Month: October 2011

False Analogy – meh

Just had a stupid Twitter spat with someone about the analogy that asking someone to predict the number of bugs is like asking a chef to predict the number of people who will fall ill.

It’s hard to discuss things in 140 characters, particularly if the individual is some kind of pedant who just can’t be wrong. I used the term false analogy, which was originally coined by the long-dead English philosopher John Stuart Mill. Mill meant that the analogy was being misapplied and did not fit where it was being used. However I was dealing with an über geek, who took the literal meaning of the term and started off a load of crap about how analogies are by definition false. Meh, I really must not assume that people know the same things I do. I’ve been using this term (without any problems) for years, but I also did study philosophy in my youth (30 odd years ago now).

My take was that we have more than 100,000 years of cooking expertise and 40 or so of software engineering. We’ve already been through the how do I cook that tuber full of starch in such a way it won’t poison me stage of our culinary development longer ago than most of us can conceive of. In software terms we’re still finding out that starch is good for us. Also most chefs, in my experience, don’t measure their success or failure by how ill their customers are. They measure by whether their customers come back to the restaurant, which is a result of the 100,000 years.

So, tried to make this point in 140 characters – mistake!

I got some stuff back about how our knowledge of germs and software are about the same. Still not true, actually, because there are a hell of a lot less people creating software than there are people who cook food and follow basic hygiene. But I can’t be bothered with this argument, I really can’t.

Some people just have to win, and aren’t short on off the cuff insults. So I unfollowed and blocked. Got better things to do, seriously!

The other thing I’ve noticed is that some people seem to have a supply of URL’s they need to make a point – and just throw them at you at any opportunity without even thinking. I’ve never used tumblr, but suspect that this is a symptom of using it. For example I mentioned there is a model for bugs. I really meant for how long things take, as in using kanban you can say things to the business like I can promise that a ticket takes between 10 and 15 days once it hits the started state. I got pointed at this(PDF) – I don’t care, all I’m interested in is shipping stuff that works. I leave the pedantic stuff full of meaningless definitions that don’t fit what I do to the people who care about this stuff.

Statistically, if you keep metrics about your kanban board, you can make predictions like this. You can employ different confidence limits depending on how important the predictability is. David Anderson talks about this in his Kanban Book at some length. But it’s not the kind of ISO crap, it’s predictably shipping useful stuff. I believe the book he’s writing at the moment covers it in more detail, because I heard this at his keynote at Agile Cambridge last month.

Statistics are interesting. You can say how long any particular job will take, just look at the distribution of how long it might take. That’s the way it rolls. But then I also studied statistics in my youth, and even have a B. Sc. in it.

Tweeting things like I’m surprised an agile coach would say this and lecturing a practising Buddhist about koans as analogies (although we don’t have them in the Tibetan tradition – and they aren’t analogies, but I don’t have the energy to bother explaining why) is more than a little insulting. But the interwebs are full of such comments.

In any event – that’s my take and I don’t care to discuss it with people who are going to make ill considered, insulting comments in 140 characters. Or I’ll have to go up to dickfinity on ya.

Plus I also changed my Twitter profile to spell agile with a lower case A.

Oh, and it should probably have been stupid hyperbole, not false analogy. But I didn’t want to start a row. More fool me.