Do or Do Not Do. There is no Try.
Master Yoda may well be an Agile master. His sage advise came back to me again in a recent discussion with some colleagues about adopting a stand-up practice for the whole company.
Quite often I find that people try to adopt Agile practices a little at a time - in this case do it with a few people to see if it works rather than the whole team. The lets-try-it-and-see-if-it-works approach. The problem is you're not actually doing the real thing, but a prototype or a simulation. And the commitment is not there in this attitude. Furthermore the actual value is often found in doing-the-whole-thing-for-real and there is no middle ground that is informative.
Its a bit like saying your going to learn to dance but only committing your left leg to trying it.
I can appreciate the fears associated with adopting a new practice especially when transitioning to an Agile approach. However there is usually a fear of "doing the right thing" at the base of it. Is it right for us? How do you do it? What if its the wrong thing? Lets try it in the small and then if its wrong the damage will be less.
So by the dancing analogy this would amount to only committing your left leg because you're afraid that you might look silly, you don't really know how to do the dance, and you might hurt yourself if you tried the whole thing - but you're willing to risk a leg.
There is no right or wrong in practice based approaches, only more valuable or not - and you can't figure that out unless you really do the practice.
I believe the lets-do-it-and-restrospect-it-later approach is best. Do the thing which believe will give you the highest value. Commit to doing it 100%, unquestioning, non-analyzing, applying yourself, by-the-book, this-and-this-only. But agree at the start how long to do it for and when to retrospect upon the practice. So for example, agree to try the stand-up practice for a given team for four weeks and the retrospect on whether it is as valuable as you thought it might be and how you might change things to improve it. During the 4 weeks just do it. Don't discuss it, don't argue about it, don't try to optimize it. Then do a brief retrospective on it.
So this is like committing yourself to learning the dance, whole heartedly and with body and soul. But giving yourself a time limit when you decide whether your getting what you want from it.
This is one of the reasons why I often bang on about Agile being about discipline. It takes discipline to commit to something and not analyze whilst doing it, but then set aside time to reflect. This attitude in itself is a valuable part of being Agile, and especially being eXtreme.