Because 100 copies of an error do not make it less an error. Consider how simple this principle is in our daily lives. We have rejected the vast majority of world religions in history for one. Why can we lean towards the one when it goes against the majority? Becuase we adhere to truth. I realize the analogy is not flawless but it shows the principle. We cannot be interested simply in "vote counting." We are more interested in accuracy. There are a number of text critical canons that are used to make decisions like this. Some are these:
1. The older is more likely to be accurate because it is closer to the originals without as much possibility for corruption in copying over centuries.
2. The shorter is more likely to be accurate because things tend to get added more than deleted.
3. The harder is more likely to be accurate because of scribal tendencies to simplify things. This canon is true "provided it is not too difficult."
Yet one of these is not sufficient. All must be balanced against each other. In sum, a text critic is looking for the reading that best explains the other ones. They look for reasons why something would get the way it was. Say for example we have manuscripts A, B, C. Assuming A is original, how would B and C develop. Assuming B is original, how would A and C develop. Assuming C was original, how would A and B develop. Metzger's textual commentary is a valuable source that shows how this principle is put into action. Perhaps a few citations regarding your verses would demonstrate:
On Mark 9:42, "The presence of eis eme is very strongly attested (actual manuscripts listed but here omitted since I don't have the characters). At the same time, however, the absence of the words fo aleph, D, and Delta, as well as the possibility that they may have come into the Markan text from the matthean parallel (18:6), casts substantial doubt upon their right to a firm place in the second Gospel. The Committee therefore decided to enclse the phrase within square brackets" (p. 86).
On JOhn 6:47, "The addition of eis eme as the object of the verb 'believe' was both naturl and inevitable; the surprising thing is that relatively many copyists resisted the temptation. If the words had been present in the original text, no good reason can be suggested to account for their omission. The reading of the OLd Syriac has been assimilated to the text at 14.1" (p. 183). PL's comments: Here we see the work of the second principle above (shorter) plus principle of "accounting" (as I call it): What best accoutns for the others? The shorter: If "in me" had been original, it is so common that surely no one would have omitted it. Therefore, its omission in some texts is very unreasonable. Had the omission of "in me" been original, it is such a common phrase that it would have been very easy to add it in unintentionally. Therefore, the shorter reading accounts for the longer. The longer reading (in me) cannot well account for how it would have been left out.
Additionally, in many places the majority text disagrees with itself. The actual texts of hte majority text are not unanimous. They are divided. In those, you must enter into some process such as these canons to make determinations.
I could go on but this should give some evidence as to how these differences are treated in the eclectic texts today. MY overall point is, It is not as easy as counting the stacks of manuscripts.
And how do you know what that reader would believe,: Christ is speaking in that verse, and would not the Holy Spirit help the reader understand that the verse is talking about believing in Christ and not believing in Santa Claus? Your objection is *really* weak. What if the reader instead read Mark 9:23, Rom 1:16, Rom 10:4, or 1 Cor 7:12 from the KJV, none of which explicitly specify what or who to "believe"?
