So, why I am on probation is irrelevant.  The purpose of this story is to help people who have an upcoming poly by sharing my experience.

First off, I definitely should not have passed.  I lied on pretty much every relevant question.  I did pass though!  To achieve this I applied a wide array of techniques, some of which are advocated by 'The Lie Behind the Lie Detector', others not.  Here is what I did in preparation for the months beforehand:

1. I read TLBTLD numerous times and looked at the other resources available on this wonderful website.
2. I practiced my baseline breathing for months ahead of time.  If you can catch yourself in the middle of a subconscious breath and determine the amount of chest vs abdominal breathing you should be able to match your conscious baseline breathing to that and not worry about the breathing so much except when it is important. 
3. I recorded my own mock polygraph questions ahead of time and practiced countermeasures while listening to these recordings.
4. I practiced countermeasures whenever I had a chance, e.g. while driving to work, on the train, etc.
5. I changed my uncertain view on the validity of polygraphs to a definite NO.  This step I feel is critical.  When you are in there, the slightest doubt in yourself or belief that these things hooked up to you can detect anything more than changes in the physiological measures they are measuring, and you are likely to fail.  It's more than believing they are bullshit, it's knowing they are bullshit.

Here is what I did in prep for the day of, and the day of:
1. The night before, reread TLBTLD chapters 3 and 4, practice everything one last time
2. The night before, smear 'certain-dri' all over my fingers.  I know this is an urban myth, but I actually use this stuff for my excessive perspiration, and I stay BONE dry, and it doesn't wash off for 48hrs with soap or anything.  So, since fingers have far less sweat glands than your armpits, I figured this might help throw off the GSR.
3. Day of, affixed a hand towel to a pair of underwear and put another pair on over them.  This was in case the guy had a pressure thing on the seat (he did) and since my preferred countermeasure is the anal sphincter clench.  Notice this measure presumes that the pressure seat thing actually can detect the sphincter clench which no evidence suggests it can.
4. Stay calm and lie like nobody has ever lied before!

Now, of course, I have no idea which of these things was necessary and which not but, in combination, they worked for me.  Truly, I believe that the most important thing was a wholehearted belief in the invalidity of a poly.  I had no fear of detection, no fear of 'what if'.  These machines are bullshit, and anybody with a decent background in science will be able to reach this conclusion.  I actually have a longer tirade about the theory behind them (type 1 and 2 errors, null hypothesis, the psychological construct of a 'lie', etc.), but that is not appropriate here.

Thanks for your time, and I am happy to answer any followup questions that have no way of identifying me further.  Oh, also, a big thanks to George W. Maschke and Gino J. Scalabrini for TLBTLD and their advice in the forums here.

The relative ease with which ABC beat the polygraph "test" speaks to just one of the many vulnerabilities of the deeply flawed polygraph process.

For some years now I have been calling on the American Polygraph Association to undertake a polygraph countermeasure challenge series, made integral to APA seminars, wherein randomly chosen polygraph operators would face a crew trained in countermeasures.

The APA has roundly condemned such an activity, with some polygraph establishment notables dismissing it as a "circus stunt."

No surprise there -- except avoiding a countermeasure challenge series suggests that the APA is actually running from science, not embracing science as they claim to be.

Why is an organization that professes to be "dedicated to the use of evidence-based scientific methods" running like hell from what would be the most telling experiment of them all?

Self-preservation.

In my opinion, such a countermeasure challenge series would plainly show that aptly applied countermeasures would be successful more often than not, rendering the "test" essentially invalid.

Given that polygraph is primarily about one thing -- money -- revealing the test's vulnerability to countermeasures would run counter to the industry's goals. 

In other words, it would be bad for bu$iness.

So, the polygraph validity charade goes on.

Meanwhile, absent a polygraph countermeasure challenge series, the APA's flattering claims of polygraph accuracy should, in my opinion, be viewed with a healthy dose of skepticism.