Aug 4, 2014

#67 What To Do with Feedback

All Rights Reserved © 2006 Thomas W. Day

Many years ago (1974) a really good dirt biker and equally creative writer named Carl Shipman wrote a book called The Boonie Book: How to ride the dirt, take care of your bike, and yourself. I found this book not long after "Zen and the Art of Motorcycle Maintenance" entered my "best book" collection. For a few years, Shipman's little book, and a few of his dirt bike products, were a regular part of my life. I used the heck out of The Boonie Book in developing my own riding style back in those days and I still think of some of the Shipman concepts when I'm teaching new riders. I try to remember those lessons at the beginning of a new riding season, every year. I probably loaned the book to someone, otherwise I'd still be reading it occasionally in those dark, depressing, motorcycle-less Minnesota winter months.

One of the many concepts that Shipman eloquently discussed in his book was that of using the design and operation of electro-mechanical servomechanisms in his description of how a motorcyclist "holds a line" through difficult terrain. I really wish I had a copy of the book, today, so I could accurately quote Shipman, but it's about as easy to find this book as it is to locate a coherent political conservative. I've had it on my Half.com watch list for more than three years (The Boonie Book, not the conservative.). The gist, however, is that when trying to stay a course, we don't attempt to "hold a line." Instead we correct our steering, often, so that we pass through the line on either side as accurately and frequently as necessary to stay on course without tightening up and freaking out when we get a bit off of the ideal path.  Today, for the zillionth time, I heard a politician, executive, media pundit, or some other ignorant bonehead misuse the "feedback" concept and it tripped a trigger in what's left of my memory circuits. The words "positive" and "negative" when related to feedback are grossly misunderstood everywhere but in control circuit design applications. When you are designing a system to perform a task as accurately and with as much quality as possible, what that system's control circuitry desperately needs to stay on course and in control is "negative feedback." If you are swinging wildly from one end of a system's steering limit to the other, you are probably suffering from "positive feedback." Positive feedback adds lots of bad data to the input of your already misdirected steering system. Negative feedback sends a bunch of that same over-amplified error back to the input, resulting in lowered gain, less distortion, less error, and more accurate steering.

In the "happy talk" world of modern politics and business, "positive feedback" (the output of "yes men" and political pundits) produces even more wrong-headed decisions from leaders in business and government.  Impossible to imagine, but they are capable of being even more wrong than their natural inclinations. What these people desperately need is criticism, doubtful analysis, restraint, and careful measurement of performance vs. philosophy. Without negative feedback, our charismatic but non-scientific leaders pick up a stupid idea and run with it like scared rabbits until they smash into the wall of reality. Then, like the ENRON morons, Lay and Schilling, they whine "nobody told me it was a stupid idea!" As if they'd have listened if someone did try to provide them with useful (negative) "feedback."

We get this sometimes in the motorcycle safety classes, too. Instructors walk a hard line between providing encouragement (positive feedback) and actual information (negative feedback) to produce a little control input to student's natural tendencies. Those natural tendencies include panic reactions, looking at the ground a few inches beyond their front tire (or staring at the speedometer), correcting the motorcycle's path of travel with ineffective body English or incorrect steering moves, and whacking away at the shift lever like they were stomping a snake with their left boot. Too much negative feedback and the new rider is discouraged, resentful, and seizes up and stops trying to learn. Too much positive feedback and the rider gets arrogant about doing things "my way" and regressively continues doing the wrong things until blood is lost and skin is sacrificed to the Gods of Roadrash. You can't top a crash for excessive negative feedback. Typically, the result of too much negative feedback is an overwhelming resistance to movement of any kind; "You can't go wrong if you don't go anywhere."

Shipman's theory of steering is really useful here. Basically, and poorly, restated, Shipman said you pick a line, steer over it (missing to the right, correcting to the left, missing it again, and returning to the right, and so on), attempting to stay as close to the line as possible, while keeping your steering input as flexible and relaxed as possible. Eventually, if you practice enough, you'll appear to be holding the line perfectly to an observer, while, in reality, your steering input will be just as precise and relaxed as required to stay in the immediate vicinity of where you want to be.

That's precision negative feedback.

Positive feedback causes more of the same (usually wrong) action to occur. You want to turn left, you point the front wheel left, the bike leans right and turns right. You want to turn left, but positive feedback tells you to provide more of the same wrong input (point the front wheel left), the bike leans harder to the right. Now, the bike is dragging parts and going the wrong way. More positive feedback, you hang your body to the left, but point the wheel further left. Your direction of travel doesn't change, more metal bits grind away, the bike is heading solidly right and you're still wondering why when you come to some kind of stop (hopefully, painlessly). That's uncontrolled positive feedback.

Criticism is like that, too. Mostly, motorcyclists pretend that they have changed their image from the bad old days of biker gangs and Marlon Brando. Get out more. You are taking in too much positive feedback and over-steering yourself into oblivion. You need to meet some ordinary, non-biking people. Motorcyclists have a lot of images, but more of them are bad than good. We're noisy, we clog up highways with our parades, we scare the crap out of people wheeling through heavy traffic, and our bikes can spew 50 times the pollution as a modern car. Yet, we hang out together, pat ourselves on the back, and pretend that everyone loves the sound of our straight pipes, is entertained by our freeway "antics," and thinks we're brave and adventurous individuals. Some people do, most people don't. Ignoring the majority opinion won't make it go away. Changing directions quickly, accurately, and as a reaction to a well-thought out path of travel is a requirement for any moving vehicle that intends to remain in control. Pushing the bar in the wrong direction because of inbred, uninformed positive feedback will just make us go the wrong way faster.

Negative feedback, opposing opinion, dissent, is a good thing. It's good for me. (I get it in your letters and pay attention, even when I disagree.) It's good for you. It's good for the country. It's good for motorcycling. You can't turn the right direction without it.

October/November 2007

No comments: