How accurate are all of the Old Norse sagas? | AskHistorians Archive Observer

First of all, it would be convenient to classify Old Norse saga into a few sub-genres, such as:

legendary sagas: on legendary heroes and their deeds before the settlement of Iceland (late 9th century)
sagas of the Icelanders, also sometimes called Family sagas: on the story of Icelandic settlers and their descendants, roughly down to the early 11th century.
king's sagas: is basically royal biographies especially of Norwegian rulers mainly from the 9th/10th to the 13th centuries.
bishop's sagas / (saint's sagas): While some sagas are translated lives of saints (hagiographies), the former is primarily vernacular gesta episcoporum (Life of Bishops, written in Latin out of Iceland).
contemporary sagas: is basically Sturlunga saga compilations, mainly dealing with the events from the early 12th to the late 13th centuries.

Of these sub-genres, modern scholars mainly employ only the latter three (king's sagas/ bishop's sagas/ contemporary sagas) as more or less possibly reliable source of the history of near-past (usually since the 12th century, 1120s onward). As for the 11th century (especially its first half), some scholars tend to avoid to take all of prose narrative in king's sagas at face value, and instead focus on the skaldic poems that the saga author cite as a kind of contemporary testimony of the event. These poems are constructed on so complex rules of meters that they could be transmitted orally in a few generations before the scribe wrote them down later in parchments. There are also some more early historical writings from the 12th century Iceland as well as Norway and Denmark that predate the 13th century major king's sagas, like famous Book of the Icelanders, so it might also be useful to check them as possible more contemporary sources of the 11th and 12th centuries.

On the other hand, it would be difficult now to validate (either to accept or to refute) most of the alleged historical figures and events narrated in sagas of the Icelanders (usually dated from 870s to 1030s) especially in Iceland, since we don't usually have much of alternative source to compare its narratives. While I don't say they are totally useless as source for the distant past of Iceland (it certainly based on the 13th century reconstructed narrative based on the "sea of orally transmitted memory" among Icelanders - see Gísli Sigurðsson 2004), they are still not so straightforward ones. It is also worth noting that alternative source to compare like law texts (Grágás) in extant form either sometimes contradicts with what saga narrates or are only transmitted in later manuscripts. Part of the problem behind Vinland sagas is that both Vinland sagas (Saga of Erik the Red & Saga of the Greenlanders) belong to this sub-genre of Old Norse sagas - authors of this sub-genre rarely cite skaldic poems as a testimony as they do for king's sagas.

In sum, very, very roughly speaking, at least in accordance with current standard of source criticism, I'd not recommend you to employ any sub-genre of the sagas (especially its narratives in prose) as source of historical events and figures before first half of the 11th century, much less prior to the turn of the first millennium.

Recent researchers those who use sagas, especially sagas of the Icelanders, tend to overcome this general problem by not dealing directly with the historicity of the events and persons appeared in sagas per se, but social dynamics narrated in the sagas.