Powers defines Bookmaker Informedness/Markedness using the dtp concept:

“We can gain further insight into the nature of these
regression and correlation coefficients by reducing the
top and bottom of each expression to probabilities
(dividing by N2, noting that the original contingency
counts sum to N, and the joint probabilities after
reduction sum to 1). The numerator is the determinant
of the contingency matrix, and common across all
three coefficients, reducing to dtp, whilst the reduced
denominator of the regression coefficients depends
only on the Prevalence or Bias of the base variates.
The regression coefficients, Bookmaker Informedness
(B) and Markedness (M), may thus be re-expressed in
terms of Precision (Prec) or Recall, along with Bias and
Prevalence (Prev) or their inverses (I-):”

Full paper link: https://arxiv.org/pdf/2010.16061

Bookmaker Markedness = dtp/ [Bias · (1-Bias)]
Bookmaker Informedness = dtp/ [rp·rn]

I didn’t quite get what DTP is. It seems to be the determinant of the Contingency Table, but I’m not 100% sure either of DTP calculus or its acronym meaning. Should I divide the whole matrix by N² before calculating the determinant?

I’m expecting a formal definition of DTP meaning and its calculus