Is Agile for Amateurs?

I think agile often creates an illusion of professionality that can lead people to collectively believe that they are doing pretty OK and have reason to be proud of themselves while in fact they are progressing at a snail pace putting together bog standard database applications. The sad reality in our industry is that a single excellent developer can outpace the productivity of an entire team of average programmers. It's not a myth and it has absolutely nothing to do with methodology or best practices. Software engineering continues to be a skill based profession where most of the skill is acquired through experience, apprenticeship, and intelligence after you leave college. Put a few smart and productive developers together and they will start building stuff almost immediately. Most of them do it in their free time as well. If you have such people around, trying to not slow them down, annoy, or alienate them is a pretty decent business strategy.

Most open source projects are run by such people and some excellent software is coming out of there. OSS and Agile are almost completely separate worlds. The OSS world is full of one man projects and accidental collaborations of a few individuals. They tend to organize as loosely as they can get away with: no sprints, no product owners, no backlogs, no stickies on a wall, none of it. Of course quality varies wildly as well and there tends to be some natural selection there where mediocre stuff simply doesn't survive very long while the good stuff tends to gain popularity and community.

Mediocre teams often need a lot of time to grasp requirements and to translate their poor grasp of the domain into sensible action. Something that is immediately obvious to a skilled engineer might take a room full of junior developers weeks to wrap their heads around. Agile is basically about smoothing this process and it introduces a lot of checks and balances based on the empirical notion that most mediocre programmers do a pretty shoddy job if left to themselves. That's good because it allows you to get stuff done with reasonable amounts of resources. Rock star programmers don't grow on trees and most businesses have to deal with the reality that the vast amount of their work force is not able to operate independently in a meaningful way. Agile is there to facilitate people working together more effectively and it sort of works well for a some people.

However, if you are a startup and your staff is mediocre, you are going to fail no matter how good your process is. That's why most successful startups start out small with a handful of really good people that generally don't need a lot of ritual and guidance to get going. Provide cash, pizza and cafienated beverages and they'll do the rest. That only gets you so far of course.

There seems to be a lot of dogma and myths surrounding agile. It's almost like mass hysteria. I've seen rooms of engineers citing success stories to each other that wouldn't survive the scrutiny in a decent peer reviewed scientific publication. Just because a lot of people believed the world was flat in the 15th century didn't mean they were right. Just because a lot of engineers have convinced themselves that scrum is the best thing since sliced bread doesn't mean it is.

Don't get me wrong, I have nothing about the pragmatic application of what are essentially sensible practices but what passes for methodology in the agile community is in my view 95% completely baseless empirically unfounded bullshit. And having a background in doing software engineering related research, my bullshit detector tends to be pretty good.

I've seen more than a few scrum and agile deployments go spectacularly wrong. You never read about such projects but they are all over our industry. You probably worked on several as well. The success stories always are about a team of developers in need of change and invariably they are coached to success by a lone cowboy agilist from a respectable consulting company that came in and save the day. We refer to that as seagull management where I work. Somebody comes in, shits all over the place and flies away before he can be held accountable. We've had a few bad cases of our management drinking to much of the coolaid and spending tons of money on consulting.

Ultimately, you need good people and you can't fix the fact that you don't have them by pretending to be agile.

I think agile often creates an illusion of professionality that can lead people to collectively believe that they are doing pretty OK and have reason to be proud of themselves while in fact they are progressing at a snail pace putting together bog standard database applications

I couldn't agree more. I'm not against agile. The best job I've ever had was done using a very agile approach, although that's not what we called it at the time. It was just our natural way to work on a very small team of extremely bright people. More recently I've seen agile applied as a rulebook that guides a rigid process for a middle of the road and much larger team. The development progress of the application has slowed to a snail's pace over the last two years, buy hey, at least we're agile! Right?!?

People matter. So many businesses repeat that to themselves and spout it as a mantra but their actions just don't match their words. People are SO much more important than process that it's not even funny, but businesses think they can overcome average people with a great process. It's a joke.