An interesting article has been making the rounds, stating that “There’s a One in Ten Chance You are Dreaming Right Now.” But something about the statistics triggered my skepticism. Can we really say that at any given moment you could be mistaking reality for a dream?

For context, the article linked to above is quoting from a new book about reality, and offers the following:

We optimistically assume that you get eight hours of sleep a night, which leaves sixteen hours during which you are awake. Sleep researchers have found out that there is a strong correlation between dreaming and being in so-called REM (rapid eye movement) sleep. REM sleep is characterized by rapid movement of the eyeballs; the brain is highly active, its electric activity resembles that of a waking brain, but the sleeper is more difficult to wake than during slow-wave or non-REM sleep. We know that between 2-% and 25% of our sleep is REM sleep. Taking the lower value and assuming that you always and only dream during REM sleep, this gives us 1.6 hours of dreaming ever night. As there are therefore 1.6 hours of dream consciousness for every 16 hours of waking consciousness, this means that your chance of dreaming at any given moment is 1 in 10 [emphasis mine].

I am not going to argue the numbers so much as I am going to contest the premises of this calculation and the conclusion.

To put it bluntly, the comparison is incorrect. You cannot compare the hours of dreaming to the hours of wakeful consciousness, as these things do not co-exist (by definition). We have to talk about the probabilities that are actually equitable. The probability that you are dreaming depends on you first being asleep (as you have to enter into REM sleep from earlier stages of sleep). So, the probability that you are dreaming is dependent on the fact that you are asleep. The only question we can ask, to salvage the comparison made above, is what is the probability that we are dreaming, given that we are already asleep? Doing the math it would be around 20%, using the numbers in the example above.

If you are not asleep, you are not dreaming. Then, by definition, during your waking hours (the 16 assumed in the example above), you have a 0% chance of dreaming. You are simply do not have a probability of being asleep “at any given moment” if you are already awake, and it hardly makes sense to think of probabilities in this way. If there is some condition on which the chances are dependent (sleeping prior to dreaming), we have to take this into account. To use another example, it would be absurd to say that you have a 1 in 11.5 million chance of getting attacked by a shark “at any given moment” if you live the entirety of your life in the desert. You first have to enter into the conditions that allow for the other probabilities to mean something (i.e., frequently swimming in the ocean).

The headline makes for an interesting soundbite, but at any given moment you are not possibly dreaming. That conclusion first has to be filtered through the probability of sleeping, and it doesn’t make much sense to say that you have a 33% (8/16 hours) chance of being asleep right now. At the moment you are reading this, you are awake, and are not dreaming, thus making your chances of dreaming at this moment 0% (not getting into discussions of reality). This should be enough to show that the phrase “any given moment” is an incorrect simplification.

But then again, maybe I am having just a particularly lucid dream…