New Yorker article about probability theory. How can we apply it to Y2k? Was started in prev thread (Poll)

greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread

"If anyone actually scrolled down this far you should be rewarded so here goes. Seriously, Last week's New Yorker has an article about a guy (Princeton professor dude) who predicts the probability of things by assuming based on his mathematical calculations there is only a 5% probability we are living in extrordinary times. Since extrordinary would include depression, fall of US etc that means there is only a 5% chance the last 3 will happen. He has never been wrong, although there is a 5% chance he will be. Read the article"

-- Jimmy Bagga Doughnuts (jim1bets@worldnet.att.net), July 15, 1999.

Jimmy Bagga I did read the article and it was fascinating. However it all comes down to your definition of "extraordinary times." Y2k is not really about extraordinary times from a human evolutionary/historical vantage point. It is an extraordinary time in the evolution of computers. How long have we had computers involved in every part of human commerce? Maybe 15 - 20 years? Even if you start from where Y2k was hatched - back in the dark ages of computing - maybe 30-35 years ago? Then the probabilities look different. Starting from a time line of only 35 years we could easily be within the 5% timeframe of an extraordinary event in the history of computing. This is a good point to bring up though. I am tempted to start it as a new thread.

-- R (riversoma@aol.com), July 15, 1999.

And so I did. Has anyone else read the article? Anyone care to do some number crunching? I will go get the "professor dude's" formula and post it so y'all can ponder its deeper meanings.

-- R (riversoma@aol.com), July 15, 1999

Answers

Song: Westside ~~ 1998 ~~ 5:02 Artist: TQ LP/CD: They Never Saw Me Coming ~~ Clock Work ~~ 1998 Writers: Tony Quaites~Mike Mosley~Femi Ojetunde~J. Sample~Robert Ford~Russell Simmons~J.B. Moore~Kurtis Walker~Lawrence Smith

This is goin out To the whole wide Westside Know what Im sayin? Yeah Break it down for me Steady Mobbin (Rock) rock on

I was just a young boy livin in the Hub City Eastside Compton, G Back in the days when Ice Cube and Eazy Had every nigga talkin bout, Boy, you cant f@#$ with me Remember Ice-T had the power (Ooh, wee) Hearin gunshot lickin by the hour When Too $hort bumped every super sport And taught us all how to ride for the West Coast

To my peoples, if you wit me, where you at, throw your dubs in the air And wave em like you just dont care From L.A. to the Bay, what you say All day, every day, any damn day Take a look around, we got the whole world locked down Goin on and on, it dont stop, wont stop, cant stop the rideda Thats why I put it down for the Westside

Didnt seem that long ago (Oh, no) We were stealin forties out the liquor store (Oh, no) Mama started trippin, so its time to go Mobbed to the park with the locsters Everybody broke, smokin roaches DJ Quik was the s@#$ Had every nigga claimin he was from the CPT Yall remember One time tried to clown We had to burn this b@#$% on down

One time for my niggas in incarceration I blaze a dime wit you for havin lots of patience Two times for my sisters at the county building I got some Westside love for all you ghetto children Three times for my niggas that done passed away I tip some gin for you and pray for better days One day everythings gonna be fine But until that day my only reply Is Westside till I die

To my peoples, if you wit me, where you at, throw your dubs in the air And wave em like you just dont care From L.A. to the Bay, what you say All day, every day, any damn day Take a look around, we got the whole world locked down Goin on and on, it dont stop, wont stop, cant stop, you could rideda Got to give it up for the Westside

From Crenshaw to Foothill Boulevard All I see is lowlows in mob cars, Daytons and Five Stars All the true h@#s is hittin switches with all the finest riches Of course the bombest b@#$%es, dont get it twisted Ask me whats a real G Show me a nigga scrappin back against the wall until his knuckles bleed (Knuckles bleed) Screamin death to all our enemies and those who dont believe West Coast livin be the s@#$ to me

One time for my niggas in incarceration (Ooh, yeah) I blaze a dime with you for havin lots of patience Two times for my sisters at the county building I got some love for all you little ghetto children Three times for my niggas that done passed away I tip some gin for you and pray for better days One day everythings gonna be fine But until that day my only reply Is Westside till I die

To my peoples, if you wit me, where you at, throw your dubs in the air Wave em like you just dont care From L.A. to the Bay, what you say All day (All day), every day (Every day), any damn day (Any damn day) Take a look around, we got the whole world locked down Goin on and on (Goin on and on), it dont stop (And on and on) Wont stop, cant stop, you could rideda You got to give it up for the Westside

To my peoples, if you wit me, where you at (Whoawhoawhoa) Throw your dubs in the air, wave em like you just dont care From L.A. to the Bay, what you say All day, every day (Every day), any damn day (Any damn day) Take a look around, we got the whole world locked down Goin on and on (Goin on and on), it dont stop (And on and on) Wont stop (And on and on), cant stop, you could rideda La, da, da, da, da, da, da, da, da

To my peoples, if you wit me, where you at, throw your dubs in the air Wave em like you just dont care From L.A. to the Bay, what you say All day (All day), every day (Every day), any damn day (Any damn day) Take a look around, we got the whole world locked down Goin on and on, it dont stop, wont stop, cant stop, you could rideda TQ gon give it up for the Westside

To my peoples, if you wit me, where you at, throw your dubs in the air Wave em like you just dont care From L.A. to the Bay, what you say All day, every day, any damn day Take a look around, we got the whole world locked down Goin on and on, it dont stop, wont stop, cant stop, you could rideda Got to give it up for the Westside

Dedicated to Eric Eazy-E Wright, Tupac Shakur Rest in peace, homies

Transcribed by: Char Star Uploaded on: September 28, 1998

-- THe phantom (butnit@hotmail.com), July 15, 1999.


The prof's name is J.Richard Gott III and he is a physicist. He correctly predicted the fall of the Berlin wall along with several other events. There is definitley merit to his theory.

Basicly he has taken Copernicas' theory about space - we are not in the center of the universe. Our location as human beings is not particularly special in the way that religious traditions of the time thought it was. Gott has applied this Copernican concept of non- special space and applied it to time. The odds are that whatever moment we happen to be in is not particularly special. Otherwise the whole concept of "special" or "extraordinary" would have no meaning.

When Gott applied this theory to the survival of the species he easily determined that this was probably not a special time and predicts many millenia for the duration of Home Sapiens.

I also need to point out that Gott's predictions are extremely broad.

Basicly you take the number of years since your starting date, divide that by 39 for your low and multiply it by 39 for your high.

So if you want to look at computers starting from main frames you will get one answer. But if you want to look at micro chips or Windows or any operating system then things begin to look a tad bleaker. Especially when you consider how FAST computers evolve. Is a pentium the same species as a 386?

I am not sure enough of certain dates to determine the probabilities myself but if anyone wants to feed me dates I will crunch the numbers.

At what point did power plants become computerized? When did micro chips become ubiquitous? When did ATMs take over banking? How long has the gold standard been history?

There is obviously much more to the article along with an explanation for the formula but I am not going to type the whole thing here. I recommend it highly. it is in The New Yorker, July 12, 1999. It is titled "How To Predict Everything" and it is written (extremely well) by Timothy Ferris.

-- R (riversoma@aol.com), July 15, 1999.


I read the article. The "39" comes from the typical statistical confidence interval of 95%, leaving 2.5% on each end.

An application, for those predicting the "fall of the US Government" in the recent poll.

Take 1776 as the starting point. The US has existed for 223 years.

Using the method, there is a 5% chance we live in "extraordinary times"; i.e., either at the very beginning of the life-span of the US, or at the very end. Divide by 39 to get the lower interval, or 5.75 years. Multiply by 39 to get the upper interval, or 8697 years. So, with 95% confidence, we can say that the US will last for between 5.75 and 8697 more years.

It was an interesting article. Not too hard to see how he "correctly" predicted some things, though.

-- Hoffmeister (hoff_meister@my-deja.com), July 15, 1999.


Here's a link to basically the same article:

http://ww w.mit.edu/people/belg4mit/services/grim.html

-- Hoffmeister (hoff_meister@my-deja.com), July 15, 1999.


Nuh-uh, every day is "extraordinary times". Last year was extraordinary, this past decade was extraordinary. Next year will be quite extraordinary, too, and next decade as well.

Issue a patently subjective premise as if fact, then support it with facile, made-to-order argument. Weak.

-- Nathan (mospam@all.com), July 15, 1999.



Try this - 1992 - Social Security Administration begins remedaition planing, most actually programming (work) starts in 1995 1994 - beginning of year 2000 remediation (a few companies.) A few people start writing. 1995 - pundits publish. budgets begin (many companies/some companies - take your pick) White House is warned! 1996 - most work begins - for those who are finishing now! 1997 - some/most/the rest begin - none finish 1998 - one agency finshed (Social Security - in Dec) 1999 - nobody else is finsihed (St Lawerence Seaway the exception?)

Therefore, years of program completion = 1 (that's stretching it!) So probability of 2000 being an exceptional year = 99% - that's past the 2.5% high and low, but you get the point.

Year since Y2K repairs started = 1999.5-1993 = 5.5 years. (Some would say 4 years, crediting the real start in 1995, others could say 3 years as the real beginning and not be wrong.)

Now, dear mathematicians: try 3 years, 4 years, and 5.5 years in your calculator.

-- Robert A Cook, PE (Kennesaw, GA) (cook.r@csaatl.com), July 16, 1999.


Upon some reflection I think that applying Gotts theory to Y2k isn't going to work because unlike the Berlin Wall or the run of a Broadway play, Y2k has an expiration date. Any fixed significant date makes his formula useless. So while it is an interesting intellectual exercise for other trends I think we can dismiss at as any kind of way to determine whether or not we are living in significant times (per Y2k).

We are living in significant times because this is a date specific event. Just as February 5th will ALWAYS be a significant date for me since it is my birthday.

-- R (riversoma@aol.com), July 16, 1999.


Robert you forget the assumption, "If there is nothing special about your observation of something, then there is a 95 per cent chance that you are seeing it during the middle 95 per cent of its observable lifetime" There is something special about your observations and the 95% assumption doesn't apply. As an aside, the 95% comes from a normal (0,1) distribution.

-- Maria (anon@ymous.com), July 16, 1999.

Well Maria - that's what I was trying to show - the "concept" that y2k-related events can be assigned to "any single year" is not really a reasonable assumption, given that the whole thing "grew" in less than 5 year period (1995-2000), and it appears that most of the resulting (critical) errors (or lack of errors!) will occur in only a matter of a few weeks in one quarter: Jan, Feb, and Mar of 2000.

So, the critical (or interesting time frames) are on the matter of weeks or months - not years. Also, it is readily apparent that the vast majority of "observeable" events will either occur close to the first weeks of year 2000 itself, or as an immediate result of events falling in the first few weeks.

Fior example: look at social problems - if they have not occurred in the first few weeks (2-3) they probably will not occur at all.

On the other hand, if there is immediate loss of services in several areas at once - and these services are not IMMEDIATELY restored (or martial law declared!) or are not restored within hours of failure, then then will be long-lasting effects (probably tragic effects) - but the proximate cause was the first few hours of failure leading to panic from an un-prepared, mis-informed majority.

If there is only limited failure in very few areas, and these areas are so widespread that they can be "physically" and "informationally" isolated - then the vast majority of the country will not see failures, and so will not respond.

Again - the failure (or "lack of failure" in this case) as an event occurred in a very short time: if no failures have been publicized (whether they have actually occurred or not) over most of the country after a few days, then most people will believe y2k is a non-event. To them personally.

___

Now, business and economic impacts are much different! I forsee NO business (or government agency) actually failing in only a matter of hours - the DOD, FAA, and a few others are the exception - their failures may be public, or may be concealed, but they will be unique amongst most agencies.

Government failure will be slow and invasive: no permits issued, no emmissions stickers, no tickets issued, tax payments lost, no insurance coverage, wrong checks transferred or issued, no payroll taxes accepted, funds transfer between agaencies or between units failing at one end or the other. VA records "gone" - or medical tranactions simple "lost". IRS records not accessible, etc.

So - would a typical user notice immediately? Would a business go "out of business" immediately? No. It would take several days/weeks to lose enough business and supplies and venders to and shipping losses to shut everything down. Walmart even has merchandise, it just could not restock its shelves from China any more - to empty them? More than a few days - the effect is there, but is spread out.

When would layoff's and cuts become apparent to secondary businesses? A few more days absolutely, a few more weeks most likely. Still short compared to the years of this "expert's" theory.

-- Robert A Cook, PE (Kennesaw, GA) (cook.r@csaatl.com), July 16, 1999.


Moderation questions? read the FAQ