RE: [scrumdevelopment] Can you do an Agile project without having a self-organising team?
- On the subject of references about high perorming teams etc., I thought that there might be some small interest in part of a conference seminar presentation I made in Bangkok recently. I have attached the P/P slides.
Date: Mon, 15 Jun 2009 17:27:00 +0200
Subject: Re: [scrumdevelopment] Can you do an Agile project without having a self-organising team?
I just recently finished reading "Wave Rider" by Harrison Owen. Highly
One of the interesting points he makes is that whenever in his
research he looked at a high performing team, he also found a highly
self-organized team. Sometimes fostered by the powers that be, often
despite them, working in the underground.
2009/6/15 Pablo Emanuel <pablo.emanuel@ globo.com>:
> most people will probably say no, but, IMHO, the benefits from Agile come
> 1st) Iterative and incremental development - deliver early, deliver often,
> get constant feedback
> 2nd) Good engineering practices - TDD, refactoring and the other good
> practices reduce the cost of change, enabling the iterative approach
> in a very distant 3rd place come self organizing teams (it's almost a
> by-product of the first two).
> Pablo Emanuel
> 2009/6/15 pieter.vandriessche <pieter.van.driessch e@...>
>> Or: are self-organising teams really a must to do Agile?
>> Or: can you do Agile in a command and control structure rather then
>> ------------ --------- --------- ------
>> To Post a message, send it to: scrumdevelopment@ eGroups.com
>> To Unsubscribe, send a blank message to:
>> scrumdevelopment- unsubscribe@ eGroups.comYahoo! Groups Links
Click Here View photos of singles in your area
- Hi Ilja,<<it seems to me that all of your examples actually are those of teams that can't handle effectively because they don't have the knowledge or experience necessary. I don't see how self-organization actually is the impediment.>><<On the other hand - if the SM has the necessary skills and knowledge to not be short-sighted - why do those skills and knowledge need to come from the outside? What if that same person where not a SM, but simply part of the self-organizing team?>><<- In the same vein, I don't see how a different form of organization would be likely to make your sports team more successful. Also, if it's that unsuccessful, what is keeping its members still spending time for it?>>To optimize any system, you need to:1) Analyse the current structure to identify points that need to be improved2) Propose actions to address the issues identified3) Implement those changesto accomplish 1 you need:1.a) access to metrics (feedback) about the system's output1.b) analytical skills (to do causal analysis - Ishikawa diagrams, 5 Why's etc)to accomplish 2 you need:2.a) process redesign skills (bottleneck elimination, etc)2.b) specific problem-area knowledge to identify alternative solutionsto accomplish 3 you need several things, mainly power and motivation.So, skill and knowledge are surely part of the equation but not everything. Iterative practices in general, and communication-intensive practices such as Scrum in particular, add a great deal to item 1.a. Experienced process-focused people (internal or external to the team) are necessary for 1.b and 2.a. Most experienced teams should already be proficient in 2.b. Most probably you will need someone with sufficient power and external to the team to accomplish 3 fully, and there's a good probability that the team alone won't do 1.b effectively, since self-anlysis always tend to be biased.To make a team fully and efficiently self-organized you'd have to:A) Make the team able to self-analyze as rationaly and dettachedly as an external agent wouldB) Develop in them the process-oriented thinking and skillsC) Give the team the necessary empowerement. The key point here is that the power is given to the team and not to the individual, e.g. if the team decides that a member should clean the floor, the individual member must submit to the team's will.(Responding your question about my basketball team, we probably lack all three, but C would definitely be a huge issue)<<Anyway, I'd be interested in knowing more about what you think *how* the key factors that lead to high performance also lead to self-organization.>>Constant and frequent feedback, knowledge dissemination, and frequent cycles of organization optimization are factors that lead to high-performance, and they naturally lead to B and to some extent to A above.<<As you care about robustness - you probably researched Complex Adaptive Systems?>>Yes I did some research on that (we had a cross-discipline Math-Biology team), and teams of 5 or 7 people usually are far from being Complex Adaptive Systems (more on that below), although we should strive to replicate their good features to externally organized systems. A system can be adaptive and robust without being self-organized (or, to use systems theory terms, without having emergent structures) - for example, the formula 1 cars can automatically adjust several parameters based on real-time measurements of the track's temperature, or fuel consumption, or tires pressure, but they're definitely not self-organized.--Start of kind of off-topic discussion on complex systems --Emergent structures are higher-level structures that emerge naturally from lower-level behaviour. One example is the V-formation I've mentioned before. Another example is the carbon cycle in ecosystems - living beings don't give a damn to the preservation and cycling of carbon in the ecosystem, they just eat each other, or absorb carbon from the atmosphere to produce sugar. But thereby they not only create a high-level structure - the cabon cycle - but also continuously enhance it. [Presumably] there's no external organizer thinking "there is so much carbon in mice, let's throw some snakes to make this carbon to move", but the abundance of prey naturally leads to the increase on the number of predators. Several other local optimizations acting on a large timespan makes mature ecosystems like the Amazon rainforest to have a much more efficient and more robust carbon cycle than a simpler ecosytem, such as a cornfield, although no one (not even the inhabitants of the forest themselves) ever intended it to happen.So, there are 2 comments that are pertinent to our discussion:1) real complex adaptive systems must necessarily have many variables in order to emergent behaviour to appear (examples are ecosytems, the economic market, a city)
2) even if the system is organized by the system's own members, *its not* self-organized in the strict systems theory sense. The system is self-organized when the system is organized *by itself*, not externaly organized by its members. That doesn't seem to make much sense, but the carbon cycle example should make it clearer: if some of the beasts decided to actively optimize the forest's carbon cycle, it wouldn't be an emergent feature anymore.So, the analogy with truly self-organized systems made in this thread to conclude that self-organized teams must necessarily be better is actually a fallacy, it's comparing two different things that happen to have the same name.-- Digression ends here --Regards,Pablo Emanuel