Forum Index » General Lightweight Backpacking Discussion » Extraordinary Claims - Extraordinary Proofs

Display Avatars Sort By:
James D Buch
(rocketman) - F

Locale: Midwest
Extraordinary Claims - Extraordinary Proofs on 10/13/2010 18:41:55 MDT Print View

Extraordinary Claims And Extraordinary Proofs

This forum and website are literally based on an extraordinary claim - "UltraLight Backpacking is possible, doable and safe, besides being fun".

The extraordinary proof is the practical experience of thousands of members and followers of this belief chalking up hundreds of through hikes on the major long trails of the United States and elsewhere in the world. It is additionally backed up by numerous published books and hundreds, perhaps, of websites of practitioners of the art offering to share their knowledge and experience.

Of course, there are really tons of doubters out there. They often don't read, and if the read they don't believe, because they have been doing backpacking the old way and simply refuse to accept that anything else could be as "right" as they have been doing. *

Extraordinary Claim of "Sweet Spot" in Compaction of Down

There has been an extraordinary claim of a sweet spot in the compaction of down where there is a "maximal efficiency" down density for warmth which is about 2.5X higher density (250% compaction) than fully lofted high loft down.

There has been no extraordinary proof.

Let me repeat that, there has been no extraordinary proof of a clear and unmistakable kind. This is following a year or two since the announcement of the "maximal efficiency" packing of down.

The data leading up to that claim has been of a sometimes fuzzy or confusing nature. Never yet have we seen a laboratory curve of the 800 fill down (or similar very high loft down) as a function of compacted density from fluffy to nearly squashed. (And plotted so that we can read off the numerical values and do reasonably accurate calculations.)

In the announcement of this alleged phenomena, the original poster (OP), cited some NIST data said to be down which showed the fundamental decrease of thermal conductivity as a function of density with 7 samples.

Then the OP showed a plot of about 550 to 600 fill down from a Natick 1990 technical report with just 4 down data points, one of which (about 2.2 lbs/ft^3) was the "Max Loft" and the second data point (at 5 lbs/ft^3) was labeled "Max Efficiency"

We will look at these initial leading arguments for the extraordinary claim in detail.


The NIST data are on the NIST website in the form of a database for CONSTRUCTION MATERIALS. The materials shown were all called "feathers" as is seen in the first two figures below. The first figure is nearly impossible to read but it was presented as evidence and the second is after a 3 minutes work to make it more readable using GIMP, a free graphics program. The "Feather" label is visible.

2 figures
Starting Figure

Enhanced Figure

Both plots are "hidden zero" type plots, that is they are exaggerated in plotting scales to blow up or magnify any small variations or trends there may be in the data points. The next figure shows an Excel plot of the data above with the zero of density and conductivity included.

Proper Plot WITH Zeros on the Axes

Now you can see that the large downward trend of decrease of thermal conductivity with density that was claimed is not really a major factor at all. There is no way that this plot should have ever been considered as extraordinary proof of an extraordinary claim. There is no reason for anyone to consider it as such. This plot was a classical example of unfortunate magnification of small differences – and it appears built into the NIST database.

Natick Data on Down (Natick TR/tr90/030 1990)

Thanks to Dr Philip Gibson, I obtained a copy of the Natick TR/Natick/TR90/030 of which he was the prime author. I have extracted the original figure which is shown along with the figure that the Original Poster presented. (below)

2 figures
Original Natick Figure

Modified Figure Showing "Max Efficiency"

The original figure compared four different insulation materials for thermal conductivity vs. density, of which down (said to be 550 fill, but the 2.2 lbs/in^3 density indicates closer to 600 in^3/oz).

The OP figure deleted all of the insulations except down, which is perfectly OK, leaving only the 4 data points for down, and added the labels "Max Loft" and "Max Efficiency" on/near the lowest and second lowest down densities.

There is supposed to be a "linear region" but that may be hard to identify from only four data points.

I have digitized the data points in the OP figure by means of a Open Source graphical digitizer (Engauge Plot Digitizer), and used these digitized data points to generate an Excel plot with proper placement of the zeros of the X and Y axes. That is below. (Hindsight suggests that I should have digitized the original Natick figure instead.)

Digitized Natic1990 Showing Zeros of Axes and "Max Efficiency" and Conductivity Values for the point

I have added as labels the data values for thermal conductivity as obtained from digitization. The lowest density points are at .22 lb/ft^3 and 0.50 lb/ft^3 with conductivities of .318 and .265 (Btu-in/hr-ft^3-DegF). These are labeled "Max Loft" and "Max Efficiency" respectively.

So, the density ratio is 0.5/0.22=2.27X and this should be the thickness ratio of a down layer compressed by exterior clothing. The Thermal Conductivity ratio is .265/.318 = .833. So, the "Maximum Efficiency" occurs when the density soars by 1.31 X, the thickness drops by 1/2.27 = 0.44 or more than half meaning that if the conductivity didn't change, the heat flow would soar by 2.27 times. In actuality the thermal conductivity does drop to 0.833 of the original value so that at "Max Efficiency" the heat loss flow is 0.833X2.27 = 1.31 X that of "Max Loft".

"Max Efficiency" therefore has HIGHER HEAT LOSS than "Max Loft".

In short, the thickness drops by a heck of a lot, but the conductivity doesn't change by as much as 20%.

So how can the indicated point actually be "Max Efficiency"?

The Extraordinary Claim (as applied to the 550 - 600 fill power down) was presented without Extraordinary Evidence.

I would need to see or hear the calculations that prove that the Natick data substantiates the claim of a point of "Maximum Efficiency" and specifically how it supports the labeling of the graphs above.

Why didn't anyone see this before? Was the use of "hidden zero" plots a part of the oversight? **

Both of the books called "How to Lie with Statistics" and "How to Lie with Charts" have negative comments on, and examples of, how misleading it can be to readers to have graphs with "Hidden Zero" points of the axes. Both books can be ordered from or may be available in local libraries or more likely University libraries. (Newspapers and magazines routinely use such “hidden Zero” plots, by the way, as they appear very convincing.)

Before I retired, I spent time finding stuff like this in other people's writings in order to help protect my employer. I had hoped I didn't have to do this kind of work again.

I claim no great experience with down - goose or duck. I claim significant experience in rooting out technical errors of logic and errors of scientific or other proof.

The lack of data on the effect of compression (density) on the thermal conductivity of down makes difficult a discussion of the "Max Efficiency" of down packing as being at 250% compression from free loft.

I sense the lack of compelling evidence here. I have additional lacks of compelling evidence that I will present in part 2 of this. Should part 2 even be needed, that is.

* In 1982, one of the junior staff at work talked to me and my backpacking buddy about how he and his wife backpacked in the Sierra Nevada Mountains in the summers. NO backpack, NO sleeping bags, NO heavy boots, NO tent. Daypacks, a down quilt, athletic shoes and a tarp in case of rain. We couldn't believe that he survived. Too bad that in 1982 I didn't accompany them of a trip. Most of the last 30 years of my life could have been changed in a major way.

** I have a relatively simple theory of how an alternative interpretation of the density data and its treatment could lead directly to the appearance of the "Max Efficiency" concept even for the 500-600 fill down data presented in this discussion.

Edited by rocketman on 10/13/2010 19:29:10 MDT.

Roger Caffin
(rcaffin) - BPL Staff - MLife

Locale: Wollemi & Kosciusko NPs, Europe
Re: Extraordinary Claims - Extraordinary Proofs on 10/13/2010 21:31:22 MDT Print View

Hi James

> Before I retired, I spent time finding stuff like this in other people's writings
> in order to help protect my employer.
I know, I know ... been there myself.

> "How to Lie with Statistics"
Classic stuff. I had a 'teacher' at work who really could explain this. But it is not entirely simple stuff.

> The extraordinary proof is the practical experience of thousands of members
> and followers of this belief
Experimental results trump theory every time. However, there are none so blind as those who do not want to see. But it's a free world.

Perhaps you could give some details about your previous employment and qualifications? Standard resume sort of stuff.

> Should part 2 even be needed,
Certainly. Bring it on.


Steven Evans
(Steve_Evans) - MLife

Locale: Canada
Re: Re: Extraordinary Claims - Extraordinary Proofs on 10/13/2010 21:48:50 MDT Print View

I didn't see (or perhaps don't recall) the original post you are speaking of but I do recall some talk that a certain amount of compression of down will insulate better than leaving it fully lofted. I'll have to do some digging to get caught up but I have to say that this type of stuff is a great read. Thanks for your work.

Dan Durston
(dandydan) - M

Locale: Cascadia
Compressing Down on 10/13/2010 22:04:27 MDT Print View

The threads/posts I recall pertaining to the compression of down argued that compressing down a bit (ie. 2-3x) isn't as bad as it's traditionally thought to be. Not that it's actually better than fully lofted down.

The argument was that the insulation value per inch of down rises as the down is compressed almost as fast as the total thickness is decreased, so down compressed 2-3x is almost as warm (ie. 80-90%) as fully lofted down. Bottom line is that for a given amount of down it will be warmest when full lofted.

Derek Goffin

Locale: North of England
Extraordinary Claims - Extraordinary Proofs on 10/14/2010 07:06:14 MDT Print View

Thank you James. I was skeptical of the original claims of down compression. I pointed out some problems with the idea that there was an advantage in compressing down, at the time, but did not force the issue. I have often felt I should have tried harder.

David Olsen

Locale: Steptoe Butte
other factors in down compression and insulation on 10/14/2010 07:23:44 MDT Print View

In the real world, other factors come into play.

For example, an insulated garment standing proud of the body
which allows air flow underneath the garment will not be
as warm as one that lays against the body. In a down garment,
compression may force the garment against the body, sealing out drafts.

Ike Mouser
(isaac.mouser) - F
the point on 10/14/2010 07:24:01 MDT Print View

So james, can you provide a summary of what you believe your findings are, i don't really have the time or effort to read/study all that. How about a conclusion and why you think its accurate.

Michael B
(mbenvenuto) - F

Locale: Vermont
compressing down on 10/14/2010 07:27:43 MDT Print View

Hopefully someone will have the courtesy to let Richard know about this thread, in case he wants to respond.

In rereading the BPL article on the issue, you are misstating the conclusion. I don't see any claim that compressing down provides an advantage. The suggestion is that down can be compacted up to 2.5 times, without losing much insulating value.

You also seem to be ignoring the measurements of down warmth in actual jackets, that seem to support the finding that jackets that are densely packed are as warm as jackets that loft higher, with similar amounts of down.

Steven Evans
(Steve_Evans) - MLife

Locale: Canada
Re: compressing down on 10/14/2010 08:05:58 MDT Print View

In rereading the BPL article on the issue, you are misstating the conclusion. I don't see any claim that compressing down provides an advantage.

This may have been my mistake. I thought I recalled reading something along those lines, but my memory has let me down before. I can't even find the original article or discussion to jog my memory...anyone got a link?

Michael B
(mbenvenuto) - F

Locale: Vermont
down on 10/14/2010 08:24:42 MDT Print View

here is the bpl article I was referring to:

"As the down is progressively compressed, the conductive heat transmission increases but the radiation transmission decreases. Over an approximate range of 2.5x density variance, the net result is approximately the same insulation value".

Steven Evans
(Steve_Evans) - MLife

Locale: Canada
Re: compressing down on 10/14/2010 09:54:48 MDT Print View

I tried to read up on the subject, but it seems I am confused now.

I can't find the graph and such mentioned in the OP.

Art ...
(asandh) - F
Re: Re: compressing down on 10/14/2010 10:31:51 MDT Print View

with all the research and investigation that's now required to be a true Ultralighter, how do any of us find time to get out into the wilderness ?

Lynn Tramper
(retropump) - F

Locale: The Antipodes of La Coruna
Re: Re: Re: compressing down on 10/14/2010 13:11:11 MDT Print View

I'm not sure what the question is here. I have understood previous BPL articles to indicate that adding more down than is needed to minimally achieve max loft will make the garment somewhat warmer and puffier, and conversely a minimally lofted article that is somewhat compressed loses little insulation efficiency. Are your charts saying something else???

James D Buch
(rocketman) - F

Locale: Midwest
Apology on 10/14/2010 19:17:52 MDT Print View

James wrote:

I realize that somehow I managed to mislabel the axes of one or more plots. And this error, or attempting to correct this error, may have garbled the text in a few critical places.

I spent a frustrating hour just a while ago trying to correct a plot, and failed.

I'm afraid that it will be a while till I can correct the clerical errors and reply to any questions that require a clear mind.

I'm very sorry to have perhaps caused anyone who attempted to carefully follow my argument any confusion. It was my error to attempt serious detailed work at this time.

Please accept my apology, and my inability to make things right quickly.
James is on the sick list and as he notes it may be some time before he is able to correct the graph. I have removed (not censored) some of what he wrote as it included needless personal details about his health.
We wish him well and hope for a speedy recovery.

Roger Caffin
Online Community Monitor
Backpacking Light

Edited by rcaffin on 10/14/2010 20:25:21 MDT.

Greg Mihalik
(greg23) - M

Locale: Colorado
Re: Apology on 10/14/2010 19:28:31 MDT Print View

This is very small potatoes in the big scheme of things.

Rest Well.
Get Better.
Good Luck.

Mike M
(mtwarden) - MLife

Locale: Montana
original article on 10/14/2010 20:20:59 MDT Print View

maybe I misread the original article, but I (like others who have already posted) came away with that slightly compressed down might not be quite the problem once thought- NOT that slightly compressed down is more efficient that fully lofted down, but again maybe I misread the article

Steve Robinson
(Jeannie) - F
????? on 10/14/2010 20:56:22 MDT Print View

Seems like we might need remedial down school on this one. The only thing I'm sure of now is don't get the stuff wet!

Ron D
(dillonr) - MLife

Locale: Colorado
Re: Apology on 10/14/2010 21:06:12 MDT Print View

James - Take care of yourself and get well, we all wish you the best.