2016-10-27: JPP ≠ SNM!
Matt Marjanovic gave public comment to DBRAC at last night's meeting;
the complete text is linked below. Judy Brennan had announced (three
hours into the
October 22nd meeting,
at 3:04:20) that PPS staff would be
evaluating an "approximate soft neighborhood model". We needed to make
it crystal clear that what she was planning to present was nothing
like the Soft Neighborhood Model.
(Click to open PDF.)
Text of Public Testimony to DBRAC, 2016-10-26
Matt Marjanovic <email@example.com>
At last Saturday's DBRAC meeting, Judy Brennan told you that her staff
would produce three scenarios for review next week, and that the last
of the three would be their, quote, "sort of independent effort to
replicate what some of you know about as the soft neighborhood model".
Judy also said "it's an approximation, it's not a replication". In
truth, it is neither. Whatever Judy and her staff are going to
present to you next week, it will be very different from the Soft
Neighborhood Model. I know this because Brooke Cowan and I developed
the Soft Neighborhood Model, and no one on PPS staff has ever asked us
how it works.
What Judy is going to present to you will be a pretty run-of-the-mill
packing scheme, which the Soft Neighborhood Model very deliberately is
not. I'm also addressing this to any board members present: Judy's
Packing Plan is not what the board asked for in Resolution 5155. What
the board wants evaluated is the actual Soft Neighborhood Model which
Brooke and I presented over a year ago at the board meeting on October
If anyone here would like to know more about how the Soft Neighborhood
Model works or what it does or, just as important, why it does
what it does, we'd be happy to come and discuss it. In the meantime,
please do not confuse it with something else. "Soft Neighborhood
Model" means something very specific, with specific qualities,
implemented with specific math.
Understanding the math is crucial. PPS has a habit of giving new and
unexpected meanings to otherwise well-known words and phrases.
Outside of this room, "over-crowded" means "too many kids per
classroom", but in here it means "too many teachers per classroom".
You avoid these misunderstandings by looking at the math behind the
words. You need to review the math used by staff --- the equations in
the spreadsheets, the database queries, the scribbles on napkins. If
math isn't your thing, find someone outside of this room to help you
work through it, and then please show it and explain it to the public.
Math is crucial here at DBRAC --- what you are tasked to work on here
is, at its heart, a resource allocation problem, complicated by
elements of psychology and culture and history. Every model you
evaluate needs to be approached with the same care and rigor, because
you won't get the right answers by accident. You'll just get it
wrong, and getting it wrong at this scale causes a lot of harm.
So, to review: (1) Judy's Packing Plan is nothing like the Soft
Neighborhood Model. (2) Make sure that staff shows you math, not just
Copyright 2016, Brooke Cowan and Matthew Marjanovic