What are the odds of 7 specific IRS hard drives crashing?

Here’s why I think there’s a cover-up happening in the IRS targeting scandal …

Lois Lerner is at the heart of the investigation.  She exercised her 5th amendment right to not self-incriminate and didn’t testify to Congress.

Then, it’s “discovered” that 10 days after the scandal inquiry kicked off, her PC hard drive – which contained the only permanent copies of emails she sent to folks outside the IRS – crashed … and all data – including the subpoenaed emails – was lost forever.

Hmmm.

 

 

Then it was revealed that the hard drives of 6 other IRS “persons of interest” had also crashed (after the investigation was kicked off.

Double hmmm.

Got me to wondering: What are the odds of that happening?

Let’s do some arithmetic …

 

 

======

First, a data point: What is the failure rate for a hard drive?

My guess was that 10% of all hard drives eventually fail.

But, since I don’t know, I did some quick digging.

A tech group called BackBlaze routinely life-tests hard drives … they keep tens of thousands of them swirling until they crash … so the sample has plenty of both old and new hard drives

Based on their latest data, about 4% of consumer quality hard drives fail annually … about  1 in 25.

• Tech note: I would assume that IRS PCs are  higher than consumer grade, so I’d expect the IRS hard drive failure rate to be lower.

Source

* * * * *

OK, let’s work the numbers …

What’s the likelihood that 7 specific IRS drives – those of the “persons of interest” – would crash in the same year?

Simple probability calculation: 4% (the failure rate) raised to the 7th power (since 7 specific people were involved) …  which works out to odds of about 1 in 6 billion.

Pretty long odds, right?

* * * * *

Let’s cut the IRS some slack ,,,

Take a worse case … the worst model’s annual failure rate in the sample is about 25%.

That gets the likelihood of the 7 hard drives crashing up to .006%  … odds of about  1 in 16,000

* * * * *

Let’s cut the IRS even more slack ….

Assume that the failure rate is 50% … about double the worst drive in BackBlaze’s sample.

Then, the probability of the 7 “of interest” hard drives crashing in .78% (50% raised to the 7th power)…. still less than 1% … with odds of about 1 in 128

* * * * *

Given the above analysis, I’d ask the IRS Commissioner two questions:

1. How many PCs does the IRS have in operation?
2. How many total hard drive crashes were reported in 2011?

The IRS has about 100,000 employees, I’d guess that the answer to question #1 is at least 50,000.

Above, we showed that a 50% failure rate would still make the likelihood of the 7 “of interest” hard drives crashing less than 1%.

And, for the numbers to square,  the IRS would have had to incur over 25,000 hard drive crashes per year.

Would have thought that SOMEBODY would have noticed that.

But, since there hasn’t been a reported epidemic of crashed hard drives at the IRS, I’m going to conclude that the industry average holds — a 4% failure rate — and that the tall tale that the IRS is peddling has odds of about 1 in 6 billion.

That is, by the numbers, why I think there’s a cover-up happening …

Q.E.D.

* * * * *
UPDATE

Believe it or not, I watched Monday night’s hearings on C-SPAN … after completing the above post.

Rep. Thomas Massie (R-KY) started to pursue a line of questioning paralleling the above.

The IRS Commissioner said: 90,000 employees; 2,000 hard drive crashes per year; “comparable to industry average of 4-5% hard drive failures per year”

OK …  that gets us back to my original calculations:

What’s the likelihood that 7 specific IRS drives – those of the “persons of interest” – would crash in the same year?

Simple probability calculation: 4% (the failure rate) raised to the 7th power (since 7 specific people were involved)

…  which works out to odds of about 1 in 6 billion.

Better chance one of us will get hit by lightning today

#HomaFiles

Follow on Twitter @KenHoma            >> Latest Posts