Betty Crocker and other corporate mascots don’t exist despite having biographies, named family members, etc. Since most corporate mascots don’t exist, we should assume any corporate mascot we hear about doesn’t exist unless proven otherwise. Colonel Sanders is a corporate mascot who did actually exist, but we know this based on evidence. We’d be wrong to simply assume he existed when so few corporate mascots are historical people.
Since most ancient demigods didn’t exist, we can’t assume Jesus did. He might have, but this needs to be demonstrated, not assumed. Since Jesus fits the Rank-Raglan reference class better than any other classification, the prior probability of him existing is the same as that for figures such as Moses, Hercules, and Romulus: about 6%. In order to be generous to historicists however, Carrier gives a higher prior probability of 33%.
If you’re afraid of math like many biblical scholars, you may be tempted to give up at this point, but keep in mind that all historical arguments involve probabilities. Carrier is simply quantifying them rather than using terms like “likely”, “probably”, etc. If the math scares you, just mentally replace 6% with “highly unlikely” and 33% with “somewhat unlikely” or something like that to help get your mind around it.
For something to count as relevant primary evidence, it must be causally connected with the facts, and it must be independent of other primary evidence. There is no archaeological evidence for Jesus, so all we have to work with are documents. (Whether the city of Nazareth existed or not at the time of Jesus has no relevance one way or the other.)
Carrier limits himself to texts originating before 120 AD since anything written later couldn’t have come from someone alive during the founding of Christianity. Since it’s unknown exactly when the texts of the New Testament were written, Carrier mostly goes with the problematic dating of the scholarly consensus. (He estimated that it would take something like seven years to properly date the documents. This is work that biblical scholars need to do, but surprisingly, it hasn’t been done yet.)
The consensus among biblical scholars is that only seven of the Pauline epistles are genuine with the rest being forgeries. The seven authentic epistles are 1 and 2 Corinthians, Galatians, Romans, Philippians, 1 Thessalonians, and Philemon. (Philemon actually may be a forgery as well, but it contains no relevant data anyway, so it doesn’t matter.) These were probably written sometime in the 50s, however, these letters do have textual variants, interpolations, and rearranged passages, so we can’t blindly trust everything they say.
The other letters attributed to Paul are too stylistically different to have been written by the same person. Hebrews was not written by Paul, but it was likely written around Paul’s time. 2 Thessalonians, Ephesians, and Colossians are forgeries, but they are consistent with Paul’s outlook. 1 and 2 Timothy and Titus are late forgeries and are inconsistent with Paul’s views.
1, 2, and 3 John, Jude, and James are of uncertain date and uncertain reliability. 2 Peter is a forgery. The biblical consensus is that 1 Peter is a forgery as well based on the gospel story that Peter was an illiterate fisherman, however the Gospels may just be exaggerating Peter’s humble origins and 1 Peter could be genuine.
Revelation’s date and authorship are uncertain, but it was most likely written in the early 90s AD and would therefor not be independent of the Gospels which would have been written earlier.
The traditional date for Mark is around 70 AD, Matthew around 80 AD, Luke around 90 AD, and John around 100 AD. However, they were probably written later than this. Also, we need to keep in mind that all have been modified from their original version.
Carrier guesses Mark was written in the 70s or 80s, but notes that there’s no evidence Mark existed before 100. (I’d think the mention in Mark 13:14 of the desecrating shrine which didn’t happen until the Bar Kockhba revolt would place Mark later than this event, but perhaps this was a later addition.)
Some scholars argue for a second century date for Matthew, but Carrier goes with the date range of 80s to 90s to err on the side of the earliest dating. Since Matthew quotes from Mark we know Matthew at least has to be later than Mark.
Luke also quotes from Mark, and may also borrow sayings from Matthew, so it’s probably later than both. There is strong evidence one of Luke’s sources is The Jewish Antiquities of Josephus and many scholars think Luke/Acts was written after 115 AD. Carrier again errs on the side of the earlier dating and goes with 90s AD for Luke. It’s important to note that there are two different versions of Luke/Acts which means it was heavily edited.
John is a response to Luke. John could have been written after 140 AD, but Carrier uses the earliest possible date of 100s AD.
Some scholars believe a hypothetical document called Q existed, since Matthew and Luke contain similar material, but it doesn’t seem like Luke borrowed it from Matthew. However, it’s more likely Luke did borrow from Matthew, so there’s no good evidence Q existed. Any other hypothetical sources such as the Signs Gospel, M, etc. can’t be used as evidence since they probably didn’t exist.
1 Clement is usually dated to 95 AD, but since Clement is unaware of the Jewish War (among other reasons), Carrier thinks it has to have been written in the 60s AD.
Carrier dismisses most of the non-canonical Christian writings since they’re of uncertain date and thus can’t be proven to be independent of the biblical writings. Likewise, any non-Christian writings we have are late, contain no new information, and are likely dependent on the Gospels.
Everyone agrees Christian documents were rewritten. Most of the redaction would have happened early, before we have copies. This is because editing, adding, and deleting text would have been harder to do later on when there were more copies circulating around. Carrier intends to give the text the benefit of the doubt, however, and assume it’s authentic unless proven otherwise.