenon involved in addition to the processes comprising "natural selection". Indeed, it would seem that there have to be some such processes involved. If one wants to say it's divine intervention, or aliens from the planet Altair IV, or cosmic pixie dust, or whatever, it's all the same. Once a theory reaches a point where "well, we don't really know how things got from here to here", all bets are off.