I don't know about analogy, but anyone who suggested that we run our servers at 100% utilization would be thrown out the door. We get worried when they're running at 80%, and panic sets in at 90%.
Somewhere, somehow, somebody decided that running our people at 100 or 110% was OK.
Local optimization almost always leads to suboptimal results. In the case of software development, this typically manifests itself as a) loss due to context switching, as team members bounce from team to team, and project to project, and b) too much partially finished work in the queue, as it's damn near impossible to level resources such that any one project has its full compliment of "necessary" skill types.
--- In firstname.lastname@example.org, "strazhce" <infobox.oleg@...> wrote:
> Hi, Roy,
> --- In email@example.com, Roy Morien <roymorien@> wrote:
> > Do you have a favourite football team? Does your boss have a favourite football team?
> > Ask yourself, and your boss, and anyone else who is in a position to influence how you work, what would be the effect on your football team, or basketball team ... no matter ... if the team lineup changed every game. I mean, major substitutions, not just because of injury or dropping a player out of form, or whatever.
> > Can you imagine how poorly the group of players )I hesitate to still use the word 'team' here) if that happened. Lack of cohesiveness, lack of team spirit, lack of game plan, lack of 'rehearsal' and training together.
> 1. I'm with you on the team spirit topic. But what about utilizing people effectively? What if you have work only for 2/5 of your HW people? What other 3 should do? They could help another team, which has more work then people.
> 2. Ok, now I'll play devil's advocate: if you've read The Goal, Goldratt says, that you should not maximize efficiency or utilization of specific machines, but throughput of the factory. Analogy to software?..