previous post next post  

The power of analytics...

Or, why there were no half-size boots available when I joined the Army, thus causing life-long foot damage.

Why?  Because McNamara's Whiz Kidz said they weren't necessary and you could save a bundle of simoleons if you didn't make half-size boots.

Pejman Yousefzadeh said over on Google+ "Why doesn't the United Federation of Planets simply outlaw red uniforms?"

If you are a trekkie, this resonates.  Pejman was being funny.  But, remembering the possibly apocryphal story of the bright-eyed young OSHA operative in Germany in the early 80's who was appalled that the laser rangefinder on the M1 tank wasn't eye-safe... heedless of how unimportant flash blindness would be to the tanker who caught what would *follow* a successful ranging...  I was struck by a non-existent study result that I could, sadly, envision coming from the right kind of people...  like heedless bright-eyed analysts with no real experience in the field they are studying (We're ORSAs!  We can study anything to death!) producing a document that followed this thread...

"The greatest single thing military casualties in OIF and OEF, both combat and non-combat, have in common is some form of camouflage uniform.  Therefore, ban camouflage uniforms.  Additionally, many non-combat casualties occur while wearing PT clothes.  Ban PT clothes."

You laugh.  No, not going to come out of any competently run analysis shop.  But not all are competently run.  I've read stuff coming from college students about to graduate that would fall into the trap of the analysis above.  The total inability (or unwillingness, for reasons good and bad) to follow the threads of second- and third-order effects.

Like - a whole lot of political, scientific, and economic analysis put forth to underpin and further an agenda, vice actually study a problem. 

The debbil is in the details.   And computers allow you to bury people in meaningless data portrayed by badly done (or deliberately misleading) analysis.

Hold that thought.  More to come on that topic later.

11 Comments

The one single common factor in every single death ever is that the person who died, was born.

Therefore, being born is shown to have a 100% fatality rate, making it the deadliest activity known to man.

Also, all of those people who died were determined to have been inhaling oxygen up until the time of their death.  A study should be commissioned to look into whether banning the inhalation or other consumption of oxygen might lower the death rate.
 
Well, I *was* kinda tryin' ta stick with the "uniform" theme...    At least women and minorities are not over-represented in the ranks of the dead, if your sample has a great enough longitudinal axis.
 


   That's the problem with computers. People try to use them as a crutch, rather than a tool.

   We had a problem with civilian bean-counters regarding our squadrons. Seems that we couldn't make any attempt to save sonobuoys, JP-5, flares, or any other expendable item. If we did, the bean-counters would assume that we didn't need that many and would automaticlally reduce our allocation for the next quarter.

   Prior to the Carter administration, squadrons were able to "bank" unused flight hours, sonobuoys, etc for use when we deployed and operational tempos increased.  The Carter folks stopped that practice and, as a result, there were times when we simply couldn't operate because we were leggaly required to NOT go over our quarterly budgets.

  That reduced our ability to locate and track Soviet ballistic missilie submarines, to assist carrier battle groups by screening them against attack submarines trying to track  them, etc.  That made the game all the more famgerous and the stakes that much higher in case DefCon was raised or the balloon went up. 

  
 
Greetings:

I was fortunate enough to have a business professor who passed along a couple of good techniques for dealing with problems.

One , he referred to as the "strategic question", the question that boiled the problem down to its crux. For today's Postal Service demise, the question might be, "Why do we need mail delivery six days a week?"

The other was what he called "staged implementation", his alternative to "betting the ranch" and what used to be known as a "pilot program" in some squared circles.  Alternatively, my mother knew it as the "proof of the pudding". 

However, that old bugaboo, correlation versus causation, will always remain a favorite of those inclined to either foment a crisis or not let one go to waste.

 
Well, seeing as how I'm all opposed to women dying in combat and all, and given that wearing uniforms and PTs clearly puts them at greater risk of death.  I hereby propose that all those young, fit, healthy Army ladies go stark naked at all times.
 
Heh.  I think that qualifies as "inside the box" thinking, Mike.
 
 Heartless:  Your assumptive logic is totall  in error - those who died were emitting carbon dioxide, a known greenhouse gas and one hazardous to human life, therefore, to prevent further casualties, the EPA is issuing regulations banning exhaling.
 
MikeD - Be careful what you wish for...  There are people of both (all, now that we're post don't ask don't tell?) genders in the army that you do.not.want.to.see.naked.  Trust me on that...  :-)
 
In my time in the Army Pogue, I saw enough of both types to say with some confidence, that the "yeah I would" to "no I wouldn't" ratio still works in my favor amongst the ladies.  And I DID specify it was due to my objection of WOMEN dying in combat.  The guys will just need to take their chances and stay clothed.

And John, I like to think of it as more "inside the bunk" thinking.  ;)
 
This study recommends forty thousand years of study by our group or everyone will die tomorrow.
 
I am a data analyst. I look at millions of records, so I can easily discard the outliers to determine trends. Then somebody comes along, looks at one outlier, and says that the sky is green, or that the laser range finder is dangerous, and I have to spend the next 6 months defending my data.

Just issue everybody those stupid green goggles that you are supposed to use when running the LRF on the DSESTS and be done, or don't look at the laser.