For whom the bell curve tolls
¸ 1999 Michael S. Hyatt
A year ago I co-authored an article with my associate, Bill Dunn, which discussed Bell Curve theory and Y2K. We suggested that the completion of Millennium Bug repair projects should follow a bell curve distribution pattern. Similar to a roomful of students taking an exam, there should be a trickle of early-finishing high achievers, followed by a growing wave which crests at the top of the bell, and finally tapering off until only the under-achieving stragglers remain.
If most businesses and government agencies could complete their Y2K projects before the end of 1999, surely our economy and society would not be devastated by the handful of straggling, under-achieving organizations. The key would be whether the bell curve crested before the Year 2000 or after.
In October 1998, with 15 months to go until the new century would arrive, we expressed concern at the absence of any major organizations which had reached Y2K compliance -- despite many firms claiming that they had begun working in earnest as early as 1995 or 1996. If the trickle of early-finishing high achievers had yet to begin, we speculated that the bell curve would not crest in mid-1999 -- as public relations spokespersons kept assuring us -- but instead would crest after the immovable deadline of Dec. 31, 1999, with disastrous results.
In other words, a year ago we were sure that time was going to run out. Unfortunately, nothing during the last twelve months has convinced us to change our view.
In fact, the situation has gotten worse. The results of a new survey were released on Sept. 30, 1999. This survey, called the Y2K Experts Poll, was conducted by CIO magazine, the Information Systems Audit and Control Association (ISACA), and Dr. Ed Yardeni's Y2K Center.
The poll results are alarming:
Fully 81 percent of large, global companies are still not Y2K ready. The trickle of high achievers has finally begun, but the cresting wave is nowhere in sight.
Forty-three percent have pushed their project completion dates from the third quarter of 1999 to the fourth quarter of 1999 (and, I wonder, when Dec. 31 comes and goes, will they move their completion dates to the fifth quarter of 1999, and then the sixth quarter of 1999?).
Twelve percent of large, global corporations admitted that they now expect to finish their Year 2000 repair projects in the Year 2000 or beyond.
Twenty-three percent are still waiting for Y2K compliant mission-critical software from third-party suppliers. (What do they think, that software is like a toaster: you take it out of the box, plug it in, and it works perfectly?)
Despite all this, 91 percent of poll respondents expressed optimism about the Y2K issue. (Reminds me of the optimism expressed by George McGovern and Michael Dukakis just before their respective election days.)
According to Dr. Yardeni, the chief economist at Deutsche Bank Securities and a noted Y2K expert, "I am puzzled that information technology professionals are so optimistic about the impact of Y2K on their organizations. Many of them are not ready yet. They are clearly expecting a victorious outcome, which may be raising complacency levels so high that people will not prepare for any possible malfunctions and failures."
Abbie Lundberg, the editor-in-chief of CIO magazine, said that she is "alarmed at the number of companies that do not expect to complete their Y2K work before the end of the year. Of those that do expect to finish in time, entirely too many expect to do so at the eleventh hour. Given the business world's abysmal track record for technology project delivery, this does not bode well."
With less than 90 days remaining until Jan. 1, those people who understand that four out of five organizations are not yet done with their Y2K projects, but still insist that that most projects will be finished in time, are asking us to believe the bell curve distribution will look something like the Washington Monument. No trickle, no slow acceleration to a crest, no gradual tapering off. In other words, no bell-shaped curve.
Instead, we are told to expect a sudden spike. But for this spike to happen, the vast majority of organizations -- regardless of size, type, or when they started working on the problem -- must suddenly finish their Y2K repair projects within weeks of each other.
I submit that this notion is untenable, unrealistic, and statistically impossible.
Bell Curve theory tells us that the distribution of completion dates should look like, well, like a bell -- certainly not a sudden spike. If, after many years of working on Y2K repairs, only 19 percent of large, global companies have completed the job with less than 90 days to go, I think it's very safe to say the bell curve crest will not occur during 1999.
A year ago we wrote, "So what can we conclude? If no major outfit has yet been able to complete (the job), if the trickle of high-achievers has not yet begun, it is obvious that the bell curve will not crest in mid-1999. It will instead crest well after January 2000 -- which means our economy and our prosperous way of life are in for some serious, possibly life-threatening, disruptions."
Twelve months ago, I expected that Y2K project deadlines would be routinely missed, especially given the time-consuming and tedious nature of software repair work. This has indeed come to pass.
However, 12 months ago, I did not expect corporate and government leaders to present such a concerted and relentless campaign of public relations optimism. This happy-faced spin campaign is, as Dr. Yardeni noted, "raising complacency levels so high that people will not prepare for any possible malfunctions and failures."
It's not too late to take some action to prepare for possible Y2K-related disruptions. Optimistic PR statements will count for nothing if the essential products and services of our modern society are in chaos in a few short months.
Let me cite one final item from the new Y2K Experts Poll: although 91 percent of respondents expressed optimism about the Y2K issue, 56 percent of these corporate leaders are personally stockpiling food. As I've said before, forget what they are saying. Watch what they are doing.
worldnetdaily.com
ml
*****
Forgotten Y2K Math: Dependencies
October 7, 1999
In an age of New Math, where feeling good about the answer is more important than getting the right answer, it's no surprise that almost nobody understands the statistical math behind the possibility of Y2K disruptions.
Beyond the dumbed-down-math agendas in public re-education camps... er, we mean public schools, there's also another reason this is overlooked: it's counterintuitive. Few people, when given the numbers, come anywhere close to the right answer.
Test yourself! If Company A has 200 mission-critical systems, and each mission-critical system has a 3% chance of failing on January 1, what is the overall chance of the entire company failing on January 1?
Stop reading and actually ponder this one for a minute. What's your guess on the answer?
Few people arrive at the correct answer: a 99.8% chance of total failure. That means choice "D" -- 100% -- is the correct (rounded) answer.
IT'S NOT OBVIOUS, BUT IT'S TRUE
Even though these answers are not obvious, they are undeniably real. Unfortunately, pointing out the answers (and the math) to most people does little good: they don't accept it because they don't understand it.
For a full description and a handy chart that shows you how this math works, click here for the "Chance of Failure" chart. We built it just for you. You can use the chart to look up the chance of failure in any multi-entity system.
Unfortunately, nobody in Washington will use this mathematical truth to describe Y2K risks: it sounds too pessimistic. But if the President's Year 2000 Council were actually working to assess the current Y2K situation, they would only need to ask two questions of each agency:
Question #1: How many mission-critical systems do you have?
Question #2: What is the chance of each system failing on January 1?
Then they only need to check our chart.
Agencies that respond by saying they have one hundred mission-critical systems and each system has only a one percent chance of failure would seemingly be reporting good news, right? But in fact, this is terrible news because it indicates a 63.4% chance of total failure.
This is precisely how so-called "good news" isn't really that good once you apply logic to the data. For a government agency to be admitting it has a 63.4% chance of outright failure is hardly comforting.
BUT IT'S ACTUALLY MUCH WORSE THAN THAT
Many agencies have far more than a hundred mission-critical systems. Government-wide, the number is reported at approximately 6,000. (It used to be 9,000+ systems before 3,000 systems were taken off the list to boost the appearance of "progress.") What are the odds of all 6,000 mission-critical systems working on January 1? Virtually nill.
But the big numbers start to come into play when considering third-party vendors. How many vendors does Ford rely on to build cars? Estimates have put this as high as 20,000. Obviously, a miracle would have to occur for Ford to not experience supply-line failures. The question, for Ford, is one of making sure the really important vendors will work -- and having contingency plans for the expected failures of the other vendors. For the legal record, we're mentioning Ford only as an example here; we don't know the actual Y2K compliance status of Ford's vendors.
CALCULATING THE NEED FOR PREPAREDNESS
These same mathematics are useful for demonstrating the need to prepare with extra food, water, cash, medicine, heating supplies, and so on. Here's why:
If you knew a nuclear missile was going to strike the United States, but you didn't know where, would you be wise to prepare for it? Would the rest of the country be wise to prepare for it? Or should everybody do nothing because nobody knows where it will hit?
Now consider water. There are over one thousand water treatment facilities in the United States. If each one has a 1% chance of failing, what is the overall chance that at least one facility will fail? The answer, of course, is 100%. (Technically, 99.9957%.) That means somebody will be without water in this country. But none of us knows which area will be hit. We also don't know the duration of the failure in question.
Given this, is it wise to prepare? Remember, there is a near-certain chance that some municipality in the country will be without water for some period of time. But who will it be? And what will be the duration of the failure?
See, there are plenty of knowns and unkowns with Y2K, but not in the way you've been told by others. The unknown is who will be hit. The known factor is that somebody will be hit. The rest is all debatable: should an entire country stockpile extra water to prepare for a near-certain water failure in one city?
-continued at:
y2knewswire.com |
