I am probably going to either annoy or bore most readers with this, but since that describes about half my posts, I figure I can risk it. I will warn you up front that it is way off topic, but then again, I post so many off topic posts I figure one more won't hurt.

What is annoying me today, and many days, is the disturbing way people misuse numbers to make their point. No, today I am not talking about bogus statistical analysis, misuse of different sampling methods, or the "hockey stick", nor anything else so complicated. Instead I am talking about the sort of tricks used by marketing departments, or by clever computer developers, to make their results seem more impressive.

For example, suppose you see a product that says it is "100% faster"? Now many may just pass this by, but my first reaction is "Wait! That's impossible!"

You see, when I think of measuring speed, it is in terms of how long it takes to accomplish something. So a task takes X seconds. If it takes 1/2X, then it is 50% faster. As you can tell from the way I think of speed, it would have to take no time at all to be 100% faster. So any claim of 100% faster or more is simply impossible.

But if you redefine speed as how many X it can accomplish in a given time, then if you can accomplish 2X, it is 100% faster. However, if you look at it in terms of my original definition, it is only 50% faster.

This isn't a problem when they claim 100% faster or more, as obviously they must be using the second definition, or using really bad math. But what if they say "It is 50% faster"? Does that mean that it takes half the time to do a single iteration, or you can fit one and a half times as many iterations into a given time? Because that second is only 33% faster by the first definition.

And that is the problem with such claims, they sound really impressive, but they don't tell you much. Without clearly defined terms they are meaningless, and sadly many terms, such as "faster", have more than one definition. Nor is "faster" the only problematic term. In the commercial arena, for example, markets have not decided if screen size for monitors is corner to corner or horizontal, nor whether it includes the bordering container or not.  Similarly, disk manufacturers have yet to agree whether a megabyte is 1,000,000 bytes, or 1,048, 576 bytes (which makes quite a difference when dealing with large disks).

Which, I suppose, does bring us back to the political arena, and makes this a bit relevant after all. For example, "budget cuts". How can two different politicians quite sincerely call the same proposal a cut and an increase? As many of you may know, because budgets in DC aren't like normal budgets. Where you take available money and figure out how to spend it, DC budgets take last year's budget, multiple all the items by a fixed growth rate, and THEN start adding money, only as a final step asking where they will find the money. So, if something had $1000 allocated in 2000, in 2001 it will have, say $1100 because of the 10% growth factored in. So, if a proposal allocates $1050, one politicians calls it growth, because $1050 is more than $1000, while the other calls it a budget cut, because $1050 is less than the $1100 required by the 10% growth factor.

All of which brings me back to my one bit of advice. Always ask "What do you mean by that?" Most people never ask, either because they think they have a "good enough" understanding, or because they are afraid that they will look foolish asking for a definition. But it is essential. Until you know precisely how terms are being used, it is impossible to have a meaningful discussion.

In some ways this parallels my earlier writing on pragmatic solutions. If you have different definitions of "working" then you will reach completely different "pragmatic" solutions. This is just a more general application.

So, please, when a debate seems to be lost in confusion and everyone seems to be talking past one another, you may want to look at what you have been discussing and ask if you are actually using words to mean different things than others are. It may make the discussion much more productive.

POSTSCRIPT

Having worked in the computer industry, I have seen a lot of this. The industry buzz words tend to be so ill-defined that often they are close to meaningless. For example "scalable", which was the term of the moment in 2000 and 2001. Does it mean that you can increase the number of users? Or of content? That you can increase the size of the application on a single server? Or that you can run on multiple servers? Or one or more of the above? Or something else? No one knew then and no one knows now. It is an utterly meaningless term. All it means is something can be increased or decreased without too many ill effects, but what can be changed, and with how much effort is not defined.

Nor is the computer industry alone in this. We may be particularly afflicted due to the confluence of techies, overly fond of jargon and acronyms, with marketrons, also far too fond of jargon and prone to take techie jargon and mangle the meaning even more, but the tech industry is hardly alone. My wife sometimes manages to utter entire paragraphs so full of hospital jargon and shorthand that I can't understand what she said. And I have been dealing with this jargon since she was in nursing school, so I can't imagine what it sounds like to a total neophyte.

Then again, I suppose jargon has its place. In some fields it has its uses. For example, knowing what p-code is helps one understand why Java isn't a true compiled language, or knowing what pipelining is (and isn't) helps you understand processor speed, as well as see through marketing hype about longer instruction pipelines. Used properly jargon can make things mroe clear and simplify discussion by creating a convenient shorthand for concepts that would take dozens of words to explain otherwise.

So jargon does have its place.  What is a problem is when we do not agree on what that jargon means. Which is why I said we need to be willing to ask one another "what do you mean by that term" when we are unsure. Otherwise jargon runs the risk of making things less clear, rather than more.

POSTSCRIPT II

What is far more puzzling, and more prone to cause confusion, is when the term causing confusion is not jargon but a common word defined in an unusual way. For example, sometimes I think the word "greed" is being used in a nonintuitve way in discussion. And it is far more pernicious as people are very unlikely to ask "what do you mean by dog" in a discussion. But if someone is using "dog" to mean "only a Pomeranian" it is important that we ask, as that sort of solipsistic definition can lead to endless confusion.