The signal and the noise

I like “Ask Marilyn.” She usually has an interesting take on a simple question. Last week in Parade magazine she addressed the phenomenon of pareidolia, which is seeing or hearing things that aren’t there.

I will often think I am hearing a song or a radio commercial when there is no source for either. When I track the sound source down it is often something like the hum from fluorescent lights or a ceiling fan. Even the rumbling of a refrigerator’s compressor will do the trick. Somehow my brain is creating a signal out of the noise.

It’s sort of like seeing a giraffe in the clouds or a dragon in an aurora. We do it all the time. Humans like patterns and we like to make meaning out of the sensory barrage we experience. Unfortunately we experience a lot of noise, that is, a lot of random stuff. We don’t like random stuff, we like order and structure, so (apparently) our minds supply the missing pieces and give us a meaningful picture or sound instead of a mess.

In radio telephony the signal-to-noise ratio is a comparison of the strength of the desired information (signal) to the background interference (noise). Think of the signal as what you desire to hear or otherwise interpret and the noise as the rest of the crap that you have to filter out.

We do this in a crowded bar when we focus on our neighbor’s words and tune out the cacophony produced by the rest of the patrons.

But pareidolia is the opposite. In this case we are supplying the signal. It’s not there! All that’s there is noise but we are determined to make sense of it and so we plug in some kind of hallucination.

Gamblers do this. Lousy gamblers, anyway. They see a randomly generated sequence like throws of the dice or fall of the cards and imagine a pattern emerging. Then they bet on an expected outcome based on that pattern. The pattern isn’t there, of course, that’s why Las Vegas casinos make so much money. Assuming the roulette wheel is “fair,” that is, every number has an equal probability of having the ball land in its slot, there is no pattern to the results. No one can guarantee the wheel is perfectly random, but it can be made random enough so that the outcomes are indistinguishable from pure randomness.

Even random number generators are just approximations of randomness. “True” randomness would have to be something natural, or perhaps I should say something physical. These days “natural” is too loaded, making one think of hand-churned butter or turd-fertilized tomatoes.


A really good physical phenomenon to use is a radioactive source. The points in time at which a radioactive source decays are completely unpredictable, and they can quite easily be detected and fed into a computer, avoiding any buffering mechanisms in the operating system. The HotBits service at Fourmilab in Switzerland is an excellent example of a random number generator that uses this technique. Another suitable physical phenomenon is atmospheric noise, which is quite easy to pick up with a normal radio. This is the approach used by RANDOM.ORG.

These folks are looking for the noise! That’s un-natural, man. People are much better at seeing the signals.

I think we all like to believe that there are meaningful patterns in things that are really closer to being random. I like crime fiction and cop shows and a well-worn trope in both is the grizzled vet who always says “there are no coincidences” or something similar. The problem is that in real life there are heaping piles of coincidences. Things that are mostly just luck (random chance can be good or bad) take on significance because of nothing more than coincidence. A quick look at the probabilities of each (independent!) event would make the coincidence just that, something remarkable but not otherwise meaningful.

I like baseball. Baseball fans attach significance to almost everything. A batter is 60-for-300 on the season but gets a start against a pitcher because he went 5-for-11 against him last season. A guy gets six hits in two games and he’s “on a hot streak” and you just know he will get three more hits in today’s game. It turns out that if you make a chart of a guy’s hits, like hash marks on a calendar each day, it will be indistinguishable from one generated by a computer that knows the player’s batting average. You could even simulate it with dice rolls. The “patterns” created by the human player won’t look any different from the “patterns” in the simulations. They will differ in the details, but overall you won’t be able to pick out a chart made by dice or by a real hitter.

So I say beware of the signal and the noise. It is hard enough to sort out what you want from the background of everything else, and now we know that our brains don’t help because they like to make shit up!

Please comment!

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

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

Facebook photo

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

Connecting to %s