Sorry, your comment does not reflect serious knowledge of the field.
Some parts are subject to distortion from having small pieces. The connection of the face to the cranial vault is one such area.
But the braincase is a different story--the various pieces have to match their curvatures in three dimensions and to join into a continuous whole. There is much less chance for a wildly erroneous result because of this.
And don't overlook the fact that paleontologists are generally dealing with multiple specimens. Not all are as complete as the ones you see in the popular literature, but that redundancy lets paleontologists study a number of different specimens, with different degrees of intactness, as a guide in assembling fragmentary specimens.
Its not just a matter of guesswork; there is a lot of hard work and solid research involved.