Saturday, November 21, 2009

SHOW ME THE DATA

My fellow Americans:

(And you are my fellow Americans, according to my strangely fascinating stat counter. Hello as well to my Australian reader, and a big hi to the Saudi Arabians looking for Penthouse. You didn't stay long. I hope I didn't say something to offend you?)

Our long national nightmare is over. I have recovered from my cold (no, Mom, it was not the Swine Flu), and Suleyman the Upchucker has recovered from his inexplicable affliction with the Projectile Vomit Pox. The cat vomit has been cleaned from my sheets, carpets, walls and shower. In further good news, I've discovered my reflexes aren't quite as bad as I feared: The speed of my reaction to anything like a cat making a whoop-a-whoop-a-whoop noise has actually been quite impressive.

My cat-vomit-stimulus-reaction speed might be seen as confirmation of Hick's experiments with choice reaction time. These experiments are often held to have interesting implications for martial artists.

The British psychologist William Hick demonstrated that given n equally probable choices, the average reaction time T required to choose among them would be roughly

T = blog2(n + 1)

The key point is Hick's prediction that the time it takes for people to make a decision increases logarithmically, not linearly, in proportion to the number and complexity of the choices with which they're confronted. In other words, it is vastly easier to react quickly if you have only one choice, i.e., two options.

Is it vomiting?
Yes, No.
If yes, get it off the bed.
If no, go back to sleep.

I can make that kind of decision very quickly -- in about a nanosecond -- just as Hick observed. If we add more choices, however, I get much slower:

Is it vomiting?
Yes, No.
If yes, either look up symptoms on Google or assess for signs of poisoning or consider the possibility of gastritis or examine color and quality of vomitis or examine vomiting cat for signs of dehydration or call the vet in the middle of the night or wait for a more civilized hour ...

Choosing among these five options, however, does not take me a mere five nanoseconds -- it takes me all week. (How much longer it takes is empirically determined, that's the "b," which is a constant.)

Now, why is this relevant to the martial arts? That's where reader Jesse Crouch, the Martial Explorer, comes in. It was Jesse who brought Hick's Law to my attention, for which I thank him.

Hick's Law offers support for something Egemen has said to me over and over: In an emergency, it is better to have a few very effective self-defense reflexes, deeply internalized, than a superficial acquaintance with a wide number of possible reactions. This is Egemen's objection to systems such as Krav Maga, which he terms "eclectic." As Egemen says:
An eclectic system tends to be less effective in a real situation because again, in the chaos and confusion of a real fight, you do not want to be standing there, scratching your head, trying to remember which technique you’re supposed to be applying (“Do I grapple? Punch? Block? Parry? Duck? Weave? Growl like a tiger?). You want to have a single set of reflexes.
Given that reaction time is profoundly important in a fight, Hick's Law would suggest that Egemen is absolutely right.

I was pretty happy with this explanation, until I studied the question a bit more closely and discovered that in fact recent research into Hick's Law -- and particularly into that constant, b -- is conflicting, and the implications considerably more ambiguous for martial artists:

Larish and Stelmach in 1982 established that one could select from 20 complex options in 340 milliseconds, providing the complex choices have been previously trained. One other study even had a reaction time of .03 milliseconds between two trained choices! .03! Merkel's Law, for example, says that trouble begins when a person has to select between 8 choices, but can still select a choice from the eight well under 500 milliseconds. Brace yourself! Mowbray and Rhoades Law of 1959, or the Welford Law of 1986, found no difference in reaction time at all, when selecting from numerous, well-trained choices.

I'm not really sure what to conclude from this. I do know that when it comes to getting a vomiting cat off the bed, I am very fast. Faster than Egemen, in fact.

But sadly, as Egemen reminds me, it is not enough just to be fast: You must also be accurate. Missing the target, even once, can leave you with a major problem on your hands.

In this case, it is the kind of problem no amount of air freshener will ever make right.



4 comments:

  1. I agree with the author on hockscqc generally. A few points:

    - I agree that Hick's law is antiquated. I also agree that its inaccurate to assume the brain works in serial mode and cannot work in parallel in processing thoughts. There is obvious research and even casually observable evidence of this. However, even computers that can work in parallel can get bogged down. Even more, this isn't just about proceedural decision making - a lot of this has to do with other variables, such as things that come from the biological level.
    Hick's Law fits better at a conceptual level of "too many choices is too much confusion (eventually)" and "the body ends up going to what it naturally does anyway, even little kids do this" - which is flinch response (See: Tony Blauer's SPEAR system).
    I don't follow Hick's from a mathematical level.
    - He mainly discusses athletes. These are professionals of a field (even if they're just kids playing a game) who have a set of expected outcomes to deal with, not people in a random situation with a almost completely random number of variables and outcomes to deal with.
    - He lumps "martial arts" and "self defense" together - two very different things. Martial arts refers to people going into known situations - competitions and war, generally - much like athletes of other sports. Self defense is a very different, much more complex world. This is one of the major issues with all martial arts instruction today - it all gets lumped together and treated as the same.
    - None of this accounts for flinch response and other things that the body does that a 'martial artist' may not expect it to do. Blauer's "system" (his system is really just a module you can throw on top of any martial arts system) very basically leverages flinch response instead of fighting it (what most fighting instructors do).

    My idea is not to dumb things down, but to make things more universal (more on that later). Hochheim and I come to quick agreement on this one I think:
    "It seems like the last 8 decades, Hick's Law has become a legacy of evolving research. But, Hick's Legacy is really telling us to train more and smarter, not necessarily to be stupid and learn less."

    In the end, W. Hock Hochheim and I come to a lot of the same conclusions - Sequential Learning, Conceptual Learning - both things I agree very much with. I wouldn't even say we're in disagreement since I don't follow Hick's from the mathematical level (and I don't think Blauer does either).

    ReplyDelete
  2. Jesse, thank you -- very interesting. Is it possible simply to explain the idea of "leveraging the flinch response?" I'm not quite sure how to understand this from Blauer's website.

    ReplyDelete
  3. I'm so glad my grandkitty Suleyman is OK now - and my daughter too..

    ReplyDelete
  4. Blauer leverages flinch response by focusing on what the body does naturally instead of focusing on complex motor skills (general martial arts technique). Basically - your body naturally flinches and can be effective. Learn to work your technique from there.

    Here's a good video: http://www.youtube.com/watch?v=jk_Ai8qT2s4

    ReplyDelete