►
From YouTube: NCSL Redistricting Seminar | Audit the Class
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
All
right
good
afternoon,
everybody
welcome
to
ncsl's
virtual
redistricting
seminar
audit,
the
class.
We
want
this
to
be
an
interactive
experience
for
all
of
you,
so
I'm
briefly
going
to
point
out
a
couple
of
key
features
of
this
page
to
the
right
of
the
screen.
Where
you
can
see
me
you'll
note
the
tabs
that
say
chat
and
q.
A
the
chat
function
is
to
allow
you
to
interact
with
colleagues
attending
this
session.
A
So
this
is
a
new
thing
that
we're
doing
in
ncsl.
We've
never
had
an
audit
the
class
session
before,
but
we
think
that,
because
redistricting
is
one
of
these
areas,
that's
changing
so
rapidly.
It's
useful
to
bring
in
people
who
are
doing
some
really
deep
thinking
about
what
redistricting
is
and
where
it's
headed
and
to
present
their
research
to
you
so
that
you
can
get
some
cutting
edge
information
that
in
a
normal
world,
may
not
be
presented
to
you
for
a
couple
years
until
it's
used
in
litigation,
for
example.
A
So
we've
split
this
time
into
two
periods
at
for
the
first
half
hour,
we're
going
to
have
dr
munduchin
from
tufts
university
again,
who
you
saw
earlier
today
and
we're
going
to
have
sam
hirsch
from
partner
jenner
and
block
who
works
on
redistricting
and
election
law
there,
as
well
as
some
other
practice
areas
and
they've
written
a
new
paper
that
I
know
that
they're
excited
to
share
with
you
and
then
for
the
second
half
of
this
hour.
A
We're
going
to
have
peter
watson,
who
is
a
former
attorney
with
the
minnesota
legislature
and
his
and
all
of
these
people
are
longtime
friends
of
ncsl.
So
we're
very
grateful
to
have
all
of
them
with
us
today,
but
for
this
first
half
hour
it's
going
to
be
moon
and
sam,
I'm
going
to
hand
it
off
to
sam.
You
first.
B
Well,
thank
you
ben,
and
I
I
really
want
to
thank
ben
and
wendy
and
christy
for
the
amazing
work.
They've
done
to
put
this
entire
seminar
together,
especially
in
the
conditions
of
the
pandemic,
that
we're
all
dealing
with.
It's
really
a
wonderful
service,
and
I
also
want
to
thank
the
legislators
and
commissioners
and
staff
who
are
taking
time
to
participate
in
the
seminar
and,
more
importantly,
for
the
work
you
do
to
make
our
democracy
function
and
function
better.
It's
an
honor
to
be
on
this
hour-long
panel,
with
with
peter
watson.
B
I've
been
attending
ncsl,
registering
events
since
the
90s
and
one
of
the
things
I
always
look
forward
to
is
learning
from
peter,
who
is
a
national
treasure
in
these
areas.
Moon
and
I
are
going
to
share
the
first
half
hour
and
talk
about
a
paper
that
we've
recently
co-authored
for
those
of
you
that
were
on
moon's
earlier
session
today.
You
know
how
amazing
she
is
and
how
brilliant
she
is
at
explaining
unbelievably
complex
things.
B
In
plain
english,
I
also
wanted
to
honor
our
two
co-authors
dara
gold,
amy
becker,
who
unfortunately
can't
be
with
us
today,
but
every
decade
in
redistricting
there
are
some
experts
who
rise
up
and
become
superstars
of
the
next
generation
and
dara
and
amy.
I
I'm
convinced
are
going
to
be
exactly
that.
They
are
wonderful
and
the
paper
that
moon
and
I
are
going
to
present
could
never
have
been
put
together
without
amy
and
dara.
B
By
the
way,
the
paper
is
almost
ready
for
prime
time.
So,
if
you're
interested
in
seeing
a
copy,
when
it
is
ready,
email
me
at
s
hersh
at
jenner.com
and
I'm
happy
to
send
it
to
you.
So
look
registering,
is
all
really
about
choices
and
trade-offs,
but
to
understand
what
choices
are
available
and
what
the
trade-offs
look
like.
You
have
to
kind
of
know
what
options
are
actually
feasible,
what
options
are
kind
of
normal?
B
What
options
are
extreme
outliers
and
if
you
think
about
like,
if
you're
drawing
a
100
district
map
say
for
your
state
house,
almost
anything
you
care
about,
presents
the
same
kind
of
of
questions.
B
B
What's
a
reasonable
number
of
competitive
districts,
what's
an
unusually
high
number,
what's
an
unusually
low
number
or
if
you
care
about
respect
for
political
subdivisions
that
may
be
in
your
state
constitution
again,
what's
you
know
we
we're,
probably
gonna,
have
to
break
up
some
counties
just
to
comply
with
one
person,
one
vote:
what's
sort
of
a
normal
number
of
counties
end
up
broken?
What's
the
rock
bottom
minimum,
you
can
do
what's
what's
really
a
bad
map
in
terms
of
just
breaking
them
needlessly
and
for
no
good
reason
to
understand
any
of
these
things.
B
You
can't
really
do
it
in
the
abstract.
You
have
to
look
at
actual
maps
and
now,
thanks
to
the
kind
of
computer
science
that
moon
and
her
colleagues
at
the
redistricting
lab
at
tufts,
can
bring
to
bear.
You
can
only
look
at
a
handful
of
maps
drawn
by
by
human
beings.
You
can
look
at
thousands
or
millions
of
maps,
and
this
is
so
exciting,
but
but
you
you
can't
look
at
all
the
possible
maps
in
the
universe.
B
B
So
if
you're
drawing
100
district
map,
you
don't
want
to
compare
it
to
what
what
an
80
or
a
200
district
map
looks
like,
because
that's
kind
of
irrelevant
you
wouldn't
want
to
look
at
maps
where
one
district
has
twice
as
many
people
as
another
district,
because
that
would
be
an
obvious
violation
of
one
person.
One
vote.
You
wouldn't
want
to
look
at
a
map
where
you
know
district
three
is
half
in
the
southeast
corner
of
your
state
and
half
in
the
northwest
corner
with
no
connection.
B
B
Let's
say
you
have
minority
voters
and
non-minority
voters
and
the
minority
voters
like
candidate,
a
and
they'll
vote
for
her
in
the
primary
and
if
she
wins
nomination,
they'll
vote
for
her
in
the
general
candidate
is
her
candidate.
If
the
non-minority
voters
also
like
candidate
a
and
this
pattern
repeats
in
all
sorts
of
elections,
then
well,
we
know
a
couple
things
once
we
know
candidate
is
going
to
win.
Two
is
we
know?
B
Typically,
if
the
minority
voters
prefer
canada
today,
the
nominated
voters
might
prefer
candidate
b
and
then
the
question
is
who's
who's
going
to
prevail,
and
when
you
have
a
districting
plan,
let's
say
you
have
100
districts
again:
the
can
the
candidates
like
canada
today
will
prevail
in
some
number
of
those
districts,
and
the
candidates
like
canada
b
will
prevail
another.
So
the
minority
preferred
candidates
will
prevail
in
some
places
and
they'll
lose
in
other
places
and
what
the
voting
rights
act
speaks
to
is.
B
Is
that
a
fair
balance
between
the
districts,
where
the
more
preferred
candidates
winning
and
losing?
And
let's
say
the
state-
is
30
minority
just
making
this
up?
So,
if
canada,
today
and
folks
like
canada,
are
winning
30
of
the
districts
pretty
consistently,
that
seems
fair
right,
30
of
the
population
of
slave
adult
citizens,
30
of
the
districts
coming
out
in
a
minority
preferred
way.
B
So
that's
really
what
the
voting
rights
act
is
about
and
then
pressing
against
in
some
ways
are
the
recent
interpretations
by
the
supreme
court
of
equal
protection
clause
and
what
they
really
say,
that's
most
relevant
for
our
purposes.
Right
now
are
stay
focused
on
actual
election
results
notice.
I
wasn't
talking
a
minute
ago
about
how
many
majority
minority
districts
need
to
draw.
I
was
talking
about
whether
minority
voters
have
actual
equal
electoral
opportunity
to
non-minority
voters.
That's
the
question
so
stay
focused
on
election
results
and
election
data
and
don't
rely
on
arbitrary
demographic
cutoffs.
B
So
the
voting
rights
act
and
the
protection
clause
are
not
about
drawing
the
right
number
of
majority
minority
districts
they're
not
about
drawing
the
right
number
of
60
native
american
districts
or
40
percent
black
districts.
It's
not
about
demographics
and
demographics
alone.
It's
about
electoral
opportunity!
B
D
And
so
I
appreciate
that
my
job
here
is
super
hard,
which
is
to
take
something:
that's
exquisitely,
complicated
and
not
only
try
in
the
first
instance
to
make
it
accessible
to
a
computer
rather
than
like
a
seasoned
voting
rights
attorney,
but
also
to
take
all
of
that.
You
know
work,
that's
in
this
new
paper
that
that
sam
mentioned
and
and
boil
it
down
into
10
minutes
right
now.
D
D
Okay,
so
I
spoke
earlier
today
about
ensemble
method,
so
here's
the
tldr,
we
have
techniques
that
have
evolved
over
the
last
decade
and
maybe
especially
the
last
five
years
for
making
lots
of
different
like
diverse
collection
of
plans
that
satisfy
the
basic
criteria,
the
basic
kind
of
threshold
criteria
for
consideration
as
plausible
plans.
D
You
can
do
this
now
with
free
public
software,
and
you
know
there's
jerry
chain,
for
instance,
which
is
what
was
used
to
make
the
map
that
you're,
seeing
that's
a
python
package
or
a
corresponding
julia
package,
and
you
could
go
play
with
them
today
and
I
hope
that
you
do
they're
fast.
So
I
mentioned
earlier
this
morning.
The
the
julia
version
recently
clocked
a
million
plans
in
eight
seconds.
That's
blazingly
fast.
I
think
you
can
call
it,
but
that's
not
that
interesting.
Just
quantity
isn't
that
interesting.
D
Also,
these
come
with
pretty
good
theoretical
properties
that
tell
you
something
about
quality.
They
tell
you
that
you
have
good
evidence
that
you're
sampling
representatively
and
it's
come
up
in
the
last
cycle.
It's
come
up
in
partisan,
gerrymandering
litigation,
but
I
think
looking
forward,
these
ensemble
methods
are
going
to
get
used
in
lots
of
different
ways.
So
let
me
talk
a
little
bit
about
folding
vra
consideration
into
ensemble
methods
right.
So
sam
just
discussed
the
voting
rights
act
hugely
important
civil
rights
statute,
but
it's
got
very
complicated
case
law
around
it.
D
That
tells
you
how
you
can
go
about
pressing
a
suit
and
there's
a
whole
range
of
multi-disciplinary
range
of
questions
that
you
have
to
ask,
and
so
to
get
a
computer
to
kind
of
make
a
snap
assessment
of
vra
compliance
is
no
easy
task,
but
we've
come
up
with
a
protocol
that
we'd
like
to
describe
here
that
really,
as
sam
said,
centers
electoral
history
and
that's
what's,
I
think
so
important
about
it.
So
this
is
just
for
emphasis.
What
not
to
do
how
how
not
to
take
the
vra
into
account.
D
So,
first
of
all,
you
certainly
can't
now
that
we're
in
a
new
census
cycle.
You
can't
just
freeze
the
last
cycles,
vra
districts,
because
they
no
longer
have
the
right
population
under
the
new
census
data,
or
it
would
be
a
minor
miracle
if
they
did
so
you're
going
to
have
to
change
those.
You
can't
just
freeze
the
old
districts.
D
It's
not
good
enough
to
just
look
for
majority
minority
districts,
because
that's
not
a
reliable
indicator
of
whether
they'll
perform,
whether
they
will
actually
provide
opportunities
to
for
people
of
color
to
elect
candidates
of
choice,
and
you
can't
specify,
as
you
heard,
a
target
demographic
share.
It's
not
just
50
that
won't
work,
you
can't
say
60,
you
can't
say
45.,
just
a
demographic
share
isn't
enough!
You
have
to
look
at
elections,
okay.
So
how
does
this
new
protocol
work?
D
It
starts
with
elections,
so
I've
got
some
images
here
of
my
home
state
of
massachusetts,
showing
you
the
precincts
of
the
state
and
color
coded
by
how
they
voted
in
these
different
races,
and
so
you
can
see
which
contest
you
look
at
matters.
Of
course,
in
new
england
we
love
our
republican
governors,
so
you
can
see
that
you
get
different
patterns
under
different
races.
D
You
want
to
take
elections
and
put
them
together
into
a
data
set,
and
it's
really
important-
and
you
heard
this
this
morning
as
well
on
the
vra
panel-
it's
not
good
enough
to
look
at
general
elections.
You
must
consider
primaries
because
people
of
color
have
to
be
able
to
get
a
candidate
of
choice
through
the
primary
and
into
the
general
in
the
first
instance.
It's
not
good
enough
to
just
have
a
preferred
candidate
win
the
general.
D
The
hard
part,
as
you
all
know,
and
have
been
hearing
about
from
a
great
many
speakers
here
at
ncsl,
is
to
locate
the
votes
in
space.
So
you
want
a
shape
file
that
tells
you
where
those
cast
ballots
were
cast
where
they're
located.
Once
you
have
all
this,
you
have
your
election
database
now.
What
now
there's
the
step
that
we
call
a
racially
polarized
voting,
so
you
have
to
identify
the
candidates
of
choice
and
look
to
see
whether
it's
the
case
that
that
that
voters
of
color
are
preferring
different
candidates
from
the
majority
group.
D
There
are
lots
of
methods
for
estimating
who
the
candidate
of
choice
might
be,
but
the
the
leading
method
these
days
is
well,
it's
actually
a
family
of
methods
called
e.I
or
ecological
inference,
and
it's
a
statistical
inference
techniques
for
for
for
looking
at
how
well,
if
you
look
precinct
by
precinct-
and
you
compare
the
demographic
makeup
of
the
precinct
to
the
voting
preferences,
it's
basically
looking
for
trends
in
so
maybe,
as
your
precincts
get
more
heavily
poc,
they
tend
to
prefer
certain
candidates.
That
would
be
a
strong
indicator.
D
So
in
the
example
that
I've
shown
you
here,
what
I'd
like
to
highlight
is
that
you
have
to
keep
track
of
uncertainty.
So
if
you
take
a
look
at
these
elections
in
texas,
what
you
can
see
is
that
in
many
instances
the
candidate
of
choice
would
be
the
same
for
latino
voters
and
for
black
voters,
but
there's
a
few
elections
where
that
may
not
be
true,
and
if
you
focus,
for
instance,
on
the
2018
primary
runoff
for
governor,
what
you
can
see
is
that
there
were
two
candidates.
D
The
way
to
read
this
table
is
that
the
first
column
is
telling
you
who
the
candidate
of
choice
seemed
to
be
for
latino
voters,
so
you're,
seeing
here
high
confidence,
a
thousand
out
of
a
thousand
draws
indicate
that
latino
voters
preferred
veldes
in
this
election,
but
for
black
voters
it
was
less
clear.
429
draws
said
valdez
and
571
draws
said
andrew
wakes,
we're
not
sure
and
whatever
techniques
we
build
up,
that
rely
on
inferring
the
candidate
of
choice.
D
We
want
to
track
how
sure
we
are
that
we've
chosen
identified
the
right
candidate
once
you've
got
your
candidate.
What
can
you
do
so?
So?
Here's
the
center
piece
of
this
protocol
it's
to
build
an
effectiveness
score
with
electoral
history
at
its
heart.
So
what
we
want
to
do
is
build
a
score
where
zero
and
one
have
a
meaning,
a
score
of
zero
means.
The
candidate
of
choice
is
never
winning
score
of
one
means.
D
The
candidate
of
choice
is
always
winning
and
all
the
primaries
and
all
the
generals,
and
we
want
to
build
a
score
that
interpolates
between
those.
So
what
I'm
showing
you
here
is
a
screenshot
from
a
new
dashboard
that
that
my
group
has
been
building
so
that
you
can
play
with
this
yourself
and
you
paint
in
a
district
here
we're
looking
at
louisiana,
and
what
you
can
see
is
that
this
pale
green
district
is
being
rated
12.4
percent
effective
and
the
the
yellow
district,
which
includes
most
of
baton
rouge.
D
There
is
being
rated
72.1
percent
effective,
but
the
dashboard
shows
you
why
it
shows
you
that
this
doesn't
have
to
do
with
the
percentage
of
black
voters
that
is,
or
at
least
not
directly,
the
percentage
of
black
voters
in
each
district,
but
how
they
voted.
And
so
here's
a
whole
range
of
recent
elections
telling
you
how
the
camp
who
the
candidate
of
choice
was
and
how
they
fared
under
the
hood.
D
We're
awaiting
these
elections
by
how
sure
we
are
that
we've
got
the
right
candidate
by
how
recent
the
election
was
and
several
other
factors
that
might
be
connected
to
how
informative.
We
think
that
outcome
is
okay,
so
we
wait
the
elections,
as
I
just
said,
and
then
we
calibrate
that
to
look
at
recent
district
effectiveness.
D
D
Okay
and
then,
once
you
have
all
that
you're
in
business,
you
can
sort
of
hitch
that
to
the
engine
of
ensemble
generation
that
builds
all
these
alternative
plans
decide
how
effectiveness,
how
high
a
level
of
effectiveness
you're
going
to
require
as
your
threshold,
we
recommend
using
at
least
60,
possibly
higher,
so
that
you
can
be
confident
that
it's
more
likely
to
perform
than
not.
D
This
is
a
user
choice,
and
then
you
run
the
ensemble
generation
technique,
while
making
sure
that
each
plan,
in
order
to
be
included,
has
at
least
as
many
effective
districts
as
the
benchmark.
By
the
way,
this
gets
back
to
a
question
that
I
remember
was
asked
in
the
vra
session
earlier
this
morning.
How
many
effective
districts
do
we
need
to
build
right
and
that's,
of
course,
a
very
delicate
question.
D
D
What
we
found
is
that
for
the
computer,
it's
no
problem
to
find.
We
ran
it
at
sixty
percent
effective.
We
ran
it
at
seventy
five
percent,
effective
at
whatever
level
we
chose
to
run
it.
We
were
finding
it
to
be
relatively
easy
to
generate
lots
of
different
alternatives
that
might
not
have
been
obvious
to
the
naked
eye.
D
It's
really
important
to
think
about
multiple
groups
and
clearly
in
texas
for
instance.
Of
course,
both
black
voters
and
latino
voters
are
numerous
enough
to
be
highly
highly
relevant
for
vra
consideration
at
every
level
of
redistricting,
including
congressional,
and
so
what
we
made
sure
to
do
was
to
build
our
effectiveness
course.
In
view
of
the
need
to
understand
how
effective
a
district
might
be
for
latino
preferences
and
for
black
preferences
in
particular,
you
have
to
ask
the
question:
do
latino
and
black
preferences
align
with
each
other
and
that
again,
that's
a
question.
D
You
can
just
interrogate
your
last
elections,
run
your
racial
polarization
and
just
like
in
that
table
that
I
showed
you
before
black
and
latino
preferences
align
a
great
deal
of
the
time,
but
not
always,
and
so
you
see
these
different
districts.
These
are
actual
numbers
of
effectiveness
that
we
rated
in
a
few
actual
current
texas
congressional
districts
and
they
perform
differently
electing
vc,
green
and
garcia.
D
Okay,
and
then
you
know,
every
single
person
that
you've
heard
from
today
has
really
encouraged
you
to
think
about
not
just
using
demographics,
for
effectiveness,
but
actually
using
elections,
and
here
this
is
a
slide
just
intended
to
drive
that
point
home.
So
what
you
see
here
is
a
scatter
plot
of
many
tens
of
thousands
of
districts
that
were
generated.
I
think
it's
100
000
in
each
picture
and
they're
located
in
space
by
their
demographics,
on
the
x-axis
they're,
partisan,
lean
on
the
y-axis
and
then
they're
color-coded
by
effectiveness.
D
The
only
point
I'm
trying
to
make
on
this
slide
is
that
if
you
just
try
to
find
a
dividing
line
in
demographics,
it
would
not
work
very
well.
In
other
words,
you
cannot
find
a
vertical
line
that
separates
the
blue
from
the
red
points.
In
these
plots
it
just
can't
be
done,
and
that's
just
a
reminder.
Demographics
alone
won't
tell
you
the
whole
story.
When
it
comes
to
effectiveness,
the
story
is
a
lot
more
complicated.
D
So
it's
not
just
constitutionally
shaky.
It's
like
poor
correlation.
If
you
try
to
rely
on
demographics,
a
lot
and
then
vote
I'll,
just
close
this
down
with
you
know,
overall,
what
are
we
finding
that
we
think
is
is
really
useful.
I
hope,
if
you
focus
on
electoral
effectiveness,
algorithms
can
build
big
samples
of
plans
that
show
you
lots
of
different
ways
to
get
opportunity
for
minority
voters
coalitionally
separately.
B
Thanks
man,
so
one
one
last
thing
that
we
found
that
was
really
interesting,
which
is
texas,
is
a
majority
minority
state
according
to
the
2010
census,
and
probably
has
become
more
so
in
the
last
decade.
B
According
to
the
most
recent
census
data,
it's
just
over
half
non-minority,
not
non-hispanic
white
among
the
adult
citizen
population,
and
it
has
36
districts.
One
thing
we
found
is
the
the
voting
in
texas.
Almost
everywhere
is
highly
racially
polarized.
B
B
But
when
we
looked
at
the
actual
map,
that's
in
place
right
now
that
wasn't
the
case
at
all
early
in
the
decade
it
had
probably
11
out
of
36
and
because
of
some
changes
in
voting
behavior
and
demographics
in
the
dallas
fort
worth
and
houston
areas
that
had
come
up
to
13
out
of
36
in
in
the
latter
half
of
this
decade,
but
still
way
short
of
that
17
out
of
36.
That
would
be
roughly
proportional,
and
the
question
is
like:
is
that
just
inevitable?
B
Is
that
just
because
the
way
you
know
the
geography
of
the
state,
the
political
geography
of
the
state,
or
is
that
an
artifact
of
just
how
these
lines
were
drawn?
And
the
answer
is
the
latter,
because
it
turns
out
that
using
the
very
same
techniques
that
moon
just
outlined,
we
could
generate
a
ton
of
maps,
hundreds
of
thousands
and
assess
them,
and
we
found
it's
not
at
all
hard
to
move
from
13
minority
preferred
districts,
minority,
effective
districts
to
15
or
16,
or
even
17,
which
is
actually
roughly
proportional.
B
So
that's
a
really
interesting
finding.
We
don't
know
if
that'll
apply
in
all
states,
but
it's
it's
a
very
interesting
finding
and
it's
an
important
finding
and
here's
why,
in
almost
all
legislatures,
federal
or
state,
historically
underrepresented
minority
groups
are
today
still
underrepresented.
This
is
true
in
the
house
of
representatives.
You
know.
Roughly,
a
third
of
all
adult
citizens
in
the
united
states
are
are
not
white,
but
only
about
a
quarter
of
the
house
is
so
that's
a
significant
underrepresentation
and
there's
at
least
some
underrepresentation
of
minorities.
B
Last
we
counted
which
was
right
before
the
last
election,
but
last
time
we
counted
in
47
of
the
50
state.
Legislatures
minorities
are
underrepresented
when
you
look
at
their
fraction
of
the
adult
citizenry
versus
the
fraction
of
districts
that
seem
to
be
controlled
by
minority
voters.
So
you
know
that
that's
that's
a
a
concern,
a
real
problem
and
it
may
not
be
an
inevitable
one.
It
may
not
be
an
unfixable
one,
so
that's
a
really
interesting
finding.
B
D
D
And
then
sam,
we
can
up
how
we
might
answer
them.
Okay,
so
I
see
a
question
from
ben
clark.
Who's
been
asking
lots
of
great
questions
today,
ben,
we
should
definitely
be
in
touch
and
he
asked
if
the
vra
analysis
is
operationalized
in
jerry
chain,
julia
yeah,
so
we've
been
building
this
into
our
chain
building
software.
D
It
will
be
there's
going
to
be
a
new
release
of
several
of
our
packages
in
the
next
week
or
two,
but
if
you
want
to
kind
of
the
dashboard
that
I
showed
a
screenshot
of
is
also
we're
getting
ready
to
go
live
and
if
anybody
would
like
to
be
an
early
test
user
for
that
we'd
love
the
feedback
so
just
be
in
touch.
If
you'd
like
to
see
how
we're
building
this
into
a
tool
that
that
advocates
and
reformers
and
attorneys
and
the
public
can
use
to
to
understand
how
this
electoral
history
fits
together.
D
Here
are
some
other
questions,
so
here's
one
sam-
maybe
you
can
start
this
off
so
adeline
wilcox
asks
in
the
2018
democratic
primary
for
maryland's
fifth
steny
hoyer
defeated
a
candidate
named
dennis
fritz,
whose
image
looks
black
to
me.
84.1
to
15.9
percent
and
defeated
a
candidate
michaela
wilkes,
who
looks
black
to
me,
64.4
to
26.7
percent.
Is
there
a
vra
case
here.
B
The
the
quick
answer
is,
I
have
no
idea
it's
it's
a
very
complicated
thing
to
say
whether
you've
got
a
valid
vra
claim
or
not,
and
one
thing
we
should
emphasize
is
the
the
method
we've
just
laid
out
is
designed
to
make
sure
that
these
ensembles
of
thousands
or
millions
of
alternative
plans
that
you
can
use
for
comparative
comparison's
sake
generally
include
maps
that
seem
to
comply
with
the
vra
and
generally
exclude
maps
that
seem
to
violate
the
vra
but
they're,
not
a
perfect
proxy
for
actually
litigating
these
cases.
B
The
way
you
do
in
court
with
a
lot
of
detailed
evidence
and
a
lot
of
specificity.
So
I'm
not
familiar
enough
with
the
exact
case
you're
raising
in
this
question,
but
you
could
never
from
the
sort
of
rough
cut
version
of
the
vra
and
equal
protection
clause.
Compliance
that's
built
into
this
model
say
definitively
that's
a
lawful
plan.
That's
an
unlawful
plan.
It's
not
built
for
that
purpose.
It's
also
not
built.
B
If
you're
trying
to
answer
questions
like
how
is
this
on
partisan
fairness
on
competitiveness,
on
respect
for
political
subdivisions,
on
whatever
it
is
you
care
about,
and
also,
as
we
point
out
the
end,
the
the
in
the
last
piece
where
I
spoke,
it
may
also
point
our
way
to
how
to
draw
districts
that
are
more
effective
for
minority
voters,
where
they've
been
underrepresented
to
close
that
gap
and
get
them
closer
to
a
fair
level
of
representation.
D
I
would
add
to
that
only
that
standing
lawyers
district
is
interesting.
It
happens
to
be
one
of
only
two
in
the
country
that
is
over
37
percent
b
vap,
but
doesn't
have
a
a
black
representative,
and
so
because
it's
a
that
district
is,
if
I
remember
right
about
38
vap
the
fact
that
that
steny
hoyer
got
84.1
in
a
primary
suggests
that
he
is
a
popular
candidate
with
people
of
color
as
well.
B
If
I
can
add
to
that
as
a
legal
matter,
what
really
matters
is
not?
Is
the
candidate
black
or
latino
or
white?
What
have
you
it's
is
this
the
preferred
candidate
of
the
voters
who
are
black
or
latino
or
asian,
american
or
native
american?
Or
what
have
you
that's
the
legal
question,
and
it
is
quite
possible
in
some
instances
that
a
candidate
of
one
race
will
be
the
preferred
candidate
of
a
community
defined
by
a
different
race,
and
we
shouldn't
make
sort
of
sweeping
assumptions
along
those
lines.
B
That
said,
the
case,
law
puts
more
emphasis
on
analysis
of
contests
that
involve
candidates
of
different
races,
because
sometimes
they're
more
revealing
than
can
than
contests
between
candidates
of
the
same
race.
D
Right
exactly
agreed,
just
it
seems
like
standing
higher,
has
lots
of
black
support
as
the
short
answer.
So
he
may
well
be
the
black
candidate
of
choice
in
some
of
those
elections.
Maybe
we
can
do
one
more
quickie
and
then
toss
it
to
peter
kendra
asks,
mr
hirsch.
What
are
the
three
states
that
do
have
proportional
minority
representation
in
their
legislatures.
B
You
know
I
apologize.
I
should
have
that
at
my
fingertips
and
I
don't
shoot
me
an
email
at
s
hersh
at
jenner.com,
and
I
will
pull
that
up
and
answer
your
question.
D
Great
okay,
peter
take
it
away.
A
A
So
let's
just
go
ahead
and
I
will
do
it,
and
I
know
that
those
of
you
in
the
and
the
comments
in
the
chat
will
be
doing
your
virtual
rounds
of
applause
as
well,
and
moon
and
sam
are
are
very
responsive,
we'd
love
to
be
in
contact
with
all
of
you,
I'm
sure
they're
happy
to
drop
their
contact
info
in
the
chat.
If,
for
whatever
reason
you
miss
it
and
you
want
to
get
in
contact
with
them
later,
you
can
email
us
at
ncsl,
just
email
elections
dash
info
at
ncsl.org.
A
That's
the
elections
and
redistricting
program
general
inbox
and
we'll
put
you
in
contact
with
them.
So
with
that
we're
gonna
turn
it
over
to
you
peter
and
let's
go
ahead,
and
what
are
you
gonna
be
talking
to
us
about.
C
I
will
say
thank
you
ben
and
thank
you
to
all
of
you
who
have
logged
in
if
you
haven't
already.
Please
enter
your
name
and
your
state
into
the
chat
box.
C
A
They're
they're,
coming
in
I'm
seeing
the
numbers
go
up
and
fluctuate.
People
are
filling
it
out
I'll,
let
you
know
when
they,
when
it
stops.
A
C
Of
rookies
a
lot
of
rookies,
but
enough
veterans
that
we
have
people
around
us
that
we
can
consult
about
how
to
do
things
better.
Terrific.
Now,
let's
take
the
second
poll
poll
number
two.
C
C
A
A
C
A
Can
see
that
a
lot
of
vast
majority
legislative
body,
members
or
staff
and
16
commission
members
are
staff
considering
the
relative
number
of
commissioners
to
legislators,
I
feel
like
that's
pretty
good.
C
A
And
peter,
if
I
could
interrupt
just
real
quickly
for
everyone,
I
know
that
some
people
tend
to
hold
their
questions
until
the
very
end.
Peter
has
specifically
requested
that
the
questions
come
adam
live.
He
wants
to
answer
him
in
the
moment.
So,
as
you
think
of
questions,
don't
hesitate,
put
them
in
that
q
a
chat,
and
I
will
come
on
and
interrupt
peter
to
ask
but
peter
for
now
the
floor
is
yours.
Okay,.
C
C
The
the
five
documents
are
all
posted
on
the
ncsl
website
and
let's
go
to
the
first
one.
I
will
share
my
screen
and
see
what
we
discover
here.
C
Now
this
is
many
of
you
have
seen
this
already,
but
if
you
haven't,
please
do
become
familiar
with
it.
This
is
the
ncsl
redistricting
website,
the
the
page
that
deals
with
action
on
redistricting
plans,
and
it
includes
both
success
rates
and
a
table
for
each
of
the
last
three
decades
and
I'm
going
to
take
you
through
the
success
rates
and
the
table
for
the
most
recent
decade.
C
So
this
is
success
and
in
line
with
what
I
said
about
my
definition
of
success
as
you'll
see
in
this
table,
success
means
the
plan
was
either
not
challenged
in
court
or,
if
challenged
was
upheld
without
change.
In
other
words,
the
plan
made
it
all
the
way
through
the
decade
and
was
used
at
every
election
for
the
balance
of
this
10
years.
C
If
you
take
a
look
at
the
ancient
history
of
the
70s
and
80s
you'll,
see
that
both
legislatures
and
commissions
were
about
as
successful
two-thirds
of
their
plans
made
it
through
without
having
to
be
changed
before
the
end
of
the
decade
in
the
1990s
commissions
upped
their
game
a
bit.
They
got
into
the
80s
and
90
percent
of
success
rate,
but
the
legislatures
fell
down
a
little
bit,
and
why
was
that?
C
In
those
days
section
five
was
in
its
first
round
of
of
being
used
in
the
way
that
it
was
for
a
couple
of
decades
and
the
justice
department
insisted
on
states
drawing
what
we
might
call
max
bot
backs.
Black
plans
draw
districts
with
as
many
african-american
majority
districts
as
the
population
would
permit,
regardless
of
whether
they
might
look
a
little
funny.
C
But
when
those
plans
got
to
court,
particularly
in
1993
with
shaw
versus
reno
justice,
o'connor
said
that
bizarreness
is
a
bad
thing
in
redistributing
plans
and
said
you
can't
do
that.
Other
plans
in
georgia,
louisiana
and
other
states
were
likewise
thrown
out
by
the
courts
after
they
had
been
approved,
insisted
upon
by
the
justice
department
in
the
2000s.
C
The
legislatures
did
a
little
bit
better
and
for
the
very
first
time
they
actually
did
better
with
their
senate
plans
than
the
commissions
did
we
move
into
the
most
recent
decade
and
we
see
that
the
commissions
fell
down
a
little
bit
now.
The
difference,
as
you
can
see,
is
that
instead
of
10
out
of
14
plans
being
approved,
only
9
out
of
14
plans
were
approved.
I
don't
know
exactly
the
reason
why,
for
that
one
planned
difference,
but
I
have
an
idea
why
the
legislature's
success
rate
went
up
so
much
now.
C
The
first
explanation,
I'd
like
to
give
is
because
legislators
and
legislative
staff
have
spent
the
last
three
or
more
decades
coming
to
seminars
like
this
sponsored
by
ncsl,
in
order
to
learn
how
to
draw
plans
that
will
stand
up
in
court
that
that's
the
favorable
view
of
that
improvement,
but
a
little
bit
of
the
polishes
taken
off
of
that
by
the
fact
that
one
reason
for
the
improvement
was
because
of
the
supreme
court's
decision
june
of
2019
in
the
north
carolina
case,
rucho
versus
common
cause,
which
said
that
federal
courts
would
no
longer
consider
claims
that
redistricting
plans
were
a
partisan
gerrymander.
C
That
saved
two
house
plans,
one
senate
plan
and
four
congressional
plans,
thus
raising
the
percentage
by
four
percent
for
senate
plans
and
seven
percent
each
for
house
and
congressional
plans
still
would
have
been
better
than
the
commissions
okay.
So
I'm
a
champion
here
for
legislatures
and
legislative
education.
C
C
Yes,
it
was
finally
approved,
but
no
it
didn't
make
it
all
the
way
through
the
decade,
in
the
same
form
that
the
legislature
had
passed
it
at
the
beginning,
same
kind
of
fate
for
the
senate
plan.
The
congressional
plan
did
a
little
better
because,
even
though
it
was
challenged
in
federal
court,
it
was
approved.
C
So
that's
a
legislature:
here's
a
commission
alaska,
it
drew
the
plan
it
got.
Thrown
out
in
state
court,
went
back,
got
redrawn
was
thrown
out
again
in
state
court,
went
back,
got
redrawn
and
was
finally
approved,
but
again
because
it
didn't
make
it
all
the
way
through
without
change.
I
give
it
an
r
for
rejected
I'd,
call
it
a
failure
same
with
the
senate
plan
and,
of
course,
the
no
congressional
district
needed
in
alaska
because
they
have
only
one
member
of
congress.
C
So
that's
how
the
table
works
and
it
looks
kind
of
bad
people
yesterday
were
talking
about.
Oh
you're,
certainly
going
to
get
sued
everybody's
going
to
get
sued
plan
to
be
in
court.
I
certainly
endorse
that
idea.
Plan
to
be
in
court
be
prepared
to
be
in
court
but
realize
that
not
everybody
went
to
court.
Delaware
is
a
state
that
the
legislature
drew
the
plan
approved
by
the
governor.
It
was
not
challenged.
Likewise,
with
the
with
a
senate
plan
not
challenged,
they
didn't
need
a
congressional
plan.
There
are
other
states.
C
Massachusetts
and
and
others
on
the
next
page
that
were
not
challenged.
In
fact,
during
this
past
decade,
10
legislatures
and
one
commission
passed
plans
that
were
not
challenged
in
court.
None
of
their
plans
were
challenged
in
court
throughout
the
decade
there
were
an
additional
nine
legislatures
and
six
commissions
that
had
some
of
their
plans
not
challenged,
whereas
some
were
anyway,
if
you're,
lucky
or
if
maybe,
if
you're
good
your
plan,
won't
even
be
challenged.
C
That
summary
information
that
you
saw
on
the
success
rates
chart
is
duplicated
down
here
at
the
bottom
of
the
of
the
actions
chart.
Now
many
of
you
know
a
lot
more
about
what
happened
in
your
states
than
I
do,
and
if
you
see
errors
in
this
chart,
something
that
doesn't
seem
right,
please
let
me
know
I'll
fix
it
and
we'll
all
be
up
to
date.
C
How
do
we
know
what
happened?
Okay,
so
we
have
the
marks
about
the
the
litigation,
but
where
does
that
come
from?
Where
does
that
information
come
from?
Well,
it
comes
from
redistricting
case
summaries,
and
this
is
the
ncsl
page
for
the
current
decade
summarizing
the
cases
there
are
similar
pages
that
go
back
to
1980.
A
Peter
we
we
have
a
question
for
you:
okay,.
A
C
No,
I
had
made
no
distinction
between
independent
and
dependent
commissions,
but
these
are
all
commissions
that
actually
drew
a
plan
they're,
not
backups
or
ones
that
weren't
needed
in
the
decade.
It's
the
ones
that
actually
drew
a
plan.
A
But
for
our
purposes
and
the
purposes
of
the
information
that
we
present
commission
is
defined
as
anything
that
is
not
the
legislature,
so
it's
a
broadest,
it's
probably
one
of
the
broadest
categorizations
that
you'll
see,
but
just
know
that
when
you
see
ncsl
talking
about
commissions
we're
talking
about
every
single
type,
not
just
the
ones
that
are
referred
to
as
independent
commissions
by
other
groups.
C
Peter,
so
what
you
see
on
this
page
is
the
current
decades
case
summaries
that
the
the
key
to
get
into
the
case,
summaries
and
you'll
see
the
callers
designate.
Whether
a
plan
was
drawn
by
a
legislature
or
commission,
the
legislature
in
purple
and
the
commission
in
yellow
and
iowa,
with
its
unique
system
of
this
nonpartisan
staff.
Drawing
a
plan,
that's
up
to
the
legislature
for
an
up
or
down
vote
is
in
orange.
C
When
we
take
a
look
at
the
congressional
map,
we
see
some
of
those
state
colors
change,
because
in
pennsylvania
and
ohio,
for
example,
and
several
others,
it's
not
a
commission
that
draws
the
plan
for
congress
for
any
of
these
states
for
your
state.
You
want
to
see
what
happened
on
either
legislative
or
congressional.
C
You
click
on
the
icon
and
I
will
click
on
north
carolina
because
it
is
the
most
complex,
more
tortured,
tortuous
process
for
a
redistricting
plan
in
the
country
this
last
decade
and
thus
its
summary
is
very
long.
C
The
summaries
include
links
to
the
various
court
decisions
that
were
involved,
sometimes
links
to
the
laws
that
were
passed,
correcting
a
plan,
there's
the
overall
summary
and
then
there's
the
case
by
case
north
carolina
had
more
than
one
case
that
went
up
to
the
u.s
supreme
court
like
cooper
versus
harris
here
and
all
the
links
and
what
ended
up
as
rucho
versus
common
cause,
which
was
up
there
twice.
C
So
there's
a
very
tortuous
path
that
their
congressional
path
plan
took,
including
after
the
supreme
court,
said
we're
not
going
to
consider
in
federal
court
claims.
The
partisan
gerrymander,
the
plaintiffs
filed
a
lawsuit
in
state
court,
challenging
the
congressional
plan.
The
complaint
and
the
various
other
important
documents
in
that
case
are
shown
here
as
well.
C
But,
finally,
by
december
of
2019,
the
court
approved
the
plan
and
it
went
into
effect,
and
you
have
all
seen
multiple
versions
of
the
north
carolina
map,
which
have
districts
that
look
really
weird.
But
this
one
is
not
so
weird,
it's
not
so
weird.
C
Now,
how
do
states
become
successful?
You've
had
a
lot
of
talk
earlier
in
the
conference
about
districting
principles.
I
don't
won't
go
into
the
details,
but
just
noting
that
a
population,
equality
and
the
voting
rights
act
apply
nationwide,
but
there
are
a
whole
lot
of
other
principles
that
the
supreme
court
calls
traditional
districting
principles
that
apply
only
in
some
states
and
only
to
some
plans.
These
principles
are
traditional,
but
not
universal.
C
Continuity
applies
in
50
states,
but
there
are
still
17
states
that
don't
require
it
for
their
congressional
plans.
For
example,
take
a
look
at
your
state
and
see
what
are
the
principles
that
apply
to
you
in
this
table?
You
can
click
on
one
of
these
links
and
with
luck,
we'll
go
to
the
text
of
the
language
and
write
to
the
word
contiguous
and
part
of
the
sentence
that
districts
be
composed
of
contiguous
and
reasonably
compact
geography
scroll
up
a
bit
and
you'll
see
the
document
that
that
is
a
part
of
it's,
not
the
state
constitution.
C
It
is
a
resolution
guidelines
passed
by
the
permanent
legislative
committee
on
abortion
in
may
of
2011.,
presumably
to
be
reviewed
in
in
may
of
2021,
or
whenever
we
get
some
population
counts.
So
you
want
to
know
the
detail.
Wanna
know,
what's
all
going
on,
that's
how
you
can
find
it,
but
you're
not
done
yet.
You
know
the
principles
that
apply
to
you,
but
what
do
they
mean?
How
do
they
work
well
to
help
you
all
out?
C
But
in
addition
to
that
handy
little
map,
this
includes
an
introduction
to
redistricting
and
an
explanation
of
why
we
do
it.
How
it's
done
explanation
of
what
gerrymandering
is
what
it
looks
like
the
equal
population
requirements
not
discriminating
against
racial
or
language
minorities
and
not
going
overboard
with
partisan
gerrymandering.
C
A
We're
gonna
run
about
five
minutes
over
which
would
take
us
to
150
mountain
350
eastern.
So
you
got
about
eight
minutes.
C
Okay,
well,
I
was
only
looking
for
three,
but
that's.
Okay,
that's
good!
That's
good!
We've
learned
that
the
the
federal
courts
say
that
partisan
gerrymandering
claims
are
no
longer
dissestable
in
federal
court.
So
all
those
plans
that
have
been
struck
down
earlier
in
the
decade
were
reinstated,
but
that's
not
true
of
state
courts.
C
So
if
one
wanted
to
look
for
a
basis
on
which
to
challenge
a
plan
as
a
partisan
gerrymander,
one
could
look
to
those
clauses.
There
was
a
law
review
article
done
2014.
That
pointed
all
these
out,
and
next
version
of
this
will
have
a
footnote
listing
those
26
states,
including
pennsylvania,
that
has
that
requirement.
A
Great
peter,
we
have
a,
we
do
have
another
question
for
you:
okay,
alexa
has
a
question
about
your
methodology,
so
she
wants
to
know
what
came
first,
your
definition
of
success,
or
was
it
reviewing
the
case?
Studies
like
how
where,
where
in
the
the
process
of
coming
up
with
your
methodology,
did
you
come
up
with
your
definition
of
success
and
are
you
planning
on
making
any
more
reports
on
success
in
the
future?
A
C
Many
many
questions
in
there.
Let
me
go
back
to
this
history
question.
How
did
this
start?
It
started.
I
think
it
was
about
1990
when
a
young
man
named
tim
storey,
was
in
charge
of
ncsl's
redistricting
effort
he's
sent,
as
we
all
know,
gone
on
to
much
better
things,
but
he
I
think
it
was
tim
who
had
the
idea.
C
How
are
we
doing
comparing
legislatures
to
commissions
with
their
ability
to
get
their
plans
enacted,
and
so
I
I
looked
at
the
old
cases
and
created
the
tables
that
you
saw
about
actions
and
made
the
computations
based
on
tim's
desire
to
know
how
legislatures
compared
to
commissions
on
getting
their
plans
all
the
way
through
the
decade.
C
These
these
tables
of
actions
may
need
to
be
revised
as
we
learn
more
and
I
know
at
presentation,
maybe
it
was
in
providence
back
in
september
of
2018
whenever
it
was.
Somebody
said
well
what
about
california
in
1990?
C
Well,
I
didn't
know
it
was
a
case
that
was
open
and
pending.
So
I
had
to
do
research
on
that
and
try
to
find
out
the
answer
turned
out.
Nobody
knew
the
answer.
What
what
ultimately
happened
to
it?
I'm
just
saying
I'm
I'm
open
to
doing
research
where
there
are
errors
that
are
discovered
or
questions
concerns.
Let
me
know
and
I'll
track
them
down,
but
part
of
that
question
is
peter.
When
are
you
really
going
to
retire
because
I
didn't
think
I
was
going
to
do
this
10
years
ago
at
this
time?
C
In
fact,
I
sort
of
gave
a
retirement
speech
at
the
january
seminar,
but
then
you
you
contacted
me
and
and
wendy,
and
here
I'm
down
enjoying
the
fine
life
in
panama
and
you
say:
well,
we
want
to
put
together
some
summaries
of
cases
and
see
what
happened
well.
A
Well
peter,
thank
you
yeah
so
is
there's
another
question:
is
there
a
link
to
the
site
peter
had
up
earlier?
Yes,
it
should
be
in
the
resource,
so
there's
a
resource
below
the
window
that
you
all
see
that
has
it's
called
installed
by
peter
watson.
Click
on
that
it
should
have
links
to
everything
that
he's
been
talking
about
and
then,
if
you
wanted
to
find
it
a
different
way,
you
can
also
just
everything
that
he's
pulling
from
is
on
ncsl.org.
A
If
you
go
into
the
redistricting
area
of
the
webpage,
you
should
find
the
links
to
all
of
them,
peter
any
last-minute
thoughts.
C
Well,
just
about
links,
I've
had
conversations
with
you
and
christy
and
others
about
the
links
in
these
documents.
Often
when
my
documents,
my
pdfs,
get
posted
on
the
ncsl
website,
the
magical
security
of
nc
itself
turns
off
the
links,
so
they
don't
work
and
we're
kind
of
working
on
that.
I
believe
not
to
try
to
correct
that,
but
we
may
be.
We
may
be
sending
emailing
some
of
these
documents.
Certainly
anybody
who
requests
them,
but
maybe
to
everybody,
sending
out
them
as
an
email.
A
That's
absolutely
the
case
and
if
anyone
has
any
issues
accessing
the
links
in
the
future,
email
me
at
the
general
ncsl
elections
and
redistricting
inbox,
which
is
elections
dash
info
ncsl.org
and
we'll
be
happy
to
handle
any
of
your
questions
from
there
with
that
peter
round
of
applause,
for
you
as
well,
the
virtual
round
of
applause
throughout
the
room
and
oh
thank
you.
Moon
is
here
as
well
and
sam,
so
everyone.
I
think
this
has
been
really
fun.
A
We've
wanted
to
try
something
different
with
this
hour
and,
if
you
all
like
it,
when
you
get
the
feedback
at
the
the
end
of
the
seminar,
let
us
know
make
sure
you
fill
that
out
and
tell
us
what
you
thought
and
if
you
want
to
see
this
in
the
future,
if
you
want
to
see
it
in
the
future,
but
you
want
something
to
be
structured
a
little
bit
differently.
A
You
know
with
the
pandemic
and
everything
going
on,
we
can't
be
in
person.
We
decided
to
just
try
some
new
stuff
and
see
if
you
all
liked
it,
so
I
hope
you
did
and
we're
so
grateful
to
all
of
our
speakers
who
came
here
today.
So
thank
you,
sam.
Thank
you,
moon.
Thank
you,
peter.
A
Some
final
notes
now
coming
up.
We
have
the
republican
and
democratic
breakouts.
Those
are
not
sessions
that
are
run
by
ncsl
they're,
run
by
the
ndrc,
the
national
democratic
registry
committee
and
the
nrt,
the
national
republican
redistricting
trust.
They
should
have
reached
out
to
you
with
information
on
how
to
log
in
if
you
registered
and
self-identified
as
a
d
or
an
r
or
they
could
have
reached
out
to
you
through
other
means,
if
you
are
a
d
or
an
r
and
you
want
to
participate
in
those
sessions
and
you
haven't
received
an
invitation.
A
Yet
the
contact
info
for
them
is
on
the
agenda
page.
So
you
can
go
to
that
through
the
platform.
I
would
also
be
remiss
if
I
did
not
mention
that
we
have
trivia
tonight
at
eight
o'clock.
Eastern
tim
story,
who
peter
mentioned
a
couple
minutes
ago,
is
very
excited
for
trivia
and
wants
to
see
all
of
you
there.
He
misses
these
in-person
interactions,
and
so,
if
not
for
me
for
tim,
please
go
over
to
trivia
tonight.
A
Just
so
you
can
say
hi
to
him
and
it
should
be
a
lot
of
fun.
We
have
a
comedian
who
and
comedian
slash
professor
who
is
going
to
be
doing
the
trivia
for
us
he's
at
middle
tennessee,
state
university.
So
with
that
everyone.
Thank
you
so
much.
This
has
been
a
lot
of
fun
and
I
look
forward
to
seeing
you
all
at
trivia
tonight
and
then
in
the
sessions
tomorrow.