Interesting info -- thanks, all!
The 4.5 lb of weight loss over 2 weeks is a very rough figure. I stepped on the scale wearing clothes, and didn't make any effort to check if it was the same amount of clothing before and after. I think the amount of water and food in your body can also vary quite a bit. A liter of water in my belly would be 2.2 lb. A day's worth of food in my belly would be roughly 1.5 lb. So with fluctuations and uncertainties of that size, I wouldn't even claim that the 4.5 lb of loss was statistically different from zero.
@Derek: I guess there are at least two different ways that one could define this figure: (1) calories per day per pound of body weight, and (2) cal/day per pound of total weight, including body, pack, and clothes. #1 was what I was originally stating and asking for. I can see how there could be arguments in favor of #2, but it would be a different thing. My body weight is about 135 lb, and my skin-out weight on the 11.5-day trip probably averaged about 20 lb, so by method #2, the figure would probably be about 16 cal/day/lb.
Compiling the numbers people have given so far, by method #1, here's what we have. Each line is cal/lb/day by method #1, followed by body weight in lb.
Ben: 15-18, 135 lb
Chris: 16-24, 143
Derek: 22, 154 [*]
Christopher Shive: 17, 180 [**]
Chad Miller: 17, 235 lb [**]
[*] recalculated as (3300*17+1400)/(17*154) to make it method #1
The results so far seem to be strongly clustered in the range from 15 to 20 cal/lb/day. To me, this seems a lot more useful than pounds per person per day, which people in these two threads have reported as varying wildly from 0.75 to 2.
Tohru wrote: "What you should do is get all the responses and plot cal/lb/day versus body weight. I'd be interesting to see the dependence..."
I don't see much of a trend in the data so far, but a lot of us are clustered around the same body weight. What's yours?
@Sam: That's an interesting thread you pointed to! Accounting for food by weight is definitely more convenient than doing it by calories, and that's the method I use a lot of the time. The thing is, the caloric density of food really varies a lot, so if we're trying to reach definitive conclusions here, IMO it's more useful to cut down on the number of variables by using calories. I used to achieve a much higher energy density with my food until someone here on BPL pointed out to me that a healthy diet should have proportions of about 50-35-15 calories from carb-fat-protein. Forcing myself to do that significantly lowered the energy density of my food, from about 2100 cal/lb to about 1800 cal/lb.