omenti

is the drug test fair and the prosecutor's fallacy

a minipost

Some sports have doping issues and there is drug testing in many of these.  There's always an issue of how good the tests are, but there's also an issue of how fair they are and how good they have to be to be fair.   The best way to illustrate this is with an example.

 

Consider a sport with 1000 players - roughly the size of major league baseball in the US .  Suppose steroid abuse is an issue and there's a test with 95% accuracy.  And assume somehow it's known (or strongly suspected) that about 5% of the players are abusers.  These are artificial numbers, but it's just an illustration.

 

What if the players are tested?  We expect 50 to be abusers and 950 to be clean.  The test is 95% accurate.. so 48 (rounding up from 47.5) are correctly identified.  So far so good.  Now consider the 950 clean players..  5% or 48 (rounding up again) are incorrectly labeled as abusers.  So of 1000 players there would be 96 positives and only a fifty-fifty chance of being right. Careers can be be ruined ..  the stakes are high.

 

A way to sort out these conditional probabilities was described the The Reverend Thomas Bayes in the mid 1700s in a paper called Essay Towards Solving a Problem in the Doctrine of Chances.  The good Reverend described the relationship between the probability an athlete is an abuser given a failed test and the probability an abuser fails the test.   Consider these probabilities (apologies for notation)

P(A) = probability a drug test is positive

P(B) = probability a player is an abuser

P(A|B) =  probability an abusing player tests positive

P(B|A) = probability a player has taken drugs if their test is positive

Perhaps the most important takeaway is that P(A|B) and P(B|A) are not the same!  It's a very common prosecutor trick to fool the jury into thinking they are - so common that it's called the prosecutor's fallacy.  But to the problem at hand.

 

We want to know P(B|A).  We already know P(B) or at least we have the 5% estimate.  So P(B) = 0.05.  This also gives us  P(not B)  = (1 - 0.05) = 0.95 as P(B) and P(not B) must add to 1.0. We also have 47.5 (I'll go with the fractional athlete here) of 1000 drug free athletes tested positive, so P(A|not B) = 0.0475.

 

Bayes derived a formula showing how all of these relate.  Sticking with our notation:

 P(B|A) = {P(A|B) *  P(B)} / {P(A/B) * P(B) + P(A|not B) * P(not B)}

                         = {0.95 * 0.05} / {0.95 * 0.05 + 0.0475 * 0.95}

                         = 0.513

This is a very unacceptable test!  Although 95% accuracy sounds good on the surface, it only finds a bit better than half the abusers and incorrectly labels a lot of players.  For fun try a 99% probability and a perfect 100%  to see what's happening. 

 

You can have much more complex conditionals, but this is the general idea.  You really have to think things through a bit..

 

 

local transportation - changing a complex system.

a minipost

There has been quite a bit of activity with escooters in cities as well as ebikes in the past year or two.  I'm hugely in favor of walking, bikes and ebikes - negative on scooters - but it needs to be recognized these are elements of highly localized transportation systems that have a lot of infrastructure, legal and social norms and other elements that tend to have a lot of inertia.¹

 

Change can happen, but meaningful change will take time.  Some areas are easier than others, but the recognized leaders in Denmark and the Netherlands required about twenty years before serious change took hold and fifty years in they're still still learning and experimenting,  The bad news is much of the geography that was determined by the car is wrong if you want change. The good news is there are a number of examples that addressed very different initial conditions and evolved to good enough solutions.  It is non-trivial and the bike or ebike is not the major component.  A lot of hard fought learning exists.  There are some folks who have written extensively on these issues:  David Levinson (his books, blogs and mailing list) and Horace Dediu (lots of work on micromobility!), and Jeff Speck (books) on walkable cities.  If you're interested in the physics and system level issues I'm always happy to chat.²

 

Normally I'd go on for pages at this point, but this is a mini-post, so check out these short videos

 

The Dutch are the world leaders (the Danes are close, but..)  - that said there's a lot of experimentation and variety in the country.  Utrecht is an excellent example near the top:

Just as important, tropical Oslo has decided to make a major change. . an example of a city in the process of doing something (there are many others and there will be an explosion of activity in the EU in the next two or three years .. unfortunately I'd only count one in North America)

___________

¹ Many systems have hidden externalities and complexities.  I've heard more than a few security experts say voting is one of the most difficult and complex systems they can think of.  It's much larger than just voting machines.  And these complex systems have often hidden externalities.  The popularity of the automobile created the suburbs and other non-optimal bits of infrastructure that have had an enormous impact on global warming..

² I've worked with a couple of governments looking at the future of transit.  I tend to start with the physics and then explore the systems.

 

 

moonshine

"Land sakes"

 

A hot Sunday afternoon almost exactly fifty years ago.  My sister and I were on the couch next to my Grandma King.  Walter Cronkite was out connection to the small spacecraft as it slipped towards touchdown on the Sea of Tranquility.  I was just a kid, but I knew this might be one of the most dramatic moments I'd witness.  About then she said those two words and I had another realization.

 

Spacings between the generations on my mother's side were large.  I never met my Grandfather - he died working as a telephone lineman for the Bell Telephone System in Utah, Idaho and Montana years before I was born.  He was born a few weeks before Custer's Last Stand.  Grandma was born in 1889 in the Territorial Utah theocracy.  She would tell stories of the first motor car she saw, of a Butch Cassidy-like thug who ran a small business protection racket, when she read about the Wright Brothers a few years after they flew, and why moving pictures were awful (the batteries emitted a stink).  They didn't have indoor plumbing until 1940, but the big event was getting electricity around 1930 .. bigger than indoor plumbing, bigger than their Ford.  We couldn't get her on an airplane.  Airfares from Great Falls to Salt Lake City were cheaper than riding the dog, but she heard about crashes for much of her adult life .. in fact real air safety didn't begin until the mid fifties, so she wasn't that far off. 

 

And here she was watching glowing phosphors on a piece of glass, listening to a human voices coming from thousands of miles away along with two that had traveled by radio requiring about one and a quarter seconds to find antennas on Earth..

 

What happened in the next few minute was breathtaking, but I found myself wondering if her life spanned greater fundamental change than any other time in history.  Would she see more fundamental change than I would?   You start thinking about what fundamental means,.  It still keeps me up at night.

 

Now there is talk about going back - sending people to the Moon and to Mars.  I'm afraid I don't see it as a big deal and consider it wasteful.

 

Others have articulated what Apollo did for America and the world.  So much was spent - nearly 4.5% of the federal budget was NASA in the mid 60s - that technologies advanced at a much more rapid pace than we'd normally see.   The computer and semiconductor industries were jumpstarted.   Secondary school science education changed and a large STEM pipeline to Universities was created.¹ There was a morale boost for a few years, but it faded against an awful war and dramatic social change on several fronts.  A moon landing, iconic as it was, faded to insignificance in a few years.  The last three Apollo missions were cancelled largely because the public lost interest and politicians responded. 

 

I'm a big proponent of space exploration and the unmanned space program has been brilliant.  Putting people into space is not only expensive, but probably wrong biologically.  We didn't evolve to live in such harsh environments.  Most of the people who have flown in space have orbited in a region where the Earth's magnetic field shields them (and us!) from radiation.   Without Earth's magnetic field it is unlikely life could have evoled on the surface or in shallow water.  Anyone who stays on the Moon or Mars will have to live underground or in elaborate radiation shields.  There are issues of water, air, food, human psychology, low gravity muscle and bone degradation and so on.   We're the wrong tool for the job.

 

People, life support systems and rockets didn't scale by Moore's Law, but computation has.  What a non-human kilogram in space or on another body in the Solar System can do has made leaps opening and now we have commercial businesses exploiting near Earth orbit and students orbiting satellites.  We can build increasingly intelligent probes that are space faring.  Plucky (I've wanted to use that word) robot explorers  are getting better and better and do almost all of the meaningful exploration of the Solar System.  We're really good at building and flying them as well as analyzing and wondering about what they tell us.  

 

There are those who sayt we have to leave Earth - that it's our destiny and we've ruined the Earth.  That strikes me as defeatist and driven by more by science fiction than logic and science.  A failure of the imagination.   Another explanation is adventure.  Adventure is fine, but you can't call it real exploration at this point.   My vote is for learning about the Solar System and beyond in a cost effective way as we work on repairing the damage we've done here.  

 

We can always look up and admire the Woman in the Moon.

 

 

I would be remiss if I didn't point out that Grandma King swore in Irish Gaelic.  Sadly I never learned.

__________

¹ I've written quite a bit about the good and bad of the approach.  Arguably it's created an over-supply of scientists at the cost of ignoring and even alienating the majority of students. It's sad it hasn't changed much since.

complete with lighter and ashtray

About a month after Apollo 8 circled the moon my Dad and I traveled the two miles to Malmstrom Air Force Base for a tour of the SAGE installation.  A number of these Semiautomatic Ground Environments were scattered around the country to detect and coordinate a response to bomber attacks from the Soviet Union.  The cost was enormous - more than the Manhattan Project - with the effort resulting in several computing "firsts".    It formed how I thought of computing at the time, blinding me to another more important revolution that was underway.

 

The SAGE building was enormous. It had to be to house the roughly supermarket sized FSQ-7 computer.  These were tube based and designed to be redundant enough that statistically likely tube failures could be repaired by airmen with replacement modules that could be swapped inside of a minute.  Outside the blast hardened building was a large cooling plant to keep everything from melting down. 

 

The big show came with an airman pointing a light gun at a large circular display with radar blips from a number of radar stations. He'd pull the trigger and get some basic information on the airplane - friend or foe id, call sign, altitude, speed and heading.   Should an intruder be spotted, interceptors would be scrambled.  In Great Falls that meant F-106s from the Montana Air National Guard about seven miles to the West on Gore Hill.  It was co-located with the commercial airport.  I believe it was the only commercial airport in the US with nuclear weapons on site.  All this was really impressive, but so was the fact that each of the radar terminals had an ash tray and a cigarette lighter.

 

I thought computers were huge expensive beasts - rare things that I might get to program some day, but certainly not anything that would be part of my life.   It turned out what took place the month before provided a hint that didn't register until much later.

 

While many of the fundamental ideas behind computing had been put together in the fifties and sixties, miniaturization and the real start of Moore's Law really came with the space program.  Simple integrated circuits had been developed, but size and weight requirements on the Apollo command module and the lunar lander called for huge advances.  The advances were borderline unthinkable, but solved with cubic money and outstanding minds.  My favorite example is the Apollo Guidance Computer.  Books have been written and I wouldn't add anything, but here are a couple of gee-whiz links.

 

The first is from Ken Shirriff's terrific retro computing hardware blog. Ken's hobby is reverse engineering ancient electronics and restoring old computers.  In this post he talks about core rope and the guidance computer.  Programs were literally woven into a rope that was a simple core memory system.  It's elegant in a hammer and tongs sort of way.  If you don't know how core memory works go and check out the photographs and skip the technical parts of the text.  It's a remarkable story involving women who could sew and sexism.  And if you have the background it's even more interesting.

 

And here's a nifty Apollo Guidance Computer Simulator.   Print out the instructions or open them in another window and play astronaut!  Just in time for the 50^th anniversary in a few weeks.

 

The dramatic infusion of R&D money built the semiconductor IC industry and established starter markets. There many other stories - like how AT&T and IBM worked together behind the scenes in the 60s and early 70s to bootstrap the necessary tool building. They did it for their own benefit, but without it we might be a decade behind where we are now.  Perhaps a more interesting question might be:  'would that have been a good thing?'  I've been thinking about that a lot recently.