Applied Nate Silver – back on track

After a couple of weeks of disappointing/distressing results it appears I’ve managed to get back on track with my long-term trend and plan. It’s taken more restraint on consumption plus constant diligence on exercise in the last three weeks. The volatility in the data has made prediction more complicated (and Nate isn’t responding to my pleas for a consult); even just getting an accurate reading is now more difficult. But now I can get back to posting my “money graph”, the most basic analysis of my progress:

weekly22-money

The highly anomalous weight, back in week 19, can now be shown with long-term results not being too skewed by those two points, given my last two weeks are where they should be and made up for the off-plan results. But the following shorter-term graph with daily fluctuations shows what I’m up against trying to make predictions for the future:

weekly22-money1

This graph, with last four weeks of daily data, shows the issue of volatility in daily weight fluctuations (as well as the ongoing issue of scale variability). Week 19 was what started my “panic”. For the preceding 18 weeks the weekly “money graph” had shown steady progress, but then in week 19 the first substantial out-of-plan fluctuation appeared. Without previous daily data I can’t know if I had this amount of volatility before and it’s surprising it would take 19 weeks for it to finally show up.

Week 20 shows the problem as well – a significant drop below trendline on day before “official” weigh-in and then another highly anomalous value on the day of weigh-in. Back then, before having more of this daily data, the two weeks in a row of off-trend values was disconcerting. But then I began to get more “predicted” values, even though week 21 was well below trendline and now the week 22 value is significantly above trend, after having dropped two days ago. Strange and difficult to project future changes or even get a solid sense of what my weight actually is on any particular day.

From the short-term data the one thing I’m relatively sure about is that the second graph is not a good prediction for next week. That would predict a value of about 193.3 for next week and I’m quite certain that is too low. The slope of the trendline shows a 3.2lbs/week drop which is significantly higher than my results (about 2.7lbs/week) for my first few months. So the longer-term trendline (first graph) is probably a better predictor and that appears to be about 194.5 for next week which actually still seems a bit optimistic to me. I have “made up” for the two anomalous week 19 and 20 results by extreme restriction in consumption that is unsustainable (and probably too extreme to be “healthy”), even though the long-term trend is now measured (by the slope of regression line) at 2.4lbs/week. So I’m going to make the more conservative prediction of

195.5

even though that’s only a drop of 1.4lbs from today (and only 0.7lbs from my best “daily”).

I’d like to decisively get below 195 since April is definitely going to mess up this long-term plan. We’re planning a vacation in early April that will certainly involve some dining out (what’s the point of going on trip and then starving all the time when there will be new food treats to try). But a trip will also drastically reduce exercise at the same time as calorie consumption is likely to rise.

At my current weight my sedentary calorie requirements are now 2070/day. My average consumption (probably 10% understated in food diary recording) is 1693 (average of last nine weeks, according to BodyMedia app). So to assume I’ll only exceed that by 300/day while on vacation that will involve dining out is unlikely. So I’ll consider it a successful trip if I gain back ONLY 5lbs.

Following the vacation I then plan a camping trip where controlled eating is even more difficult (although I won’t be tempted by fun/good food as much). So my prediction will be that I’ll be lucky to be at 200lbs by the end of April, or higher than I am now. That means it will be more like end of May before I see any significant drop from where I am now and thus probably postponing reaching my target (185lbs) until June.

And I’m not even sure that 185 is a good enough target. That is a weight I haven’t had for four decades, even with a triathlon during that time (I’m now about the same weight as I was when I did triathlon but I definitely had more muscle mass then, esp. upper body due to all the swimming). And it may take getting down to 180lbs to get body fat level to the right place (now around 22%, I want to target low teens which is probably the best I can hope for at my age).

And then the question is whether I will have achieved my real goal for all this. While losing weight is good on many levels (my aging joints will be happy) my real purpose is to get my blood glucose levels low enough where I’d need any meds and possibly eliminating the meds I take now (reduced in half already) for blood pressure and cholesterol. I’d ideally like to get med-free and then stay that way for the remainder of my life. That’s a good but tough goal. So the real test of all these graphs and keeping stats is to address age- and obesity-related health issues, not just some personal vanity to look like I did when I was rowing crew in college.

So I’ll make a more daring “prediction” (really goal) and that is to achieve 180lbs by my birthday. That’s still a long way off so I’ll get plenty of opportunity for more boring posts on all this.

So let’s hope I have good news to report in the next two Sundays before I fall off-the-wagon in April.

About these ads

About dmill96

old fat (but now getting trim and fit) guy, who used to create software in Silicon Valley (almost before it was called that), who used to go backpacking and bicycling and cross-country skiing and now geodashes, drives AWD in Wyoming, takes pictures, and writes long blog posts and does xizquvjyk.
This entry was posted in comment, musing, prediction and tagged , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s