Matt Waymeyer @ http://expositorythoughts.wordpress.com reviews the reliability of Scripture in regards to manuscripts. He notes:
The Abundance of Existing Manuscripts
"By way of comparison, only ten manuscripts of Caesar’s Gallic Wars exist, the earliest dating 900 years after Caesar; only eight manuscripts of Thucydides’ History of the Peloponnesian War exist, the earliest dating 1,300 years after Thucydides; only eight manuscripts of Herodotus’ History exist, the earliest dating 1,300 years after Herodotus; and only two manuscripts of Tacitus’ Histories and Annals exist, the earlier one dating 700 years after Tacitus. As W. Edward Glenny notes, “the number and early date of the NT manuscripts give us great confidence that God’s Word has been preserved in these documents.”
The Insignificance of Most Variants
Daniel Wallace notes:
In that two percent, support always exists for what the original said—never is one left with mere conjecture. In other words it is not that 90 percent of the original text exists in the extant Greek manuscripts—rather, 110 percent exists. Textual criticism is not involved in reinventing the original; it is involved in discarding the spurious, in burning the dross to get to the gold.
The Preservation of Primary Doctrines
D.A. Carson draws a helpful analogy:
In my judgment the degree of uncertainty raised by textual questions is a great deal less than the degree of uncertainty raised by hermeneutical questions. In other words, even when the text is certain there is often an honest difference of opinion among interpreters as to the precise meaning of the passage. Few evangelicals, I would like to think, will claim infallibility for their interpretations of the Scriptures; they are prepared to live with the (relatively) small degree of uncertainty raised by such limitations. The doubt raised by textual uncertainties, I submit, is far, far smaller.