Commenting on Bart Ehrman's Misquoting Jesus, Stanley Porter says (among other things):
Ehrman notes, as I already mentioned above, that John Mill [c. 1645-1707] examined around one hundred manuscripts and found thirty thousand variants. Now, Ehrman notes, with over 5,700 manuscripts, we have somewhere between two hundred thousand and four hundred thousand variants. Or, as Ehrman likes to say, we have more variants than words in the New Testament. This sounds rather shocking; in fact, it is sensationally so. Mill had on average three hundred variants per manuscript. Is that a lot? Given that some of these are minor variants, others changes in word order, and others obviously slips of the pen, I think not. However, with roughly 5,800 manuscripts and, for the sake of argument, four hundred thousand variants (the largest number selected), this means only seventy variants per manuscript. With 5,800 manuscripts and two hundred thousand variants, that reflects only thirty-five variants per manuscript. So in fact, the situation with variants is getting better with the discovery of new manuscripts, not worse. Eherman should be applauding rather than disclaiming. Another way to look at these statistics is to recognize that, on a conservative estimate, 80 percent of the text is established (some say 90 percent or more), regardless of the textual variants present in the manuscripts. If textual variants are distributed equally throughout manuscripts–they may or may not be, but there is no other way to examine this, and some of them, such as spelling, transpositions, and accidental scribal errors, almost certainly will be–this means that, if there are four hundred thousand total variants, there are only eighty thousand in the part of the New Testament that is not established, or an average of only fourteen variants per manuscript in the disputed portion; or if there are only two hundred thousand total variants, only seven variants per manuscript in the disputed portion. This is manuscript production–remember, the copying of ancient manuscripts was done by hand–that nearly rivals that sometimes found today in modern print! Ehrman's comments, then, are a clear instance of unwarranted sensationalism. Of course, the way to treat variants is not simply to average them, but there is no need to sensationalize and exaggerate the situation so as to engage in fearmongering. After all, besides those mentioned above, many if not most of these variants will be unique variants, probably (on the basis of the distribution of dates of manuscripts, in which the vast majority are late) in later manuscripts, with little impact on the text; others will simply be the repeating of similar types of errors, again with little impact. This no doubt accounts for why in his treatment of the subject Ehrman returns to the same relatively limited number of examples of textual variants.
In Misquoting Jesus Ehrman's sensationalism, besides a few incidental examples, begins with the story of the woman caught in adultery (Jn 7:53-8:11) and the ending of Mark's Gospel (Mk 16:9-20)…Ehrman is misleading on at least two fronts. First, he makes it seem as if many, if not most, of the textual variants are ten to fourteen verses in length, as these two passages are, when he knows better. In fact, most of the others that he discusses in the book are a word or a phrase in length. This latter length is far more representative. Second, Ehrman gives the possible impression that the scribes, in changing the text, deleted two valuable early passages, when quite the contrary is true. Later scribes, for whatever motives, added later material, but material that on the best textual grounds was never originally there in the first place.
An examination of several of these other shorter examples, however, shows that Ehrman is on thin ice to claim that there is radical and gratuitous change of the text of the New Testament. I will not treat accidental errors, because to know that it is an accident assumes that we know what it is not to have the accident. In other words, where such occurs, the original is easily discernible. I will also not treat intentional changes where it is clear why the change was made–for historical, theological, or factual reasons–but where the original or unchanged text is easily restored, or where the Byzantine tradition is the only one that supports it. S. Porter, How We Got the New Testament: Text, Transmission, Translation (Baker, 2013), 65-69.