Python kind of had three things that really mattered, an culture of “one correct way”, an invisible but conservative type system, and an style guide build into the language an dictated from the language designers in the form of significant whitespace and PEP-8.

Ruby could have been a contender but still had some of the “there is more than one way to do things” left over from perl(ruby was kind of an pragmatic alternative to the “real soon newer” of perl6) and it’s a bit more dynamic than python in how it handles variable internally.

And the field of academic data crunching was hurting badly from the fact that the two only viable options for most entry level researches was Excel/VisualBasic or PERL so everyone wanted some level of structure which the python community led by Guido provided, once people begun to settle on bumpy and pandas.

And pythons popularity definitely grows from the fact that it’s what non programmers learned as an alternative to Excel/SPSS/SAS/Matlab(all of whom are expensive proprietary tools) in classes focused on analyzing data.