The relationship of plays to their sources has always been important evidence of chronology, authorship, and the derivation of textual variants. Such evidence has been particularly important to studies of Shakespeare’s early plays. But for centuries source scholarship has been based on random anecdotes: a scholar reading one text notices something about it that reminds them of another text. We can now re-evaluate those anecdotal findings by testing them systematically against digital databases. Such tests establish that Margaret's long speech at the beginning of Scene 2 of The First Part of the Contention is based on a passage in Hall's chronicle, whereas the variants in the Folio text of 2 Henry VI instead draw upon Holinshed's chronicle. This evidence supports revision rather than memorial reconstruction. Likewise, the links between the Contention speech and Edward II are best explained by Marlowe's authorship of both.