►
Description
In this episode of the Whiteboard Series, our co-founder Illia Polosukhin (https://twitter.com/ilblackdragon) speaks with Alex Chapman (https://www.linkedin.com/in/iheartsemicolons/) who's a Tech Lead at Augur (https://twitter.com/AugurProject)
Learn about Augur: https://www.augur.net
Location: Nomadworks, Midtown, New York, NY | Consensus Week 2019
Download the video here: https://near.ai/61abd
------- About Whiteboard Series ---------
The Whiteboard Series with NEAR is a set of video interviews with Founders who are building scalable solutions for Blockchain. We deep dive into their story, what they're developing and how it is helping the ecosystem.
Follow the latest from NEAR Protocol on,
Discord: https://near.ai/discord
Medium: https://near.ai/medium
Twitter: https://near.ai/twitter
GitHub: https://near.ai/github
#WhiteboardSeries #FutureIsNEAR #Consensus2019
B
Inauguration
myself,
Alex
Kevin
lead
Evan
auger
at
the
forecast
foundation,
we're
a
group
of
software
developers
right
now,
working
on
a
prediction
market
that
runs
on
aetherium,
so
I
guess
high-level
overview
of
how
augur
works.
Yes,
yeah
sure
and
I'm
not
great
at
drawing
too
much
in
a
whiteboard,
but
I
can
kind
of
give
like
an
idea
what
the
pieces
are
just
a
really
high
level
view.
I
actually
never
spell
augur
like.
B
B
So
trading
component
of
augur
kind
of
you'd
expect
in
a
prediction.
Market
people
can
make
markets
by
themselves.
You
know,
but
we
don't
actually
make
any
markets
totally.
This
head
applies
people
just
come
in
and
make
their
own.
You
can
make
a
market
on
something
like
political,
like
who's
going
to
in
a
particular
election
sports
games
or
something
more
sophisticated.
Like
some
sort
of
a
specific
business
hedge
you
can
make
with
scalar
markets,
which
is
something
we
have
prediction
markets
to
track
like
a
particular
stock
price.
So
there's
a
lot
of
different
functionality.
C
C
B
Can
make
markets
and
exchanges
all
the
time
using
smart
contracts?
What
actually
sets
order
apart
is
the
fact
that
we
have
a
decentralized
oracle,
so
decentralized
oracle
means
when
someone
makes
one
of
those
markets
like
an
election
market,
for
example,
and
it
passes
the
deadline
where
the
election
results
have
come
in.
The
block
team
needs
to
know
and
the
contract
needs
to
know
how
to
pay
things
out.
So
people
who
bet
on
the
winner
need
to
get
that
the
money
that
they
deserve
and
the
people
who
didn't
bet
on
the
winner.
B
They
don't
get
to
get
anything
a
lot
of
the
ways
that
people
solve
this
without
using
specialized
Oracle,
typically
revolve
around
just
a
trusted
party
going
in
and
punching
in
who
won
and
then
the
results
come
out.
The
problem
with
that
is
having
a
centralized
party
in
prediction.
Market
has
a
lot
of
risks
associated
with
it
pretty
much.
Every
prediction
market
in
history
has
been
shut
down,
attacked
or
has
ended
up
being
some
sort
of
like
an
exit,
scam
situation,
so
the
decentralized
oracle
is
really
the
super-complicated
t,
curse
or
a
part
of
augur.
D
B
I
said
I'm
bad
at
white
board
by
the
way
so
I'm
just
switched
to
talking.
I,
don't
know
your
guy
anything,
yet
we
have
a
token
called
rep
when
a
market
is
supposed
to
end.
The
market
creator
had
designated
when
they
created
the
market.
Someone
to
report
so
someone
to
go
in
and
do
that
manual
process.
I
subscribed
where
they
go
in
themselves
and
say
yeah
this
guy
won.
B
The
difference
is
that,
once
those
results
are
up,
there's
a
time
period
where
anybody
who
owns
rep
can
go
in
and
say
that's
not
true
and
there's
a
preceding
sequence
of
disputes.
That
happened
where
the
amount
that
they
had
to
put
up
gets
higher
and
higher
and
higher,
which,
if
that
was
the
only
system
we
had,
which
is
the
way
a
lot
of
other
decentralized
Oracle's
work
would
be
a
Keynesian
beauty
contest,
which
just
means
if
someone.
B
And
whoever
ends
up
having
the
most
just
wins
and
as
a
rational
game
theory
participant,
you
should
just
go
with.
Whoever
has
the
most
rep
yeah,
which
means
we
wouldn't
actually
be
a
good
decentralized
oracle.
It'd
just
be
a
Keynesian
beauty
contests
system.
So
the
interesting
thing
about
our
descendant
Oracle
is
at
a
certain
point.
We
actually
fork
auger
within
the
ethereum
blockchain,
so
not
like
an
aetherium
fork
like
an
actual
system
where
our
contracts
are
capable
of
forking
into
two
different
auger
systems.
C
B
B
Maybe
Bernie
I
don't
know,
and
actually
there
could
be
all
three,
because
any
one
can
make
it
was
a
categorical
market,
for
example.
Anyone
can
can
say
I
would
like
for
that
child
universe
to
become
its
own
auger
once
the
pork
occurs.
The
intern
with
that
is
rep
holders
have
to
choose
where
the
RIP
goes
right
right.
B
They
have
to
burn
a
lot
of
rep
to
do
that
and
that's
kind
of
the
the
crux
of
our
whole
security
model:
the
fact
that
they
they
have
to
make
their
rep
go
to
a
universe
where
it
is
probably
worth
nothing.
So
all
the
security
calculations
we
do.
Everything
in
the
white
paper
is
basically
based
on
this
notion
of
them
having
to
burn
more
in
rep
than
the
markets.
Open
interest
is
so
even
if
they
got
all
of
the
money
from
the
market.
They've
been
overall
lost
money
by
performing
that
attack.
B
A
B
D
B
B
B
B
The
categorical
is
it's
that
the
number
of
shares
that
we
have
to
generate
and
when
you're
trading,
the
number
of
shares
we
have
to
send
to
people,
including,
like
emitting
logs
for
all
those
transfers
and
just
all
the
associated
costs,
with
having
that
many
shares,
makes
anything
beyond
eight
kind
of
get
a
little
wild
I
think
it
was.
There
was
actually
the
cost
of
just
creating
a
market
was
too
high.
Originally,
we
could
actually
probably
increase
that,
but
there
would
probably
be
negative
side
effects.
B
A
B
Multiple,
you
can
go
long,
multiple,
which
means
make
multiple
trades.
The
key
thing,
though,
is
if
you
short
so
the
way
that
other
thing
works
in
the
underlying
contract
system
is
all
of
our
sorts
or
synthetic
shorts.
So
it's
easy
to
understand
with
yes,
no
there's
a
yes
share
in
a
no
share.
If
you
short,
yes,
if
you
say,
sell
that
you'll
get
no
shares
yeah,
it
makes
more
sense
categorical.
B
If
there's
eight
candidates-
and
you
say
I-
want
a
short
Trump
yeah-
that
actually
what
the
system
has
to
do
is
give
you
shares
of
all
the
other
candidates.
So,
if
sort,
what
does
get
into
with
the
synthetic
sort
problem
with
reversing
positions
is,
if
you're
reversing
that
means
it
has
to
Train
for
eight
chairs
per
person
twice
and
has
to
admit
logs
for
all
those
and
do
all
of
the
accounting
for
all
those
individual
chairs.
So
what
is
the
problem
happening?
Is
a
single
trade
if
they're
both
reversing
positions,
costs
like
3.5
million
gasps.
C
B
A
C
B
Anybody
else,
so
we
could,
we
could
have
that,
but
it's
one
of
those
things
where
I
suspect
people
using
it
would
be
like
what
this
is
broken.
What
is
going
on
this
can't
be
right.
Yeah
categorical
is
a
data
type.
You
can
have
eight
outcomes
max
on
those
and
in
scalars.
Scalars
are
not
used
very
much
in
v1,
I
hope
they're
used
more
in
v2,
because
these
are
things
are
the
neatest
type
of
markets,
so
scalar
markets,
it's
like
a
margin,
Coles
kind.
A
B
You
basically
tracked
that
price.
So
the
way
we
have
everything
set
up
in
the
UI
and
it
works
differently
underlying
on
the
chain,
but
everything
in
the
UI,
for
example,
if
you
thought
it
was
gonna,
be
150,
you
could
buy,
buy
it
anywhere
on
a
hundred
fifty
you
would
actually
pay
that
amount.
So
if
you
someone's
going
to
sell
it
a
hundred
forty-five,
you
could
buy
a
hundred
forty
five.
If,
at
the
end
of
market
resolution,
identity
150,
you
would
make
five
dollars
and
that's
how
the
payouts
actually
works.
B
The
payoff
system
for
everything
is
actually
actually
works
as
a
scalar
market.
Interestingly
enough
yeah
it's
and
we
could
go
through
all
that,
all
that
how
that
works
on
contract
but
yeah
the
the
scalar
payout
there's
actually
two
shares
along
in
a
short
and
there
payout
is
kind
of
on
a
range
and
that's
how
that
helps.
B
It
isn't
those
are
the
three
classifications
and
market
and
again
on
chain.
We
have
three
different
creation
functions,
but
they're
actually
all
the
same
thing
and
they
all
work.
The
exact
same
all
the
math
is
the
exact
same
on
all
of
them.
Oh
yeah
I
mean
this
is
just
it
right,
it's
much
simpler,
so
you
can
imagine
I'll
actually
go
into
it.
So
there's
like
a
thing
called
payout.
B
We
call
them
payout
numerators,
which
is
a
misnomer
right
now,
actually
in
the
contracts
it
has
to
do
with
how
we
payout
invalid
markets,
but
in
yes-no
you
can
measure
in
there's,
there's
a
I
think
the
first
one
of
the
yes
share
and
the
second
one's
a
no
share.
Don't
quote
me
on
that
I
always
forget
which
one
it
is,
but
so
there's.
B
B
There's
this
concept
in
the
contracts
called
num
ticks,
which
everyone
kind
of
dislikes
cuz
it's
complicated,
but
we
have
not
thought
of
a
better
way
to
conceptualize
and
program
this
notion,
but
the
basic
idea
is:
when
people
are
trading
and
we
will
get
chairs
paid
out,
there
has
to
be
some
notion
of
how
many
ticks
they
trade
in
seek.
Imagine
if
you
were
buying
something
like
on
an
exchange,
and
you
want
to
set
a
price
for
a
share.
B
You
might
think
it's
worth
eighty
nine
point
five
cents,
but
and
that
wouldn't
be
possible
if
that
exchange,
supported,
1000
ticks,
I,
think
where's
the
hundred
texts
yeah
1000
days
because
they're
busily
when
a
thousand
units
have
increased
you
could
have.
You
grew
from
eighty
nine
point,
five
to
eighty
nine
point
six
day
and
point
seven.
If
they
didn't
want
to
deal
with
fractions
of
a
cent,
you'd
only
have
a
hundred
ticks.
You
can
only
you
only
deal
in
that
many
ticks,
so
our
our
underlying
system
for
yes/no
in
categoricals
has
10,000
ticks
right.
C
B
Yeah,
so
the
issue
is
this
is
the
payout
is
also
based
on
those
gnomes,
and
that
makes
it
work
out
really
really
nicely
with
making
sure
that
all
of
the
the
funds
are
actually
like
covered,
and
we
don't
have
to
do
any
weird
rounding
or
division
or
anything
like
that.
We
just
limit
what
you
could
have
actually
traded
and
what
you
get
paid
out.
Scalars
are
more
complicated
scalars.
We
let
people
do
custom
amounts
because
ten
thousand
takes
does
not
always
apply
to
every
scalar
market.
B
You
could
have
something
very
complicated
if
you,
what
your
trading
Berkshire
Hathaway
shares
or
something
you
know
like
it,
doesn't
yeah,
it
has
to
be
custom.
Categoricals
worked
the
same
way,
but
obviously
there's
more
there's
like
for
each
category
yeah
right.
So
this
is
Trump
Hillary.
It's
Bernie.
A
B
A
C
D
B
Can
imagine
if
rap
is
essentially
worthless,
they
don't
care
if
they
have
to
burn
it.
If
there
was
a
lot
of
oh
I,
for
example
in
in
the
Democrat,
which
is
actually
possible
so
like
as
a
very
specific
example,
if
the
if
the
open
interest
and
that
got
up
to
let's
say
three
million
and
the
market
cap
of
v1
rep
was
less
than
three
million
yeah,
then
just
by
all
the
rep
and
resolve
the
market.
B
B
B
B
B
Jerk
edges,
Trump
and
yang
I'm
thinking
dude.
He
said
he's
come
to
the
conference
he's
talking
on.
Was
it
I
got
a
I
got
a
Johnny
hanging,
so
so
the
total
Alliance
market
is
six
million
dollars.
So
say
you
and
I
off
chain.
Just
I
have
a
bag
of
cash
say
I'm
I'll
give
you
I'll
bet
you
a
billion
dollars.
The
yang
wins
problem
is
there's
nothing.
The
chain
can
do
about
that.
There's
nothing
our
our
security
can
do
about
that.
B
We
can
never
know
that
so
there's
no
way
that
any
systems
we
create
or
any
mechanism
that
we
create
could
account
for
that.
Basically,
the
best
we
can
do
is
and
it's
what
the
white
paper
does
is.
It
says
we
don't
think
they'll
ever
be
more
than
this
much
parasitic,
oh
I,
based
on
the
oh
I
within
our
system
and
that's
kind
of
that's
the
best
we
can
do
so
so
reading
the
white
paper
and
with
that
amount,
I
believe
the
amount
of
we
we
limit.
C
D
B
A
B
And
agreeing
to
use
our
Oracle
for
that
winners.
Other
options
too
yeah,
like
I,
said
it
kind
of
gets
into
a
thing
where
it's
purely
pedantic
game
theory
is
I.
Believe
it's
entirely
possible.
There
could
be
a
lot
more
personally
and
things
would
not
actually
break
again
like
that.
There's
more
reasoning
on
that:
what
if
they
have
other
markets
and
they
want
those
to
resolve
correctly
yeah-
and
you
know,
presumably
showing
that
the
Oracle
resultant
badly
would
have
other
negative
side
effects
for
them
and
potentially
their
other
markets.
Yeah.
B
B
B
But
v1
v2
and
Marcus
there's
not
a
lot
of
competition,
not
a
lot
of
high
quality
market
craters,
so
people
can
charge
higher
fees
and
people
seem
to
be
okay
with
that,
don't
mean
for
trading
on
things
that
have
you,
but
even
as
high
as
like
four
percent
fees,
which
I
think
is
probably
way
too
high,
but
but
that
just
goes
to
the
market
painter.
So
when
two
people,
the
way
that
we
do
fee
fee
settlement
is
whenever
a
trade
would
make
people
sell
both
their
shares.
B
So
if
I
have
gang
chairs-
and
you
have
Trump
chairs
and
we
both
want
to
close
in
our
positions
and
we
get
matched
in
a
trade,
it'll
sell
that
back
into
the
system
and
extract
these
I
see.
So
we
take
these
in
that
case
and
we
take
fees
in
the
case
when
the
market
is
over
and
you're
just
getting
your
payout
from
the
system.
Those
are
the
two
case.
If
you
take
these,
if
you're
just
trading
back
and
forth
shares
in
tokens,
we
actually
don't
they're.
B
D
B
A
A
B
And
the
way
that's
implemented
because
of
smart
contracts,
is
you
just
specify
an
address
yeah
so
really
savvy
people
are
very
quickly
going
to
learn
that
they
can
just
specify
their
own
address
and
yeah,
but
you
know
for
a
lot
of
people.
That
is
the
case.
I
mean,
arguably,
you
can
sign
up
for
like
the
Amazon
affiliate
program
thing,
but
nobody
does
so
in
practice.
It'll
be
actually
use
that
way.
Yeah
the
other
fee
is
the
reporting
fee.
B
B
B
It's
kind
of
central
to
secure
you
the
system,
but
it
is
a
function
of
the
price
of
rep
and
the
öin
the
system,
and
this
is
how
we
make
secure
that
whole
system,
where
we
try
to
guarantee
assuming
a
certain
amount
of
peracetic
interest,
that
if
someone
does
incorrectly
resolve
a
market,
they
will
overall
burn
money
and
we
do
that
by
adjusting
reporting
fee.
So,
if
you
imagine,
the
open
interest
on
the
platform
is
see,
I'll,
try
and
write
this
out.
So
if
the
open
interest,
the
platform
is
10
million.
B
B
Sort
of
reputation
base,
like
you
would
by
reporting
tokens
for
the
right
to
report
on
markets.
Maybe
we
get
assigned
to
you
and
there
were
a
lot
of
iterations
where
kind
of
the
game
theory
folks
were
progressively
more
and
more
cutting
everything.
I
think
this
is
not
necessary.
This
is
not
necessary,
so
it's
sort
of
it's
less
reputation,
took
it
reputation
based
now,
I'm
sure,
there's.
B
B
Right
so,
let's
actually
make
that
a
hundred
nine
if
ret
market
cap
was
1
million,
then
obviously
something's
wrong
with
the
security
model
right
now,
rep
is
not
worth
enough
to
secure
the
system
and,
if
there's
actually
there's
a
multiplier
that
we
have
I
think
for
v2
we're
changing
it
to
the
rep
market.
Cap
should
be
sorry,
I
do
I
should
be
five
times,
I
believe
the
rep
market
cap
that
might
be
adjusted,
but
we
have
a
formula
basically
for
for
adjusting
that
and
what
happens
is
when
the
rep
market
cap
is
too
low.
B
A
So
I
order
your
feet.
So
look
just
for
my
understand
for
my
visual
understanding,
so
you
have
let's
say
market
so
as
part
of
it
there's
some
amount
of
eath
locked
it
locked
up,
and
this
issued
some
yes
tokens
and
no
tokens.
We
can
freely
trade
this.
It's
like
a
20
you
can
before
this
finished.
You
can
also
just
like
pretty
much
burn
it
and
take
out
ease.
Thank
you,
like
both.
Both
people,
yeah
sure
show.
Yes.
A
B
A
B
A
B
So
the
prices
that
they
pay
for
their
shares
are
not
actually
something
the
system
ever
takes
into
account,
so
whatever
they
paid
for
them,
it's
it's
something
that
MV
one.
We
don't
actually
record.
We
we,
we
have
a
middleware
component
that
reads
logs
and
tries
to
kind
of
show
them
what
they
paid
for
it
and
in
v2
we
actually
do
we
record
that
on
chain,
not
for
purposes
of
payouts,
but
just
purposes
of
showing
them
what
their
profit
losses
and
everything,
but
the
way
the
system
works.
B
B
The
other
position
is
no,
but
once
they
pay
that
money
to
the
system
think
about
it,
they've
they've
bought
purchased
a
complete
set,
basically
of
yes
and
no
so
they
bought
all
the
outcomes
that
exist
in
that
market
together
then
they
must
pay
whatever
a
complete
set
is
worth,
but
it's
irrelevant
what
each
of
them
paid
it
all
the
matters
that
the
system
got
enough
to
pay
out
enough.
Basically,
this
amount.
B
So,
interestingly,
the
amount
of
complete
set
costs
is
always
the
number
of
ticks
by
the
way
to
which
again,
one
of
those
nice
properties
that
works
on
the
chain.
So
together
they
came.
Let's
say
they
just
bought
one
share
of
each,
so
they
collectively
had
to
pay
ten
thousand,
but
the
person
about
the
yes
may
have
only
paid
one,
the
person
who
bought
no,
maybe
one
who
paid
$9.99
or
whatever,
but
then
when
they
cash
out.
B
B
Well,
so
the
price
is
defined
by
whatever
people
are
trading,
so
there's
actually
an
order
book
on
on
chain
and
people
going
in,
say
what
they're
willing
to
trade
for
so
just
like
any
other
exchange.
It's
it's
not
actual
something
we're
like
it's
not
like
a
actual
betting
market,
where
somebody
defines
the
odds
or
anything
like
that.
It's
just
a
open.
B
C
A
See
so
pretty
much
yeah
take
out
the
yes.
If
I
want
to
take
no
position,
I
take
out
the
yes
share,
I
go
to
some
exchange,
I
sell
it.
Somebody
buys
it
right
and
then
that
defines
the
odds
right
like
whatever
position
I'd
like
sell
it
here,
that's
kind
of
odds
and,
frankly,
to
take
yeah
and
then,
if,
for
example,
it
doesn't
close,
I
can
go
close
this
up
and
say:
hey.
They
work
great
okay,
cool.
A
A
B
B
B
Whether
Trump
were
you
anyone
I
see
so
the
D
are
currently
in
v1
has
three
days
to
do
that
in
v2,
they're
gonna
have
one
day
because
we
found
that
if
they
do
show
up
at
all,
they
show
up
within
one
day
if
they
they
don't
show
up
within
one
day,
they
don't
show
up
at
all.
So
so
there's
a
few
actually
interesting
if
they
like,
if
they
lost
their
key
to
I,
don't
think
in
v1
there's
a
way
to
transfer
that,
so
they
would
the
shadow
look.
B
B
In
this
case,
the
dr
shows-
and
they
say
yang
and
then
in
the
case,
where
there's
no
show
so
the
three
a
limited
past
or
in
be
to
the
one
day
limit.
It
goes
into
something
called
open
reporting,
so
are
an
open
reporting,
the
no
Shobana,
which
is
something
that
the
market
crater
has
to
put
up
at
market
creation.
Time,
which
is
rap,
gets
used
for
the
open
report.
B
You
exactly
and
the
open
reporter
there
is
Li
entitled
to
it
once
the
market
finalizes
presuming
they
they
report
it
correctly,
so
the
initial
bond
goes
up
and
up
and
up
until
it
reaches
kind
of
equilibrium
that
that's
the
idea
of
it
and
that's
based
on
that
something.
The
system
records
the
system
records,
how
much
people
don't
show
up
and
the
more
people
don't
show
up
relative.
B
A
B
It's
not
market
specifics
for
everyone,
but
yeah,
and
it's
sort
of
a
mechanism
just
to
adjust
the
the
market
cap
that
the
the
no-show
is
very
specifically
just
people
aren't
showing,
which
is
making
markets
resolve
slower,
have
me
adjust
that
a
little
bit,
so
there
are
no
show
the
o.r
gets
report.
Let's
say
the
o.r
reports
for
trunk
I,
don't
know,
let's
say
in
this
in
this
world,
yang
wins
so
so,
regardless
of
whether
the
O
R
or
the
D
are
reported
at
this
point,
things
are
gonna,
be
the
same.
B
B
Just
to
be
technically
accurate,
what
ends
up
happening
is
when
the
dr
does
show
up
the
market
here
gets
the
no
Chabon
back
and
then
the
designate
reporter
has
to
put
in
a
new
bond,
but
in
practice
that's
just
them.
Putting
the
notion
bond
and
just
like
what
happens
with
the
the
open
border
I
see.
So
at
this
point
yang
one.
So
a
few
things
can
happen.
The
market
can
resolve
with
this
outcome
for
that
to
occur.
B
This
has
to
go
at
least
7
days
with
nobody,
putting
up
a
Rhett
bond
to
challenge
this
and
say
the
Trump
one.
It's
actually
a
little
longer
than
seven
days,
there's
a
cadence
of
Windows
that
go
by
called
in
v1
fee
windows
and
in
v2,
they're
called
the
spiit
windows
and
there's
run
a
seven-day
cadence.
So
it's
possible
that
if
this
occurred
at
the
beginning,
it
has
to
go
a
full
one
of
those
windows.
B
So
if
this
was
reported
at
the
beginning
of
a
previous
window,
it
could
be
almost
14
days
for
it
to
actually
resolve
and
there's
ways
in
v2
that
that's
going
to
give
people
opportunity
to
dispute
it
for
it
to
give
people
who
have
rep
and
who
are
looking
at
markets
time
to
actually
find
it.
If
the
the
window
is
too
short,
we
risk
I,
guess.
B
B
I
go
every
Wednesday.
Wednesday
is
my
my
day
to
look
at
markets
and
pick
out
what
I
want
to
I
want
to
do.
It's
gonna
be
work
a
little
different
of
you
to
there's
still
a
cadence
like
that,
but
there's
other
mechanisms
we
have
to
speed
up
reporting
because
you
know
time
is
adding
up
here.
We
have
three
days
and
up
to
14
days
and
additional
other
time.
It
all
makes
for
really
slow
resolution
time.
B
B
So
so
now
the
market
is
finalized.
Anyone
can
there
is
a
actual
contract
of
claim
trading
proceeds.
You
can
pass
it
in
a
market.
It'll,
look
and
see
what
shares
you
have
see,
what
to
pay
you
for
them,
destroy
the
shares
and
give
you
the
money
from
the
market
and
then
pay
feeds
for
you.
The
other
thing
that
could
happen,
though,
let's
say
somebody
just
beats
this.
B
And
they
say
no
yang
one
when
you
dispute,
what
you
have
to
do
is
put
up,
there's
actually
a
formula
for
how
much
rep
it's
for
purposes
of
this.
We
can
just
say
it's
twice
the
previous
bond.
It's
actually
different
than
that
and
it
especially
when
it
goes
back
and
forth.
It
ends
up
being
a
different
amount
than
that.
But
just
for
simple
to
guess
it's
twice
the
amount
of
rep
that
they
put
up
so.
B
It
says
so
the
are
acted
the
o.r
or
they
are
rather
it's
always
gonna,
be
the
envy
one.
It's
a
little
complicated,
it's
the
initial
report
size,
but
for
simplification
purposes
and
because
in
v2
this
is
how
it
work.
It's
just
the
notion
bond.
So
if
the
no
Shabana
was
one
rep,
then
the
amount
staked
on
Trump
right
now
is
one
so
to
dispute
that
you'd
have
to
put
up
to
rep
nicely.
Then
this
can
continue
and
continue
and
continue
if
it
doesn't
go
this
period,
nobody
disputing
it
until
so.
A
B
B
So
what
the
contracts
actually
do
is
when
the
market
is
finalized.
It
looks
at
all
of
the
losing
bonds,
all
the
ones
that
weren't
for
they
were
arguing
basically
for
the
outcome.
That
did
not
end
up
being
the
finalized
outcome
and
just
takes
that
rep
and
then
puts
it
into
the
bonds
for
the
for
the
winning
outcome.
B
So
when
someone
comes
in
and
redeems
their
bond
with
the
system
for
a
sample,
this
yang
bond
they'll
just
receive
that
rep,
so
it
liquidates
all
of
the
losing
rep
and
gives
it
to
all
of
the
winning
bond
actually
and
that
actually,
the
way
that
we
have
the
formula
set
up.
That
I
was
mentioning
which
it's
actually
more
complicated
in
again
and
then
just
two
times
the
previous
bond.
But
it
guarantees
a
50%
ROI.
B
Right,
if
you
think
you're
correct
it's,
actually
it's
extremely
profitable
to
all
right.
If
you
think
that's
the
truth,
really,
not
not
you
guys
are
like
yeah.
The
reporting
is
actually
it's.
It's
weirdly
nuanced.
Actually,
just
the
concept
of
what's
true
yeah
there's
some
interesting
markets.
We
should
talk
about
actually
but
yeah.
If
you
ended
up
disputing
before
you
wrap.
C
B
Yeah,
it
admits
very,
very
interesting
markets,
one
of
the
one
of
the
first
markets
and
actually
I
think
the
the
biggest
dispute
that
ever
occurred
I
think
it
got
up
to
either
some
around
like
80
thousand
dollars
with
a
rep
before
actually
a
little
finalizing
was
a
will.
The
weather
be
nice
for
the
Bastille
Day
Parade
tomorrow
and
I
was
that
was
a
description
and
there
was
a
link
to
a
weather
site
that
did
not
use
the
language
good
or
nice
yeah
it.
B
It
was
odd,
actually
seeing
a
good
disputed
so
much.
The
the
marking
crater
was
very
very
adamant
in
our
discord
that
it
was
a
valid
market
and
that
people
should
be
able
to
find
a
shelling
point
on
what
is
nice
or
good,
but
it
end
up
resolving
to
in
it
invalid,
which
is
arguably
what
it
should
have
resolved
to
invalid
is
interesting
to
know,
saying.
B
C
B
This
is
nice
in
some
ways
and
easy
to
understand
in
some
ways,
but
it
has
a
lot
of
really
negative.
Weird
side
effects
and
not
every
market
that
doesn't
even
pay
out
obviously
is
invalid.
There's
also
like
a
flag.
You
have
to
provide
that
says.
I
intend
for
this
to
be
invalid.
You
can
imagine
a
scalar
market
there's
actually
a
fairly
decent
chance.
It
will
end
up
just
being
paying.
B
B
The
beginning
and
be
one
no
right
which
is
actually
my
problems
and
it
you
can't
even
really
glean
it
easily
from
what
had
the
order
books
trading,
because
this
is
sort
of
weirdly
baked
into
the
price
of
things
like
if
I
pay
0.44.
Yes,
that
a
signal
to
the
market
is
invalid,
maybe
because
if
it
is
invalid,
then
I'll
actually
make
money
on
that
I
as
I'm,
paying
less
than
the
amount
that
I'm
going
to
get
from
payout.
But
it's
really
difficult
to
factor
that
in
based
on
this
payout.
So.
B
And
it's
it's
it's
tricky
and
it's
especially
tricky
because
invalid
markets
are
sort
of
a
conundrum
in
v1
these
people
enterprising
individuals.
Again
everything
is
decentralized,
so
people
are
gonna,
find
everything
eventually
and
figure
it
out.
I
figured
out
that
they
can
create
markets
with
really
tiny
subtleties
that
mean
that
they
should
resolve
to
invalid
surfers.
Who
might
make
a
market
like
what
will
the
price
of
a
theory
and
be
at
the
end
of
April,
15th
and.
C
B
B
A
B
If
you,
if
you
imagine,
reporters
kind
of
as
a
whole
whole
body,
if
they
don't
want
to
get
in
situations
where
they're
hurting
themselves,
just
because
of
when
people
report
it
or
something
like
they're,
very
intolerant,
for
example,
of
api's
that
go
down
or
provide
different
results
to
different
viewers,
there's
a
problem
with
ESPN,
because
we've
other
people
who
report
in
Europe
and
their
Google
report
in
the
United
States,
and
they
see
different
ESPN's
and
yeah
there's
they
do
not
display
in
a
way.
That
is
easy.
People
actually
discover
this.
B
This
is
one
of
the
weird
subtle
invalid
markets.
Was
they
referenced
ESPN?
But
this
is
kind
of
an
issue
for
v1
because,
like
you
said,
there's
no
there's
no
curation.
Obviously
the
thing
is
completely
decentralized.
We
actually
run
ball.
All
we
just
make
the
software.
The
best
thing
we
can
do
is
add
things
to
to
a
reference
UI,
basically
for
viewing
the
on
chain
data
that
can
help
people
there's
a
few
of
those
that
can
be
added
in
v1,
like
looking
at
the
spread
in
a
particular
market
like
is
it
oddly
gearing
towards
this?
B
Instead
invalid,
it
will
be
a
tradable
outcome
on
every
market,
regardless
of
whether
you
specify
that
or
not
so
you'll
always
be
able
to
trade
invalid.
The
the
neat
thing
about
that
is
you:
can
you
can
obviously
trade
invalid
with
that
I'm
going
to
do
weird
calculations,
but
also
it
gives
us
sort
of
a
it,
gives
everyone
an
insight
into
whether
the
market
is
invalid
or
not
yeah,
and
it's.
B
It's
a
the
cool
part,
you
so
sue,
so
you
could
use
the
invalid
outcome
to
hedge.
You
could
also
like
one
of
the
things
we've
discussed
is
building
filters
into
the
reference
UI
that
look
at
what
in
Val
is
trading
at
like
if
there
are
Ani
any
bids
for
invalid,
say
above
0.1
or
something
you
can
imagine
that
we
could
have
to
filter,
have
a
default
or
hey.
B
A
C
B
B
Maybe
meaningfully
yeah,
if
it's
trading
at
point,
nine,
like
that
and
there's
gonna,
be
some
formula.
Obviously
there
yeah
because,
like
we
don't
want,
like
a
single
dust
order,
four
point:
nine
to
hide
a
market,
but
you
know
if
there's
significant
volume,
whatever
that
is,
maybe
some
portion
of
the
overall
volume
on
the
market
trading
for
prices
above
a
certain
level,
then
yeah,
you
can
say
with
some
confidence.
Actually,
the
market
is
invalid.
That's
something
everyone's
looking
forward
to
a
lot
in
v2,
because
there
are
a
lot
of
very
silly
markets.
Yeah
env2.
B
Actually
another
really
interesting
thing
about
reporting
is
the
reporters.
You
know
they
they're.
Their
goal
was
to
essentially
to
report
truth
and
to
distinguish
these
invalid
markets
and
handle
things
like
that.
But
really,
at
the
end
of
the
day,
their
goal
is
to
serve
traders
really
to
make
traders
confident
that
markets
will
resolve
how
the
traders
expect
there's
actually
a
more
subtle,
weird
notion.
We
weren't
really
sure
how
was
going
to
play
out
in
practice,
but
there
was
market
that
actually
kind
of
exercised.
B
This
notion,
which
was
made
by
someone
who
famously
makes
a
lot
of
invalid
markets
to
I,
think
this
was
actually
an
accident,
but
it
was
on
the
the
2008
2018
house
election.
It
was
whether
Republicans
would
maintain
a
majority
and
the
market
ended,
I,
believe
December,
December,
18th
or
something
somewhere
in
the
middle
of
December.
The
interesting
was
even
though
the
the
elections
happened
before
that
they
don't
actually
take
a
majority
until
the
next
year
and
the
way
that
the
work
market
was
worded,
it
was
will
they
take.
B
A
majority
adds
a
point
of
time
right
and
it
ends
at
December,
19th.
So
technically,
Republicans
still
had
a
majority,
even
though
they
they
actually
lost
seats.
The
Democrats
and
the
Democrats
took
the
majority
from
the
elections.
Yeah
yeah,
but
not
really
right.
Right
and
traders
had
actually
put
I
think
1.5
million
Noi
into
this
market,
so
it
was
getting
heavy
traction
and
there
was
some
debate
about
how
it
would
resolve
actually
because
it
was
sort
of
the
spirit
of
the
market.
Obviously,
what
traders
had
intended
it
for
to
be?
B
It
should
resolve
that
Democrats
won
yeah,
but
if
it
was
sort
of
too
to
the
letter
of
the
market
in
the
description
and
the
end
time,
then
it
should
be
either
Republicans.
It
was
actually
even
some
legal
argument
should
be
invalid,
because
they're
saying
because
this
was
known
before
before
the
market
got
made,
this
shouldn't
actually
qualify
as
a
valid
prediction
market
and
should
be
made
invalid.
A
B
There's
I
mean
there's
plenty
of
other
valid
reasons.
Things
could
be
invalid,
yeah,
but
it
instantly
it
did
it
resolved
sort
of
where
people
had
hoped
it
would
for
traders
with
Democrats
I'm
so
resolved
in
the
spirit
of
the
election,
who
won
the
election,
which
was
interesting
to
see
and
nice
that
the
the
reporters
could
kind
of
find
a
way
to,
amongst
themselves,
figure
out
what
would
be
best
for
traders
and
for
the
future
of
the
platform,
which
is
what
they
kind
of
have
to
do.
Yeah.
B
Makes
this
the
Oracle
problem
so
interesting
because
it
can't
it's
not
something
that
can
really
easily
be
solved
with
like
presentable,
just
simple
programming.
It
may
be
a
machine
learning
and
extremely
fancy
AI
or
something.
But
a
lot
of
these
questions
are
sort
of
weirdly
human-like.
What
what
did
humans
intend
by
this
market?
Yeah
I
was
one
of
the
more
interesting
yeah.
B
So
so,
let's
say
that's
where
the
home
was
getting.
If
the
dispute
just
continues
and
continues
and
continues,
it'll
eventually
be
some
round.
Let's
say
it's
the
twentieth
round,
and
this
varies
depending
on
you
know
what
the
new
shabanam
was
originally
on,
how
people
dispute,
but
once
it
reaches
a
certain
threshold
it
will
cause
a
fork.
B
A
B
A
So
yeah
so
pretty
much
originally
there's
one
universe
and
then
you
like
fork
it
into
let's
say
because
of
this
need
for
it
into
a
new
one.
So
it's
just
a
clone
event
where
rap
holders
pretty
much
here.
It's
rep
holders
off
like
yang
the
pretty
much
got
everything
cuz
take
on
Trump
and
here's
a
Trump
holders
got
everything
from
Yang
sort.
B
Of
so
so,
what
actually
ends
up
happening
is
I
mean
this
other.
There
will
be
a
trump
and
yang
universe
in
the
in
the
yang
universe,
when
they,
when
all
the
people
who
staked
on
Yang,
move
their
rep
to
that
universe.
It
will
mint
right
for
them
in
that
universe,
as
if,
though,
they
had
won
the
dispute
yeah
and
similarly
in
the
front
peanut
verse,
it'll
mint
rep
for
them
as
though
they
had
won
this
view
in
those
respective
universes,
but
it
does
actually
take
it
from
the
other
universes.
B
Also
need
to
decide
so
the
markets
actually
don't
decide.
The
markets
just
go
to
the
winning
universe,
so
the
markets
are
forced
to
go
to
those
winning
universes,
which
is
again
why
this
security
calculation
considers
the
entire
open
interest
of
the
whole
system
instead
of
adjust
the
open
interest
of
like
the
particular
market.
B
That
is
working,
because
if
someone
wins
a
fork,
all
UI
and
the
system
moves
into
that
universe,
they've,
even
though
they
don't
win
that
market
they've
effectively
want
all
those
which
which
in
a
way,
makes
sense
as
well,
because
if
they
have
the,
if
that
force,
whatever
it
is,
has
the
ability
to
win
that
fork.
Then
they
have
the
ability
to
win
it
in
future
forms
as
well
yeah
exactly.
C
B
B
B
B
Right
or
if
it
was
like
a
sports
game
with
a
very,
very
odd
call
to
it
or
something,
and
they
you
know
they're
people,
sports
traders
are
like
no
we're
trading
in
this
universe
where
it
was
evaluated
this
way.
Instead,
because
that's
the
way
we
like
to
bet
on
things
so
I
could
imagine
some
sort
of
a
scenario
like
that,
where
maybe
the
people
who
bet
on
a
specific
subsection
of
things
and
they
want
the.
B
Into
Zeus
right
like
it
in
the
bastille
day
market,
actually
there
was
a
just.
You
know
one
in
the
best
deal
day,
will
it
whether
it
be
nice
for
the
best
deal
they
parade?
You
can
imagine
maybe
they're
people
who
like
to
bet
on
weird
hard
to
define
things.
Maybe
they're
people
who
want
such
a
market,
for
example
to
resolve
is
yes
when
it
was
sunny
outside
and
they
could
have
their
own
universe.
And
then
people
who
don't
want
to
bet
on
strange
in
difficult
to
word
markets
would
go.
C
A
B
B
The
thing,
if
you
create
a
market
in
both,
though
I
mean
you
have
to
split
rep
into
both.
You
have
to
split
your
open
interest
into
both.
You
have
to
bless
the
quiddity
because
you're
splitting,
so
you
could,
if
you
had
to
create
immediately,
but
it
costs
money
and-
and
presumably
you
are
burning
half
of
your
money
unless
both
of
those
universes
end
up
being
legitimate,
yeah
yeah.
So
it
is
disruptive
and
it
would
be
good
to
avoid
it.
But
it's
important
that
it
exists
in
functions
for
the
game
theory
to
actually
hold.
A
B
A
B
So
it's
sort
of
in
the
same
vein
when
people
report
on
a
scalar,
for
example-
let's
say
it's
the
price
of
Apple:
what
dispute
for
that
could
get
particularly
messy
if
there's
not
some
very
specific
instructions
on
how
to
actually
resolve
it.
Yeah
the
the
argued
method
of
doing
that
is
point
to
a
very
specific,
singular
source
like
like
a
API
feed
or
something
that
will
give
you
a
historical
time
period
so
that
there
can
be
no
debate
about
what
the
actual
number
is.
A
B
Was
a
fork?
Yes,
after
you
redeem
your
your
money,
they
would
fork
in
those
universes
there's
and
that
would
be
bad,
obviously
I
guess
alien,
like
that's
extremely
unclear,
there's
I
mean
there's
some
just
sort
of
rational
processes.
That
would
probably
go
into
that
to
prevent
something
from
happening.
So,
for
example,
if
you're
really
disputing
that,
obviously
there's
probably
a
shelling
point
somewhere
in
here
where
people
can
kind
of
agree
like.
B
Obviously,
if
both
of
you
had
money
on
this
or
even
if
you're,
just
reporters,
it's
doubtful,
people
would
argue
about
amounts
that
small
and
if
you
actually
had
a
why
to
protect
in
that
market,
it's
doubtful
you
would
make
more
by
disputing
that
small
amount.
Then
you
would
just
going
okay,
that's
close
enough
own
make
enough
money.
It
I
just
feel
like
that,
would
have
to
be
driven
by
something
kind
of
irrational
and
it's
difficult
to
prevent
people
from
doing
irrational
things
because
they're
irrational.
So
some,
like
my
guess,
is
in
such
a
situation.
B
A
B
Has
to
work
that
way
too,
because
otherwise
there's
sort
of
a
attack
in
the
forking
procedure.
Where
you
could
imagine
somebody.
Let's
say
this
was
a
categorical,
for
example,
and
interestingly,
Biden
and
Murray
are
also
running
they're,
actually
like
five
own
candidates.
They
decide
to
be
independent
and,
let's
say
yang
wins
somebody
want
to
attack
the
system,
could
just
go
by
and
Bernie
by
demurred
anybody
burning
and
not
let
anybody
else
do
any
disputes,
and
if
that
was,
it
was
actually
yang
that
one
there
could
be
no
true
child
universe.
B
So
we
have
to
assume
that
attack
could
have
occurred
and
let
people
make
in
a
universe
that
could
have
possibly
been
disputed
or
could
have
been.
The
truth
there
has
to
be.
There
has
to
be
a
true
universe,
basically,
so
to
do
that,
we
have
to
a
lot
of
people
to
make
any
universe
that
could
have
been.
B
B
Try
spring
into
it
and
it
gets
even
more
complicated.
If
you
imagine
a
world
where
there's
been
multiple
Forks
so
say
you
know
the
the
election
market
Forks
and
then
in
the
yang
universe.
They
have
an
argument
over
how
many
Twitter
followers
yang
has
or
something
of
that
sort.
Then
is
a
sub
fork
there.
So
I
live.
You
know,
if
you're
using
augur,
you
would
be
using
I
believe
in
the
truth,
where
yang
wins
and
had
a
million
Twitter
followers,
you
know,
but
it
gets,
it
gets
more
more
cop.