Numb3rs (2005–2010): Season 2, Episode 7 - Convergence - full transcript

A series of home invasions become increasingly violent. Charlie's most famous work, The Eppes Convergence, is scrutinized by a former rival.

What do you say, Larry?
The Eagle, the Shark?

Oh, I think, the
Eagle. More stability.

Let's go, Bro. Show
us how it's done.

Nice. Ah, that close.

Nice. All right. Watch this.

All righty, go for it.

Oh, yeah.

Oh!

Oh. Nice!

Nice throw, Brother.
Sweet, right?

You know, Don, you
and I are very alike.



We're both focused on large,
possibly unattainable goals.

Me trying to explain all the
workings of the universe...

Yeah, and what am I doing?

Trying to take all the unworkable
workings and put them in jail.

Good luck to both of us.

Eppes.

Yeah. All right.
Okay, I'm on my way.

Sorry, guys, gotta go. You
know, work. I'll see you, all right?

Yeah. Well, there he goes,
back to the unattainable.

Hi. They killed the father
when he tried to resist.

So it's just like
the third robbery?

The same guys. Ski masks,

physical descriptions match
the other home invasions.

This puts us up to
seven in six months,



so they're coming
at faster intervals.

They go after
any specific items?

Yeah. Jewelry, paintings,
high-end furnishings.

The deceased is
Dr. Richard Bloom.

He and his wife arrived home
with their 15-year-old son, Jordan,

after a school track meet.

Three guys then rushed
them at the front door,

and bound them
up with duct tape.

Did you see this? Yeah.

The son put up a struggle,

and one of the burglars hit him.

The father tried to stop him,
and they threw him in the pool,

hands and feet still bound,
and laughed while he drowned.

Daughter's name is Jill.
She was at a friend's house.

Her friend's mom brought her
back home, saw something's wrong.

She called 911.

Daughter didn't see anything.

Thank God.

Right now, someone has
to tell her, her Daddy's dead.

We all use math every day.

To predict
weather, to tell time,

to handle money.

Math is more than
formulas and equations.

It's logic.

Math is more than
formulas and equations.

It's rationality.

It's using your mind to solve
the biggest mysteries we know.

So, we got seven
home invasions, right?

This is before LAPD asked
for our help. This is after.

What we know is that
they... they do have weapons,

although they haven't fired them,
so we have no useful forensics.

And they all live in
exclusive neighborhoods,

they all have high-end,
expensive valuables in their homes,

and they all have
security systems.

But the attacks take place
just as they're coming in,

and the alarms
have been turned off.

Who are they?

They're both fathers. They both
fought back to protect their families.

I mean, we can't even figure out
how these guys are picking the homes.

No. So far, there's no
commonalities between the families.

You know, other than the
fact that they're wealthy.

Let's try data mining.

It's a system for
analyzing information.

Army Intelligence uses that

to identify possible
terrorist attacks.

I design an algorithm
based on these crimes.

The algorithm
then takes the data

and identifies correlations.

All right. I'll get you everything
we have on the families.

Yeah, where they live, where
they go to school, shop, dine out,

who cleans their pools,
who fixes their cars.

I'll need more than that.

More than everything
on the families?

More. On everything.

All crime stats for the past
six months for the entire county.

Wouldn't that much data make it
harder to find what you're looking for?

The opposite. More data means
more chances to find something.

It's like when you're trying to
put together a jigsaw puzzle.

You start with a few pieces,
and the rest are in the box.

All the pieces you have should
eventually fit in your puzzle.

All the pieces and nothing
more come in the box.

But with a real-world problem,

that's just like trying
to solve a jigsaw puzzle

when all the pieces you need are mixed
in with pieces from many other puzzles.

Now, when you grab a few
pieces from the box, most won't fit.

You gotta go
through the entire box

and pull out the pieces
that fit your puzzle.

The algorithm
goes through it all,

pulls out what fits together.

All right. Well, I'll get
you our reports, okay?

Hey, guys. Hey, I've got a
complex data analysis problem.

Either of you free?
B-Both of you? I prefer both.

I... I have some,
a couple of hours.

Charlie, have you seen the guest
lecture schedule for this week?

No, I haven't. Why?

Oh, don't... Damn it. Don't tell me I've
gotten stuck hosting another reception.

No. Uh, some guy's giving a talk on
the asymptotics of random matrices.

Hey, you've done some
distinguished work there, haven't you?

The guest lecturer's name is...

What the hell is that
son of a bitch doing here?

To whom are you referring?

Marshall Penfield.

That's the guy. The
guy that's giving the talk.

Oh, really?

Marshall and I, we
attended Princeton together.

Every chance he got,
he slammed my work.

And the only keg
party I ever threw,

he stole the keg.

Marshall.

Hey, Eppesy. You
coming to the talk?

It's gonna be a good one.

I just heard about
it. Ever so excited.

Wondering what
your subject could be.

Oh, same old, same
old. Hi. How you doing?

You found a new approach?

Only because I recall
your earlier attacks

on my analysis of
Plancherel measures.

You made an ass of yourself.

Yes, but unlike you, I
continued to work in the field.

And I made some
interesting findings.

Elaborate.

A flaw in your work.

I'll spare you the details. I'll
save it for the talk, but, bottom line,

your big breakthrough, the
Eppes Convergence, it's wrong.

It was nice meeting you.

These things, most of them
have been in our families for years.

I don't care about any
of this. But Richard,

he was only trying
to protect our son.

I know, Mrs.
Bloom. I'm so sorry.

He just couldn't bear
to see people hurt.

It's why he became a doctor.

I just think if maybe you could tell
me anything about one of these pieces,

it might help me find
your husband's killer.

There's a lamp. It was,
uh, made by Dirk van Erp.

It's one of a pair.

Richard got the first one
for me as a wedding present,

and he spent years
trying to find the other one.

He finally found it.

He said they belonged
together, just like we did.

So this lamp's
worth, like, $100,000.

Which, I imagine, you
can't just sell anywhere.

Yeah, that's what I'm thinking.

We should check into whatever
markets deal with this kind of stuff,

see if there's any
buzz. Mmm-hmm.

Penfield hates me.
It drove him crazy.

I became the youngest person
to ever publish a major paper,

and when I gave my first seminar
on the Eppes Convergence,

he shows up with a
button that says, uh,

"Don't believe the hype."

Okay. Well, that's a
little confrontational.

But, I mean, you... you
were a star in your field.

You... You got to expect
this kind of attention.

This man has singled me out.

No. Your status
has singled you out.

I mean, people... people are scrutinizing
your work because of your reputation.

You must understand, that's why
he's doing it. To ruin my reputation.

The Eppes Convergence made me.

If it's wrong,

then what am I?

Well, a talented theoretician
with an ego problem.

I'm sorry?

Hello? Yeah,
I'll be right there.

I gotta go. Okay.

The computer lab says the algorithm
for Don's case kicked out some findings.

Yeah.

Well, wisdom and genius,
rarely present in equal abundance.

No, I don't remember where I got
this lamp. Probably some estate sale.

Wrong answer.

Hands behind
your back right now.

What's going on?

You know what's going on. You
can tell us where you got it here,

or you can tell us
down at our office.

That's after we walk you through
your showroom full of customers.

Wait, wait. Okay.

Where did you get it?

I was introduced to this
guy a few months ago.

H-His name is Roley.

He told me he got the lamp
from a bankruptcy procedure.

But you knew better, didn't you?



Colby.

Your brother is starting
to freak me out a little bit.

What, just a little?

Hey. What's up?

Calendars, wonderful
analog mathematical tools.

They create a universal
timeline for societies

by interpreting
astrophysical movement.

What are you talkin' about?

Calendars aren't based on
easily divisible factors. Okay?

You know, they originally
tried to make them like that.

360 days divides quite evenly,

but it doesn't match with
what you see in the night sky,

so ancient calendars would add
an extra month every eight years,

or tack on five extra days.

So, what, these were all the
dates from the home invasions?

Right. Once I got the data
mining algorithm up and running,

I ran a Fourier analysis on the
chronologies, the dates, the times.

We've actually been lookin' at trying
to find a timeline pattern, Charlie,

but we can't come
up with anything.

Because you weren't looking in
the same way my analysis does.

The red days are known attacks.

The blue days are
when the analysis says

there should be
attacks, but there aren't.

And what we must
ask ourselves is,

how do you know there aren't?

Wait, what do you mean? I mean, no
homes were hit on those dates, right?

Let's assume that these same
guys commit other types of crimes.

You mean, not
just home invasions.

So, you're saying, what, if we find new
crimes, maybe we find new evidence?

I think you should be looking
for crimes committed by gunmen

that fit the same description, that
occur in these days that I've marked blue.

And, Charlie, the pattern, can it tell
us when the next attack would be?

Patterns like this are...
are clearer in hindsight.

I mean, they're... they're not
really used as... as predictors.

I mean, there's an
erratic frequency here.

I mean, you know, there's several
robberies a couple days apart,

and then there's one or two hits,
and then an even shorter break.

Just point us in a direction.

This is my best guess.

It's possible that the
next attack could occur

in three days.

I checked the incident
reports on the days

where Charlie said there
might have been related crimes.

I found three carjackings
done by three guys in ski masks

that match the physical
descriptions of our home invaders.

Our guys go for the high-end
goods, though, Megan.

The cars stolen were a Mercedes,
Range Rover and a Porsche.

Yeah? Any of them found?

No, which means either they
were repainted and given new VINs

or they were sold
outside the area.

These guys seem to know
how to move expensive items.

And there's somethin' else.

We've never had a gun fired
in any of the home invasions...

There was a gunshot
in the carjacking.

Yes, there was. At
the Porsche robbery,

one of the members of the crew

shot a round into the
air to scare the owner.

We find that bullet, we might
be able to make a match.

You wanna find a bullet fired
straight up? It's not gonna be easy.

Well, it's not impossible. Just find
the cops who responded to the call.

They would've done a
neighborhood canvas,

tried to ID the
round that was fired.

I think we should be talking to
the people whose cars were jacked.

We should be running down the
lead from the antique store owner.

What's the fence's
name? Roley. I'm on it.

Good. All right

A bad time?

No, come on in.

I'm just sitting here, waiting for a
new run of the data mining program.

Emergence theory.

The whole is greater
than the sum of its parts.

Individually, ants have
limited capabilities.

But the colony can be viewed
as a single complex organism.

But you guys aren't here
to talk about ants, are you?

No, actually, we
have a little bet.

See, David here seems to think
that you might be able to help us

find a needle in a haystack.

A bullet in Brentwood,
to be exact.

Checking the time
periods where you thought

there might be related crimes,
we found three carjackings.

One of them, there was gunfire.

We find that bullet, we may
be able to match it to a gun.

This bullet, was it fired parallel
to the ground or at an angle?

According to the
crime report, into the air.

We're not sure
of the exact angle.

What type of gun?

Well, we're actually not
certain about that, either.

Okay, we're looking
at at least two variables.

Any handgun's going to fall into
a certain range of muzzle velocity.

Now, we're assuming that the gun was
pointed up, not exactly vertical? Okay.

So, that'll give
us a radius of...

Two to three thousand
feet. Roughly half a mile.

I mean, there are a lot of
assumptions built into these calculations,

but I suppose that
can't be helped.

Agent Sinclair, Agent
Granger, this is Dr. Penfield.

Sorry to interrupt. I just heard the
familiar staccato clacking of chalk,

I thought it might be
something interesting.

You're still coming to
my presentation, right?

I might get something wrong
about Hermitian random matrices.

You know, I've moved on to sequences
with orthogonal symmetry, man.

Oh, really? Mmm-hmm.

Well, that's funny, 'cause
that's the exact same thing.

Resonances in
quantum chaotic systems.

That's a simplistic analysis.

You always were
testy when challenged.

You know, I was talking
about that with Amita, um...

You were, were you?

Yeah, yeah, last night.

We were, uh, we were hangin'
out, you know, got a little dinner,

a long conversation. That
doesn't bother you, does it?

Marshall, do whatever you like.

Just you remember, Amita's
a sharp mathematician.

So, no matter how hard you try,

you're never gonna get her to
believe that this is six inches.

I bet with you that
subject's never come up.

Ooh, math fight.

Hey, that was carjacking
victim number three.

Just like in the home invasions,

these robbers
targeted specific items.

They went right for his watch,

which was hidden
underneath his jacket.

Carjacking victim number two,
they made give up a $55,000 ring.

All right, so they're
being followed, right?

Forced to pull over. No, no.

Carjacking victim number
one locked her keys in her car.

She called AAA
on her cell phone,

and ten minutes later,
these guys showed up.

Victim number two got lost, entered
the destination into her navigation system.

And victim number
three ran out of gas.

He called his onboard assistance
service and spoke to a real operator,

but the robbers got there
first with a can of gas.

It's like they're eavesdropping
on their cell phones or something.

Maybe, but that doesn't explain their
knowledge of their personal possessions.

And they all have
different cell phone carriers.

I mean, they're tapping
into 'em somehow, right?

Yeah. The question is,
what's the common link?

Just don't push the wrong button

or else we'll be looking for
that bullet the old-fashioned way.

It's okay, all right, Charlie's
already programmed

the trajectory equations
into the calculator.

Yeah, well, we have
the origin point of the shot

and the equations will
generate high probability areas

for a search.

It's you plugging in those
variables that has me worried.

Hey, man, it's only
two variables, okay?

We got muzzle velocity

and inclination of the weapon
when fired. I can handle it.

Yeah, well, at least I've used
a graphing calculator before.

Really? You have?

High school, man.
Why? You didn't?

I went to high
school in the Bronx.

Get your gear. We found Roley.

Roley. He's the fence that sold that
stolen lamp to the antique store, right?

Yeah, we're settin' up a buy.

Hey, Charlie.

Hey. What's goin' on?

Oh, it's all my work from
my, uh, city planning days.

You see, I'm taking it all down to
the offices that we're leasing now.

You're really starting the consulting
company with Stan Fischer, huh?

Why not? You know, I miss work.

Now I can be my own boss

and I can concentrate on the
aspects of design and planning

that I always enjoyed myself.

It's a big investment.

Oh. I have money.

Thanks to you buying this
house at the asking price.

So you're not
worried about the risk?

Well, that's business.
In fact, that's life.

You know, I never
really wanted to retire.

But when your
mother got sick, uh,

there really wasn't much choice.

And now, I have a choice.

I'm really happy for you, Dad.

You okay?

You seem a little, uh...

Stressed. Yeah.

Why? What's wrong?

Marshall Penfield,
fellow Princeton graduate,

has come here to rip
apart five years of my work.

And it's made me realize that I haven't
focused on anything new since then.

I... I've been wasting my
time with this stuff for Don.

Oh, come on. You don't
mean wasting your time.

You're right, I don't. I... I'm
not wasting my time, I just... I get

so easily distracted.

Yeah, well, it's more like
you're easily fascinated.

I need to devote myself to
something significant in my field.

Important to other
mathematicians.

Yeah, well, you know
what I think, Charlie?

I think your life's work should
be what you find important.

Are you the lady Jay
sent over from his store?

Yeah. He said you get him
some of his best inventory.

I'm Roley.

This here's Timmy. Hi.

Sorry to make you come out so
late, but, uh, you know how it is.

Oh, I've been to
worse neighborhoods.

So, let's see if we
could do a little business.

You, uh, brought money, right?

Oh, I brought better than money.

FBI! FBI! FBI!

Get down!

FBI!

You stop right there! Stop! Get down
on your knees! Hands behind your head!

Your right hand.

You know where
this came from, right?

I'm really not one to
ask that many questions.

No? Maybe you ought to rub it three times
and find out where it's about to send you.

This is a list of everything
that was in your storage unit.

And this is a list of everything that
was stolen from seven home invasions.

They're kind of similar.

I had no part of that.

Having this stuff
makes you a part of that.

I'm telling you, I am
not that kind of guy.

You... You got a nice painting or
a good piece of period furniture,

I can find you a buyer.

Who's not gonna ask
too many questions.

That's right.

I find people with money

who are interested in
deals on luxury items.

And that's my niche.
Killing? Not my niche.

Well, how'd you
get all this stuff?

It wasn't from any
crew, it was just...

It was two guys.

And how'd you meet them?

They came up to me at a bar,

said they knew me by reputation.

And what makes you think
that these guys aren't robbers?

One of them was dressed
better than my lawyer.

And the other guy, he
was just a total geek.

I'm gonna need their names.

Mr. Brown and Mr. Gray.

That's... That's all I know.

Then I feel sorry
for you, Roley.

W-Wait... Wait a minute.

All right, these guys, they're
not barehanders, all right?

They're way too
business. How so?

For one thing, they gave
me a list, like this one.

With everything they were
gonna bring me weeks in advance.

Said it was so I could
locate buyers ahead of time.

Roley's description of the men he
dealt with don't match what we have.

So then, we... we could be
looking for five different guys?

Maybe not five different guys,

but maybe two guys that case it

and three guys that
do the actual robbery.

I'm telling you, the
cell phones tie into it.

Yeah, but every single
person uses a different service.

And how would they
know what's in the homes?

And what about the data mining?

I'm working it. The analysis
finds all kinds of links

and I have to gauge
which ones are meaningful.

For instance, the
data mining found

that all the victims buy bottled
water. Well, that's meaningless.

When Army Intelligence
tried to use data mining

to spot commonalities
between terrorist attacks,

one of the people it picked
out was Condoleezza Rice.

Okay, so it's thorough,
but it's not discriminating.

I mean, whatever
you can do, Charlie.

I mean, you're the one who said
we're talking about three days here.

Yeah, I know.

Variable two is muzzle velocity.

We're working with
a nine millimeter,

but these guys
like to play rough.

Now, I'm thinkin' they're
packing something stronger.

Yeah, but muzzle velocity changes
with the grain of the bullet, anyway.

The heavier the projectile,
the slower it goes.

Right. Okay, let's just say, you're
a guy into invading people's homes,

and you're getting off
on messin' them up.

Right. What's your
weapon of choice?

Semi-automatic with a big clip.

Lots of stopping power.
.45 ACP, 180-grain rounds.

That's a muzzle velocity
of 1,200 feet per second.

Oh, yeah, man, plug that one in.

The development of any
type of representation theory

is usually split
into two stages.

The first is a description of
all elementary representations.

The second consists
of constructing

certain natural
representations of the group,

and decomposing them
into irreducible ones.

What we see when we get
deeper into the Eppes Convergence,

is a small and, until now,

undetectable
conceptual roadblock.

It's an expression, floating
through these equations

like a tiny blood clot,
until it lands here,

at the infinite-dimensional
simplex,

which causes a stroke
to the entire theory.

Maybe these guys
are into revolvers.

You know, a nice little
.38, or .357, maybe.

I'm telling you,
man, they're physical,

they're pumped,
they're flush with cash.

These guys buy a gun, it's
gonna be a .40 caliber 45 HK.

David, we can talk muzzle
velocities all day long.

But if this joker fired
straight up, it doesn't matter,

because the bullet
would've landed right here,

within a 500-foot radius
of where the guy fired from.

The question is,

do you think it's possible
that he actually fired straight?

Yup, I do think it's possible.

Yeah? Based on what?

Based on I just
found the bullet.

Charlie?

He does have a valid point
about the end-point process

on a one-dimensional lattice.

Right.

Penfield's blood
clot to my brain.

Or was it the torpedo in the
engine room analogy? I forget.

Hey, Amita, there you are.

You still up for Pie n' Burger?

Sure, let me grab my stuff.

Hey, Eppesy.

Heard you got a substance-abuse
problem with lemon meringue.

It's a gateway pie, you know.

Why don't you, uh,
come meet us out front?

You know, there's a whole
crowd of people going.

I'll pass.

What's the matter?

I can't make any
headway on this FBI case.

And there's my life's work

going up in flames.

Well, you're spreading
yourself really thin.

You're spending a lot
of time with Marshall.

He's an interesting thinker.

It's not a big deal, Charlie.
You've got your FBI work,

your teaching, your research...

Yeah, and you've got
your astrophysics work,

your... your work
with Dr. Keppler.

I know. I'm focused on my work.

I'm not sure that I wanna
focus myself socially right now.

I guess I can understand that.

You sure you don't
want some pie?

No, I... Go eat pie.

Please, I need to think.

Okay.

Do you two realize
what's in here?

Well, in your case, a lot
of unanswered voice mails.

I can tell you that.

After 9/11, the FCC mandated that
all cell phones be able to locate people

in the event of an emergency.

Oh, right. That, uh,
GPS locator chip

has to be put in every
new phone by 2005.

And newer phones can be used to
track people within a range of 30 feet.

But you're still gonna need
access to classified data.

Are you saying that these
robbers have figured out

how to intercept encrypted
material from a military satellite?

No, that's impossible.

Listen, somehow they've
gotten that GPS information.

They've gotten into the phones.

Well, that doesn't explain
how they knew who to rob.

Look, this is a
two-part problem, okay?

Part one, how do
they know who to rob?

And part two, how do they
know where the victims are?

The phone is the key to
part two, locating victims.

Part one, how they know who
owns what, I still can't tell you.

But you can be positive that all
the families will have it in common.

Hey, Charlie, you
seen my cell phone?

I been looking all
over for it, I can't find...

What are you doing with it?

We'll put it back together.

You can't use one of your own?

Well, I refuse to allow the
small, quiet moments of my life

to be invaded by these
devices of distraction.

Which means you don't have one?

And I need mine.

Look, Charlie, I'm not a
homebound retiree anymore.

I happen to be
starting a new business.

And having a cell phone
might come in handy. Right?

Look, you'll have
it back in one hour.

Dad, we need to
look at your GPS chip.

What do you need to do that for?

Every new phone comes
embedded with this tiny chip

that acts as a global
positioning receiver.

And it's my theory that
these home invasion robbers

are using them to
track their victims.

How would they do it?

Well, a cell phone's
essentially a radio.

Whenever it's on, it's
sending radio signals

to let cell towers know
where the phone is

so a person can move
from one cell zone to another

without losing a call.

At the same time, the GPS chip
receives signals from outer space,

specifically, from at least
three geosynchronous satellites.

Using the geometric method called
"three-dimensional trilateration,"

the GPS chip uses the
signals from the satellites

to calculate its exact
location on Earth.

The data generated
can be used by police,

roadside assistance,
emergency services,

tracking truck fleets,

and even lets parents keep
tabs on their teenage children.

Ah, it's too "Big
Brother" for me.

I'll stick with my
rotary dial-up.

But intercepting these signals

is really beyond the
grasp of most criminals.

If all my years of working at Urban
Infrastructure has taught me anything,

it's that there's always somebody
who gets into the system.

Charlie was right.
Three days on the button.

Married couple,
74 and 77 years old.

From the evidence, it looks like
the man struggled with the robbers.

Yeah, they beat him to death, got the
woman over here. They strangled her.

Doesn't look like they
put up much of a fight.

They're acting more frequently,

and killing more readily.

I am crossing my fingers. This has
to be the bullet from the carjacking.

Not a lot of bullets
lyin' around Brentwood.

No, it's more of a slash-your-throat
kind of neighborhood.

Well, we're running the slug
through the ATF Bullet Trap database,

hoping to match the gun it was
fired from to a previous crime.

We need this lead.

And after eight hits, they're
gonna think they're invincible.

That we can't catch them.

They're gonna be more
reckless, more dangerous.

Bingo.

Yeah, I heard on the
radio, two more people killed

by these robbers
Don is trying to catch.

Yeah. It's bad.

And the data mining is not yielding
any connections you can work with?

Well, it's hard to say. You
know, one commonality

that the analysis keeps
highlighting is cell phones.

But, you know,
everyone has cell phones.

Then there's a second
commonality. It has no defined value.

The algorithm keeps pointing out

that the families all have rare,

collectible items,

and all were insured.

You really should get
this program to Amita.

She is the best coder we've got.

I don't think this is
where the problem is.

And Amita's off
chilling with Penfield.

Well, he is a brilliant
mathematician

and, I assume, a
sparkling conversationalist.

Are you serious?
Oh, you're great.

You're... You're just another
member of the Penfield fan club, right?

Well, I do understand
his analysis

of the Eppes Convergence
was a bit pointed.

I'll admit, he's
a bright fellow.

He's got all sorts of insight,
but he's so full of himself.

He's so annoyingly confident.

"Certainly in extending
the calculation for S

"to the infinite unitary group,

"it would be obvious to anyone
that my genius would mandate

"upon my death the inclusion
of my brain in the Smithsonian."

I would actually prefer someone
take my brain on a road trip

like they did with
Einstein's. But...

Hey, Charlie.

Hello.

Charlie, Marshall and I went
to your office looking for you,

and, uh, we kind of saw
what was on the chalkboard.

Yeah, you had a problem
that was of interest to me

'cause it relates to my main
field, Set theory. Oh, well...

This is it, isn't it? This is the...
This is the same thing right here.

Could be.

Why?

Marshall's new research
at Princeton is Inset theory,

in something he calls
"deep current" sets.

What are those?

Chalk.

Connections between
groups that are hard to detect,

but, uh, that affect
all the other sets.

Kind of like how currents
beneath the ocean's surface

can influence the weather
patterns in the atmosphere.

There's something
underneath we can't see.

Things that seem to
have no connection

are all linked by
a common factor.

Marshall, have you developed
any methods of analysis

to help you define the
unidentified connections?

Oh, yeah, sure.

Even if I can't tell you where the deep
current is, I can tell you where to look.

But, Charlie, all eight families

have different
insurance companies.

Sure, but the data
mining found that

there's a powerful connection
between the victims.

And it's not something
we could see.

It's an unseen set,
a deep current set.

I'm sure that means
something. Right.

The victims' families all
had high-end valuables,

like the Dirk van Erp lamp that
you guys tracked down, right?

Right, that the robbers
all knew about in advance.

The victims' families all had different
insurance companies, but my analysis suggests

that the insurance companies
have something in common.

Like what?

Well, apparently, when
covering rare, high-cost items,

insurance companies
often purchase a rider,

a backup policy
from another firm.

Yeah, insurance
riders, of course,

which wouldn't show
up on a criminal report,

because the customers don't
even know they have them!

Hey, Colby, Charlie's
got a good idea.

Check out if any of the insurance carriers
bought any riders from the same firm.

If I'm right, you'll find a company
common to several victims

and someone with access to
that company's list of clients...

And would know how to
hack into cell phone signals.

Megan, get a list of local
employees with criminal records

from the wireless
companies, all right?

I have that.

You're fast.

Thank you.

Don, I already looked at three of
them. They used the same underwriter.

It's a BTL Limited, so I'm gonna
check and see if I can get employee lists.

Good. Great. Now, compare
that list to Megan's list,

and maybe we'll
find one guy on both.

No, we're looking for two
guys, Mr. Brown, and Mr. Gray.

Remember, one will know cell
phones, one's gonna know insurance.

This is the FBI's
version of data mining,

finding links the
old-fashioned way.

Oh, yeah, by hand and by luck.

Here we go, Herbert Quilty,
service tech for Vertech Wireless,

two years ago pled guilty
to selling stolen cell phones,

served three months
in county jail...

What's his link to the
insurance company?

Same time as Quilty was in jail,

a guy named Thomas Maynard was
serving weekends on a DUI conviction.

Maynard was a supervisor
at BTL Underwriters.

How did Roley put it? "One
dresses better than my lawyer,

"and one is a
total geek." Right.

Freeze! DAVID: Hands
behind your head!

Down on your knees! Down!

Herbert Quilty, you're
under arrest for murder.

All right.

I had that wine at the
Mortons' last week.

FBI! Get on the ground!
On your knees! Down!

Okay.

Hands behind your back, Maynard!

You're working with some
nasty dudes, Maynard.

Where'd you meet guys like this?

Weekend lockup seems
a bit tame for stone killers.

No evidence connects
me to any crimes.

Can't control these
guys, can you?

Thought you were getting
into the robbery business.

They switched the
game on you, didn't they?

You set it up so they
could stalk their victims,

in their homes, in their cars.

That puts you right
in the middle of things.

Walter Gordon. His
friends call him "Demento."

Leader of a San Fernando
Valley Skinhead gang. Hmm?

Never seen him
before in my life.

Yes, you have. What
happens to you now?

What do you think
he's gonna do to you

the next time the two of
you run into each other?

And what about your family?

If you're locked up,
and he's still out there?

They weren't supposed
to hurt anyone.

That's all on them.

I picked people from the company
files. Quilty rigged the phones.

With GPS to the node signal.

Look, Demento's crazy.

I tried to keep
him to a schedule,

but he wanted to
move faster and faster.

He decided to pull
these carjackings.

Said he wanted to be known
for selling expensive wheels.

Nice timing on the photo.

Forensics matched the
bullet we found in Brentwood

to one shot through the head
of a fast food worker in Tarzana

about two years ago.

Witness in that case ID'd
Walter Gordon, aka Demento.

What I don't understand
is how did Quilty's phone

let the robbers know
where the victims were?

Well, Quilty must have
put a sniffer on a node.

A node is something that takes
all the calls coming into a cell tower

and then relays those
signals to where they're going.

A sniffer is an electronic
device that detects specific calls,

then sends that signal to a
phone the robbers must have.

And you think Quilty's phone
could do that? I think so.

All he needed was
access to the cell towers.

Because he worked at a
wireless company, he had that.

And the gear he required,
well, he could buy that

at any electronics
store, couldn't he?

All right, so Maynard
gave Quilty phone numbers

of people who
owned valuable things.

Right, and Quilty had his
sniffer search for those numbers.

When those signals came in,

the locations went to the
phone that he gave the robbers.

They could follow the victims,
move in as they arrived home.

Walter Gordon, aka Demento,
has no known address.

LAPD's been looking for
him since the Tarzana killing.

Okay, so Maynard and
Quilty set up Gordon

with phone numbers
of potential victims.

Well, see if they
picked any new victims.

I mean, if we can't find Gordon,

let's find out where he's going.

Come on. Take me to your home.

There's the Volvo.

This family won't
know what hit 'em.

I got the driver,
you get the girl!

Out of your damn car!

Get out of the car now!
- Get out of the car!

Drop the gun! Drop it!

Don't move! Get your
hands behind your back!

Come on. Get him up.

- What the hell is that?
- Shut up.

Can't you see the
baby's sleepin'?

Marshall. Eppesy.

I just thought you
might like to know

that your insight on
the subset commonality

helped the FBI catch
three murderers.

Wow.

That's, uh...
That's... That's great.

So, thank you for your help.

You're welcome.

I... I hope you understand,

I... I didn't work on the Eppes
Convergence because of you.

I did because, uh...

Well, because it's great math.

You know, back at Princeton, you
were always the guy with the deep ideas.

This... This one
just fascinated me.

Well, I have to admit,

your work shows
tremendous insight.

Oh, thank you.

Though I've taken another
look at your analysis.

Oh, no. I know that tone.

And I think I've found a way to
address the flaw that you've identified.

That's the basic idea.

That... That negates the whole
need for the flawed section, then.

Well,

the great Professor Charlie
Eppes has done it again.

No, that... that's hokum.

This approach was only
apparent because of your work.

So I'm gonna call this
the Penfield Variation.

Great

way to make me
feel like a real ass.

Door's open.

Hi, Mr. Eppes.

Hi. Hi.

Charlie's supposed to
come with us to a seminar.

Oh, he's out in the garage.

Hey, Charlie, your friends...

Here he comes. He's been
out in the garage all day.

Hey, guys, you know what?
Go on ahead without me

maybe I'll catch
up with you later.

'Cause I'm a little caught
up with something right now.

What are you working on?

A unified theory of
the neural network

in higher cognitive functions.

The math of the brain.

I'd better stick to it, too.
It might take me a while.

Say, several decades.

Attaboy.

Yeah, what is it they say,

"The journey is more
important than the destination"?

So true.

You're doing a jigsaw puzzle.

Oh, yeah, it's 5,000 pieces.

My wife used to
do them all the time.

So what do you say you skip the
seminar, and we'll order in some pizza?

I love jigsaw puzzles.

And I love pizza.

All right.