Sunday, July 20, 2008

Counting Sheep

A few months ago, I started to wonder about the mystery of my sleep patterns. Some weeks, I would get by with five hours each night, and get lots of work done. Other weeks, I would sleep eight hours a night, and still be exhausted during the day. I had the idea that perhaps by collecting data on my sleep patterns and how awake I felt, I could figure out a mathematical model that would be able to correctly predict how tired I would feel each day. I reasoned that if I had such a model, I could use it to figure out how much sleep I needed to get in order to get things done, because I find it very frustrating to plan a late night of work, only to find myself too tired to do anything.

So, I collected data on myself for several weeks, and played around with some simple models, and sadly, none of them seemed to match the data at all. I was complaining about it to Drew Davidson, and he suggested I read Counting Sheep by Paul Martin. He wasn't sure it would help, but he said it was interesting anyway.

And he was right! It was interesting. It is basically a guided tour of the science of sleep, trying to touch every aspect. It doesn't get very deep, but it gives a little bit of info about everything. It's also full of interesting quotes and examples from history, literature and mythology, which keeps it from being a dry read.

Some of the fun facts I learned in Counting Sheep:
  • People who are early risers tend to have more stress chemicals in their blood (that explains why you early people act like that!)
  • People who have frequent nightmares are easier to hypnotize. I will be thinking about that for a long time.
  • We don't just dream during REM sleep -- there are also dreams during NREM sleep, but they are a bit different.
  • I knew when you were falling asleep, it is called the "hypnogogic state", but know I know that waking up is the "hypnopompic state."
  • I learned the horror of Ondine's Curse.
  • If your eyes don't keep moving, oxygen doesn't get to the iris. This may be part of the reason for REM sleep.
  • There are two kinds of sleep apnea: "obstructive" (the normal kind, where your throat collapses while you sleep), and "central", where some kind of brain damage makes you stop breathing in your sleep.
I found this book a nice complement to Stanley Coren's Sleep Thieves, which has a different set of facts, more oriented towards why sleep is important. Counting Sheep is more of a general overview of the history and science of sleep. And though it was fun, I still don't have a mathematical model that can predict when I'm too tired to work. If you have one, please let me know!


  1. "And though it was fun, I still don't have a mathematical model that can predict when I'm too tired to work. If you have one, please let me know!"

    This paper may interest you even though it's mainly regarding Slow Wave Sleep (NREM Stages 3 and Stage 4) during the first 12 months of life. Nonetheless, I go into a little detail about how things we learn during the day are temporarily stored in the hippocampus and then transferred to the neocortex during NREM Stages 3 and Stage 4. Being a statistical programmer NREM sleep seems to be a bit like running a batch job overnight for my bank and updating all the Oracle and DB2 tables..

  2. Sounds cool... which paper do you mean?

  3. "which paper do you mean?"

    This guy is the Chief Medical Officer for NASA:

    "A reassessment of the SIDS Back to Sleep Campaign" by Dr. Ralph Pelligra also discusses this issue: