Timebox Development has been found to produce extraordinary productivity at DuPont, where it was initially developed. DuPont averages about 80 function points per person month with timeboxing, compared to 15 to 25 with other methodologies (Martin 1991). Moreover, timebox development entails little risk. System evaluation and possible rejection is an explicit part of the practice, but after its first few years of use, DuPont had not rejected a single system developed with timeboxing. Scott Shultz, who created the methodology at DuPont, says that "[a]ll applications were completed in less time than it would have taken just to write the specifications for [an application in] Cobol or Fortran" (Shultz in Martin 1991).