Not Widely Accepted Practice

Rejecting something for being NotWidelyAcceptedPractice is an example of letting your actions be directed by a PurposeSupportingCliche, it's far better to think for yourself.

Recently, in an internal workshop on commenting (when to, when not, what makes a comment worthwhile, what makes most comments worse than useless) a colleague challenged a couple of suggestions with the (correct) claim that they were NotWidelyAcceptedPractice". Now, these suggestions might be beneficial, they might not, they might be appropriate, they might not. But given the shoddy state of much of the industry, that they are not widely accepted is pretty much neither here nor there. As I pointed out in that session, it's almost (but not quite) a recommendation by itself. Indeed, if our team were to restrict itself to majority practices, I went on, we probably wouldn't be doing version control, nor UnitTesting, and certainly not PairProgramming or a bunch of other things that we get a lot of benefit from. And we would be wasting our time, and client's money, on a bunch of stuff of no use to anyone. If we worked the way the majority of the industry works, we'd be doing our job very badly indeed. And yet, as Jeff suggests above, there is a tremendous complacency around, great surprise over and resistance to calls for change. -- KeithBraithwaite

While not commenting on the examples given above, the argument NotWidelyAcceptedPractice is not necessarily an AntiPattern. One disturbing trend repeated often in our field, is the ease with which IT shops and the like embrace new-fangled ideas and such, peddled as SilverBullets (either by vendors looking to make a buck, or egotistical career-minded individuals within the organization looking for some way to increase their visibility), without so much as an ounce of serious critical review. A new idea (new to an organization) should be considered if and only if any of the following is true: 1) It has been shown to be a BestPractice (or at least a good practice) for the field as a whole, or for similar organizations in the same specialty (in other words, it is "widely accepted"); 2) It is new to industry but has quite a bit of academic research to back it up, or 3) the team/department acknowledges that they are trying the new and unproven - they are the ones conducting the research.

In short, NotWidelyAcceptedPractice is frequently a good defense against snake oil. Of course, how one distinguishes between new-but-promising methodologies like ExtremeProgramming (especially in its early days, before it got an established track record) from the snake oil, is an unsolved problem.

Further, failing with a NotWidelyAcceptedPractice is not AnAcceptableWayOfFailing.

"...how one distinguishes between new-but-promising methodologies like ExtremeProgramming (especially in its early days, before it got an established track record) from the snake oil, is an unsolved problem."

No it's not, you just use your brain and decide for yourself, a solution as old as man.

The problem with "using your brain" is if you ask n different practitioners (all of them who are well-versed in the subject, and that n answers are indeed possible), you may get n different answers. Many spectacularly wrong decisions have been made as a result of intelligent, well-meaning individuals "using their brain", "trusting their gut", or otherwise depending on their judgment/expertese alone, rather than searching for imperical evidence. This is not intended to devalue experience (in many cases, it is all that is available); instead it is intended to point out that dependence on experience is not a reproduceable or falsifiable technique--it is highly dependent on the individual in question. And many people's experience leads them to think IfItsNewItMustBeBetter; others experiences teach them to avoid that which is NotWidelyAcceptedPractice.

In short, NotWidelyAcceptedPractice is frequently a good defense against snake oil.

I disagree. NotWidelyAcceptedPractice is not a good defence against snake oil - simply because it does not discriminate based on the quality of what it is used against. Proof: assume everyone is using snake oil. In this case, the argument supports the continued use of snake oil. The argument, I think, is only useful as a defence against change. Thus, it is only useful in cases where change, in general, is harmful. If you believe that, in general, the IT shops are changing too quickly, and that, as a result of changing, they suffer, then this concept is a useful one to adopt. If, instead, you believe that IT shops adopt too many bad ideas, then you should instead support an argument which encourages them to think before changing (as opposed to an argument that encourages them not to change at all).

Note that, though often used as such, NotWidelyAcceptedPractice is not a valid counter argument to IfItsNewItMustBeBetter. Both arguments are fallacious - they argue against the value of a concept based on its age, or general acceptance. They both pretend to say something about the quality of a new concept, but are not. They are instead commenting on the preparedness of the world at wide to accept said new object. One means to say yes, we're willing to change, and the means to say, no, we are not willing to change, while neither is considering the question of: is this concept offering benefits which are worth the price of changing? (The fallacy occurs because instead of saying what they mean about being prepared to change, both arguments instead say: the new concept is good/bad.)

See: FallaciousArgument