Farson contends that in flat structures with self-organizing teams the team member will weed out the the stronger member. In hierarchical structures people will do the Darwinian approach and weed out the weaker members.
As David Anderson notes this is obviously a problem in an Agile context. I’ve always followed the opposite approach. My two rules are:
- At least two developers per project. One developer is just too risky and doesn’t share knowledge.
- If possible at least one tech lead per project who’s top responsibility is to bring up the rest of the team. As developers grow in experience this is becoming easier to cover.
I haven’t seen developers go after the stronger team members on our Scrum projects. I’ve actually seen more of the Darwinian behavior. What I’m really looking forward too is I’m able to staff whole project teams soon with strong experienced developers. I just need to convince senior management not to swap out 80% of the technologies everyone’s mastered again and start much of the learning process over.
As an old fan of Survivor I can see that given the context of a million dollar prize cooperation tends to lead to voting off the stronger members first where possible, because they’re real threats. I just don’t think that’s how Agile projects usually work.
Considering contrarian ideas helps expand your vision, so it might be worth picking up the book. Some crazy ideas really work better than expected.