I propose a survey. Gary North has posted a link to the Y2K Circle of Dominoes Engine at y2knewswire.com. This web-based software gives you the opportunity to draw your own conclusions on what will happen with Y2K, based on your own assumptions. With many people saying Y2K is "no big deal" and others predicting a total collapse, wouldn't it nice to run your own numbers for a change?

For those who would like to, please post here the results you obtain from the engine. It should be interesting to see the range of results we get, and if they exhibit a typical bell curve distribution.

My results: 100 percent chance of total collapse.

-- Prometheus (fire@for.man), August 04, 1999

Answers

87.5 with very conservative input.

-- rambo (rambo@thewoods.com), August 04, 1999.

58.5% with conservative input

-- Bob (bob@bob.bob), August 04, 1999.

Even answering as low as 1% to each and every question to this "engine" it still came up with a 24.5% chance of failure.

I think this is a faulty "engine".

Seen this sort of thing before with the cash calculator on some other site. Looks like loaded dice to me. We dont need this sort of crap confusing people even more.

-- Bean Counter (beancounter@beano.com), August 04, 1999.

68.3% with extraordinarily conservative input. Thought I'd see what the percentage would be if I acted like I believed the media/government reports on y2k.

-- Libby Alexander (libbyalex@aol.com), August 04, 1999.

Hmm. 24.5% with two "5" and "1" and "0" for the rest. impact was set at "5".

I doubt the overall system is so easily modeled.

Keep your...

-- eyes_open (best@wishes.net), August 04, 1999.

I looked at the engine, but I don't buy its model. As an alternative, take a look at

Y2k Dependency Map

Techno-Economic Engine

-- bw (home@puget.sound), August 04, 1999.

94 % for total..........

-- kevin (innxxs@yahoo.com), August 04, 1999.

The Y2K Dependency Map is flawed. It is far too simplistic. Take a look at the "Hospitals" rectangle near the bottom. Are we to believe that banking and Medicare are the only things affecting hospitals? What about the availability of electricity? How about the availability of pharmaceuticals, or disposable medical supplies? If the streets are unsafe, how many hospital staff will choose to stay home to protect their families?

Or take a look at the "Farming" rectangle. Are we to believe that farming's only dependencies are fuel stations and agrichemicals? How will Cargill or Monsanto get the hybrid seeds for next spring's crops to the farmers all over the country if the railroads are clogged because of switching errors, or if the railroad can't get diesel fuel? How will farmers get replacement parts for their tractors and field equipment if John Deere shuts down because two percent of its thousands of suppliers have gone belly up? How will farmers be paid for their produce if the banking clearinghouse system has crashed, or if bank runs have shut the banks down?

The problem with the Y2K Dependency Map fails to show that, in one way or another, every industry relies on all the other industries. The Circle of Dominoes Engine does a better job of taking the systemic nature of Y2K into account. When the Dominoes Engine provides frightening results, you can either claim it's broken or rigged (shooting the messenger), or you can re-evaluate your assumptions about how interconnected our industries are. Saying "I don't like these results, therefore it's wrong" is the same thinking that makes people wrongly reject Y2K in the first place: "The breakdown of society is too terrible for me to handle emotionally, therefore it won't happen. Bill Gates or the federal government will fix it."

-- Prometheus (fire@for.man), August 04, 1999.

Erratum:

The problem with the Y2K Dependency Map is that it fails to show, that in one way or another, every industry relies on all the other industries.

-- Prometheus (fire@for.man), August 04, 1999.

99.4

-- sarah (qubr@aol.com), August 04, 1999.

3% chance of total collapse.

Many of you need to read the questions again. All of the questions were stated at the ability of one small disruption to cause total collapse. It is no wonder that your numbers are so high. If, in the questions, any of the items have the ability to cause Total Collapse then they would multiply against the numbers to give you an artificially high result. Example, the ability of one solar flare to stop all communication world wide is really low. It is a lot less than 1%. Read the questions again with a new view.

The engine is probably a good guess of the real possibility of total collapse. A total collapse with what has been repaired to date will not happen. Also a localized power disruption(which everyone believes is a possibility) is not a total system collapse. The same is true for the phone system. Also, the trucks(equipment) not being able to start in a coal mine will not stop the rest of the industry from producing coal. Not all of our oil comes from the mid east region, either(mideast termoil). They did not say anything about Texas, Alaska, Mex., Ven. ect... which all product large quantities of oil for the US.

Re run your numbers again and re-evaluated the small items causing "total" collapse. This will change your numbers

-- Ned P Zimmer (ned@nednet.com), August 04, 1999.

I must be getting more optimistic. A few months ago, I was at 98% Today I was clear down to 86%.

By 12/31, I'll probably think there's an 80% chance of going Milne!

-- Dog Gone (layinglow@rollover.now), August 04, 1999.

27.2% -- That was putting 1% in the "total collapse" questions (and most of the others as well). Biggest single input was 10% for cash woes at banks... which I consider conservative.

-- M.C. Hicks (mhicks@greenwich.com), August 04, 1999.

It is simply a lesson in statistics. Certainly the choice of "Achilles Heels" is flawed to the extent that there are redundancies (which amplify the gloom) and many omitted factors (which attenuate the gloom). However, what it shows to those who struggle with probabilities is that if there are alot of ways we can be screwed by y2k, the chances of getting screwed gets really high. If you can personally identify 20 different (independent) ways that there is 1% chance of catastrophe striking, then there is (wait while I calculate this...) 18.3% chance of a catastrophe. If there are 100 low (1%) probability ways that catastrophe can strike, then the overall probability goes to 83.4%. Given all the ways we can be screwed (IRS, FAA, rails, meltdowns, currency collapses, Saudi desalination collapses, bank runs, and countless other infrastructure and global issues (with probabilities that may be higher than 1%), the simple statistics paint a compelling picture.

-- Dave (aaa@aaa.com), August 04, 1999.

Given all the ways we can be screwed (IRS, FAA, rails, meltdowns, currency collapses, Saudi desalination collapses, bank runs, and countless other infrastructure and global issues (with probabilities that may be higher than 1%), the simple statistics paint a compelling picture.

Exactly. This is why the engine works. Read the explanation they give and you'll understand. With so many points of failure, it's practically a given that the system will collapse.

-- (its@coming.soon), August 04, 1999.

I have seen this engine now for quite a while and I haven't seen any legitimate debunking of the basic theory that underlies it all. The problem here, for anyone that takes the time to seriously input some probability guesses is not that it comes out hard to understand, but rather that it comes out hard to *believe.* Is that such a surprise? Most of the population doesn't believe all this interconnected danger business, without even trying to put some thought and numbers to it.

-- Gordon (gpconnolly@aol.com), August 04, 1999.

Gordon,

Exactly. The very point I was trying to make above.

Zimmer,

Maybe you need to reread the questions. You stated, "All of the questions were stated [as] the ability of one small disruption to cause total collapse." Not so. The instructions specifically ask you to "assign a percentage chance to each of the following events occuring within 30 days of January 1, 2000." Nothing more to it than that -- what is the chance of each event happening in January, 2000.

The instructions do not state anything like "assign a percentage chance that event X will cause total collapse." The only time "total collapse" is even mentioned is after you have assigned the values, and the algorithm does its work. This is textbook probability theory at work, and although the Domino Engine is not overly sophisticated, it's fundamentally valid.

-- Prometheus (fire@for.man), August 04, 1999.

Prometheus,

You need to do a little more careful reading, too.

Note the "iron triangle" comment: "Virtually all entities require these four, even if links between them are not explicitly shown".

Of COURSE it's simplistic! It's only one crummy sheet of paper, for gosh sakes.

-- bw (home@puget.sound), August 04, 1999.

Any "domino" model is utterly flawed. It concentrates on the multiplicative effects (A causes B and C and D) and ignores network redundancy effects (for Z to fail requires that W, X and Y all fail at once). A correct analysis would include both, except that you rapidly discover that it's a monster problem. And then you'd have to include time factors, like exhaustion of stocks (negative) and things being mended or bypassed by human adaptation (positive).

Network analysis is a complex, largely unsolved problem, even for fully deterministic things like telecomms. Economics is "the dismal science" because prediction of the behaviour of a network that has human beings as its principals is probably impossible.

-- Nigel Arnot (nra@maxwell.ph.kcl.ac.uk), August 05, 1999.

The value of both the domino site and the dependency map are that they help people think about the problem. Two drawbacks of the domino site are that you have to have web access and go to that site to run it, and the logic is a black box so you don't understand what the engine is doing. The dependency map lets you juggle these variables in your head, and helps you envision the links. Neither one is even close to a full picture.

By the time you get to the domino site, you already get it. The dependency map (and the back side of that page, linked above), is good for DGIs, because it requires no PC, no web address. I can hand this out to a couple dozen PC-phobics in a church meeting room, and within 30 seconds they can see that we are in serious trouble.

The Techno-Economic Engine (second link above) DOES take into account the time factor and the community attitude, both before the rollover and afterward.

I think the human brain can absorb much more information, and more complex linkages, than we are able to represent in any one medium. Both the domino site and the dependency map can really do no more than give a skeleton for those links, and let people go on from there.

-- bw (home@puget.sound), August 05, 1999.