Long-tails--->Your view

Alexander Karatis · 05-16-2006, 11:54 PM

I know we have a couple of folks into this stuff...

What in your view is that which generates long-tail/pareto distributions?

Pretty broad-should make for some interesting discussion though...

Joseph Hart · 05-17-2006, 06:41 AM

Are we talking 6 Sigma here? The most defects that are found in a process? I am pretty sure I missed it.

Paul Findley · 05-17-2006, 08:48 AM

Questions like this are a good example, I suspect only a few PPM (people per million) will have a real quality response. I don't consider myself one of these elite few.

Alexander Karatis · 05-17-2006, 08:58 AM

No we are talking about power laws, that ubiquitous trend that whenever you leave a system free of contraints or regulation, it self organizes according to a very specific pattern that if allowed, produces a long-tail distribution.

Think nature, free-markets, amazon.com's contribution to sales by product code, wealth in a society....You name it.

I have written Chris Anderson of Wired fame with this same question, who is now writing a book on long-tails, and who pointed me toward this other book-->http://www.amazon.com/gp/luc/click.html/104-0122853-2891976?%5Fencoding=UTF8&id= SBBcEqffvUR5NiZvAc09CtTKbvZl8RZ6dngjmCY64kDsxzoz75 tOrdDP7beza8lsTU%2BQ1Epm6W1bHs ytdl8197HuivHn3BcGDFoZeqer3wdWD3QpaaGnCI%2BRD7BAgH sazYNz6V42unJjPDXfThoOIgc%2FUT RGKFSyeYHIjQxj%2BmzJ8BikLHOkv9SJSXq7%2Bhy%2FSCZqLE vekvar3lu3sqx0TruWdcotTH%2F%2B okHSJKJjnrc%3D

His blog is http://longtail.typepad.com/

Six Sigma I believe is based on power laws.

Robert Wolf · 05-17-2006, 12:41 PM

I like 'em:lol:

Eugene R. Allen · 05-17-2006, 01:30 PM

You know that gesture where you pass your hand over your head and don't even duck because whatever you are talking about has passed so far over your head it wasn't even close? This is like a satellite, in orbit, around Venus. I'm with Robb on this one. I just have to pretend you're talking about something else.

Paul Findley · 05-17-2006, 01:53 PM

Steal this thread...

I asked my friend Norm about this topic. He said he is just an average guy, and just to centered to even comment on the subject.

Sorry.

Don Stevenson · 05-17-2006, 04:53 PM

Six....sigma....pareto.....analysis....cold chills....creeping...up....my....spine

I used to work for a big multinational that was into all that stuff. Great theories but try getting a 60 year old fitter to "buy in"

However the best thing that company ever did for me was send me to some podunk, middle of nowhere town for a 2 week six sigma type course because i'd been chasing this cute girl and she just wanted to be friends, however after 2 weeks without me she was a lot keener and we are still together.

The course was rubbish though and I made lots of trouble and got a bad report from my boss for my efforts.

Joshua Newman · 05-17-2006, 10:06 PM

Chris was previously on the advisory board of the film company that I run, and I spend lots and lots of time thinking about the Long Tail and its application in the world of media.

But, oddly enough, I believe that Pareto distributions are also perhaps the best illustration of why CrossFit is so effective.

A traditional training approach - doing the same thing again and again - certainly causes you to improve at that thing, yet also yields ever diminishing returns in that improvement.

CrossFit, on the other hand, by varying workouts constantly, allows you to simultaneously follow multiple skill curves all at once, yet returns to benchmarks and builds from base movements sufficiently to ensure that you progress at least through the 'meatiest' part of each curve.

The essential CrossFit insight, then, is that for generalists the compounding of the beginning of a large number of curves yields far more than chasing further and further down the Long Tail of a smaller number of fitness skill curves in the same amount of time.

So, in short: get some, go again.

(Message edited by jnewman on May 17, 2006)

Alexander Karatis · 05-17-2006, 10:34 PM

Six Sigma, TQM...all that stuff definitely turns me off.

It usually happens this way-Management theories get ahold of a great general idea and kill them with "How tos" or "XXX Made Simple"... I swear, if I see "Sun Tzu on Business", or "Boyd applied to the Market" one more time, I'm going to scream.

But this just has too many implications to be avoided. And I can never get over the "why?" question if I can reach a state where I have a basic understanding of it.

In this case, the obvious first answer is that power law are ubiquitous in free systems because the are the more efficient means of distribution. There exist point like these in Game Theory matrices (like Prisonner's Dillema, or Rock Paper Scissors) where the payoffs of all players are best distributed to suit both the whole and each individual.

And then again, why is it the most efficient system? This also has profound implications for diet and fitness-(I believe DeVany has worked a lot with them).

In an evolutionary sense, we can obviously state that good adaptation (one that fits many different criteria-or is better "all-around") is what makes abundant species, and likewise rarity is a result of poor adaptation? But why have the long-tail anyway? Why didn't all the less than "good all-around" species go extinct, but instead remained rare? Are they a test pool that improves the whole group? Why have hits and niches in nature, and not just hits? I can understand that the ratio, when taken as a finite number is just that! (i.e. the most efficient).

I know that if I put all possibilities in a matrix, I'd end up at this column and row-so I'm not interested in that-explanations to that branch to the metaphysical or the plain dumb.

But what purpose has this law served? Why can't it be altered and produce better results?

BTW, Paul, I might have taken this the wrong way but...No offense...the whole "average guy" thing sounds kinda snark to me. No one is pretending to be more than just that here, so it serves no purpose.

BTW, if you guys find the topic inappropriate or just plain tiresome, we can always skip it, end it!:wink:

(Message edited by ApolloFS on May 17, 2006)

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
How long after being ill?	Mike Gray	Injuries	4	07-17-2007 07:37 AM
MT. View Fire workouts	Randy Nelson	Workout Logs	45	10-10-2006 08:17 AM
Running: how long is "long distance"?	Rachel Bennett	Fitness	16	06-01-2006 05:31 PM
What's the view with belts?	Seral Mehmet	Starting	18	02-06-2006 07:08 AM
Dan John's View on Stretching	Antti Akonniemi	Fitness	30	11-25-2005 05:59 AM

05-16-2006, 11:54 PM	#1
Alexander Karatis Member Profile: Join Date: Jun 2003 Location: Athens Attica, Greece Posts: 786	I know we have a couple of folks into this stuff... What in your view is that which generates long-tail/pareto distributions? Pretty broad-should make for some interesting discussion though...

05-17-2006, 06:41 AM	#2
Joseph Hart Member Profile: Join Date: Mar 2005 Location: Lino Lakes MN Posts: 281	Are we talking 6 Sigma here? The most defects that are found in a process? I am pretty sure I missed it.

05-17-2006, 08:48 AM	#3
Paul Findley Member Profile: Join Date: Jan 2006 Location: Longmont CO Posts: 1,258	Questions like this are a good example, I suspect only a few PPM (people per million) will have a real quality response. I don't consider myself one of these elite few.

05-17-2006, 08:58 AM	#4
Alexander Karatis Member Profile: Join Date: Jun 2003 Location: Athens Attica, Greece Posts: 786	No we are talking about power laws, that ubiquitous trend that whenever you leave a system free of contraints or regulation, it self organizes according to a very specific pattern that if allowed, produces a long-tail distribution. Think nature, free-markets, amazon.com's contribution to sales by product code, wealth in a society....You name it. I have written Chris Anderson of Wired fame with this same question, who is now writing a book on long-tails, and who pointed me toward this other book-->http://www.amazon.com/gp/luc/click.html/104-0122853-2891976?%5Fencoding=UTF8&id= SBBcEqffvUR5NiZvAc09CtTKbvZl8RZ6dngjmCY64kDsxzoz75 tOrdDP7beza8lsTU%2BQ1Epm6W1bHs ytdl8197HuivHn3BcGDFoZeqer3wdWD3QpaaGnCI%2BRD7BAgH sazYNz6V42unJjPDXfThoOIgc%2FUT RGKFSyeYHIjQxj%2BmzJ8BikLHOkv9SJSXq7%2Bhy%2FSCZqLE vekvar3lu3sqx0TruWdcotTH%2F%2B okHSJKJjnrc%3D His blog is http://longtail.typepad.com/ Six Sigma I believe is based on power laws.

05-17-2006, 12:41 PM	#5
Robert Wolf Member Profile: Join Date: Oct 2002 Location: Chico CA Posts: 2,670	I like 'em:lol:

05-17-2006, 01:30 PM	#6
Eugene R. Allen Affiliate Profile: Join Date: Sep 2004 Location: Tacoma Washington Posts: 1,713	You know that gesture where you pass your hand over your head and don't even duck because whatever you are talking about has passed so far over your head it wasn't even close? This is like a satellite, in orbit, around Venus. I'm with Robb on this one. I just have to pretend you're talking about something else.

05-17-2006, 01:53 PM	#7
Paul Findley Member Profile: Join Date: Jan 2006 Location: Longmont CO Posts: 1,258	Steal this thread... I asked my friend Norm about this topic. He said he is just an average guy, and just to centered to even comment on the subject. Sorry.

05-17-2006, 04:53 PM	#8
Don Stevenson Affiliate Profile: Join Date: Jan 2005 Location: Sydney NSW Posts: 1,038	Six....sigma....pareto.....analysis....cold chills....creeping...up....my....spine I used to work for a big multinational that was into all that stuff. Great theories but try getting a 60 year old fitter to "buy in" However the best thing that company ever did for me was send me to some podunk, middle of nowhere town for a 2 week six sigma type course because i'd been chasing this cute girl and she just wanted to be friends, however after 2 weeks without me she was a lot keener and we are still together. The course was rubbish though and I made lots of trouble and got a bad report from my boss for my efforts.

05-17-2006, 10:06 PM	#9
Joshua Newman Affiliate Profile: Join Date: Jan 2005 Location: New York New York Posts: 160	Chris was previously on the advisory board of the film company that I run, and I spend lots and lots of time thinking about the Long Tail and its application in the world of media. But, oddly enough, I believe that Pareto distributions are also perhaps the best illustration of why CrossFit is so effective. A traditional training approach - doing the same thing again and again - certainly causes you to improve at that thing, yet also yields ever diminishing returns in that improvement. CrossFit, on the other hand, by varying workouts constantly, allows you to simultaneously follow multiple skill curves all at once, yet returns to benchmarks and builds from base movements sufficiently to ensure that you progress at least through the 'meatiest' part of each curve. The essential CrossFit insight, then, is that for generalists the compounding of the beginning of a large number of curves yields far more than chasing further and further down the Long Tail of a smaller number of fitness skill curves in the same amount of time. So, in short: get some, go again. (Message edited by jnewman on May 17, 2006)

05-17-2006, 10:34 PM	#10
Alexander Karatis Member Profile: Join Date: Jun 2003 Location: Athens Attica, Greece Posts: 786	Six Sigma, TQM...all that stuff definitely turns me off. It usually happens this way-Management theories get ahold of a great general idea and kill them with "How tos" or "XXX Made Simple"... I swear, if I see "Sun Tzu on Business", or "Boyd applied to the Market" one more time, I'm going to scream. But this just has too many implications to be avoided. And I can never get over the "why?" question if I can reach a state where I have a basic understanding of it. In this case, the obvious first answer is that power law are ubiquitous in free systems because the are the more efficient means of distribution. There exist point like these in Game Theory matrices (like Prisonner's Dillema, or Rock Paper Scissors) where the payoffs of all players are best distributed to suit both the whole and each individual. And then again, why is it the most efficient system? This also has profound implications for diet and fitness-(I believe DeVany has worked a lot with them). In an evolutionary sense, we can obviously state that good adaptation (one that fits many different criteria-or is better "all-around") is what makes abundant species, and likewise rarity is a result of poor adaptation? But why have the long-tail anyway? Why didn't all the less than "good all-around" species go extinct, but instead remained rare? Are they a test pool that improves the whole group? Why have hits and niches in nature, and not just hits? I can understand that the ratio, when taken as a finite number is just that! (i.e. the most efficient). I know that if I put all possibilities in a matrix, I'd end up at this column and row-so I'm not interested in that-explanations to that branch to the metaphysical or the plain dumb. But what purpose has this law served? Why can't it be altered and produce better results? BTW, Paul, I might have taken this the wrong way but...No offense...the whole "average guy" thing sounds kinda snark to me. No one is pretending to be more than just that here, so it serves no purpose. BTW, if you guys find the topic inappropriate or just plain tiresome, we can always skip it, end it!:wink: (Message edited by ApolloFS on May 17, 2006)