Shakespeare Statistics
The Authorship of Wily Beguiled
auf deutsch
All data were generated with R Stylo.
See: 
The rolling classify investigation uses the classifiers nsc (nearest shrunken centroid) [lines 1-81], svm (support
vector machine) [lines 83-162] and delta [lines 164-243], which are respectively based on the most frequent function words (mf1w) and on the most frequent character bi-
and trigrams (mf2c and mf3c). A precision of the results again resulted from the methodological step
of using a large set of reference texts which are single-authored and well-attributed.
The use of only a few core texts from larger corpora prevents possible impairments due to the quantity of text.
The window sizes chosen were 1000 to 8000 words at intervals of 1000 words. The respective overlap was each time
250 words lower than the window size, so that a comparability with
rolling delta files was achieved. According to their mathematical kernel
the classifiers have different weights. Nsc has a low decision level, svm has a very high one
and is therefore to be preferred. Character trigrams are usually more accurate than word and bigram classifications.
In his 1922 thesis, Baldwin Maxwell reports on the anonymous comedy Wily Beguiled, which was entered in the Stationer’s Register on 12 November 1606. He records that most critics, however, assume an earlier date of origin and refers to a possible connection with Wylie Beguylie, which was performed at Merton College, Oxford, in 1566/7. He also sees references to the Parnassus Plays, which were performed in Cambridge around 1600, as well as to other plays of this period. On the dating of the comedy, set by Malone at 1596 because Thomas Nashe had spoken of "[the] tricke of Wily Beguily" in Have With You to Saffron Walden, he arrives at Fleay's observation that Romeo and Juliet and The Merchant of Venice were clearly parodied, but not Shakespeare's later plays. Hales and Nicholson would have argued for "in or after 1601", citing the heated dispute between Ben Jonson and his fellow playwrights as the reason. Maxwell finds allusions to Jonson in the text, especially his plays Cynthia's Revels and Poetaster, which supports the latter dates, since Cynthia's Revels was performed in the autumn of 1600 and Poetaster in 1601. Through the attribution of authorship by critics (Bernardi → Shakespeare, Dyce, Fleay, Penniman, Gause → Peele) and his own investigations of alliteration and the quality of Latin phrases, Maxwell finally arrives at Marston, who was at odds with Jonson at this time. But: ”The verse of Wily is most obviously not the verse of Marston ; it is far more lyric and full of more elaborate conceits.” Another of Maxwell's areas of investigation is comedy, which he traces back to Gammer Gurton's Needle, among others, leading him to conclude that Wily Beguiled is a reworking of an earlier comedy, as indicated by patchwork passages in which an editor retained or only slightly altered the original text, but also added lines to make improvements. If the non-preserved text of Wylie Beguylie (1566/7) is the source text, it should have contained many Italian elements, according to Maxwell, for the conventional characters of Wily Beguiled also have the pedant, the nurse, the parasite and the disguised villain. Considering the Rolling Delta and Rolling Classify results under these auspices, a perspective gains weight according to which John Lyly could have done a reworking of Wylie Beguylie, which in turn was revised by John Marston, who was also involved in the Parnassus Plays and by Thomas Dekker. Both were antagonists towards Ben Jonson in The War of the Theatres .
Compare this evaluation with the results of Rolling Delta
or the General Imposters method.