The New Scrum Game
Ken Schwaber and Jeff Sutherland created the Scrum Framework based on The New New Product Development Game (TNNPDG) & empirical process control in the 90s. It was oriented around a team building a product. It's origins are illustrated in Figure 1.
Figure 1: The origins of the first incantation of Scrum for software development.
It's success has led it to be used in IT and across organizations. Unfortunately, this first incantation of Scrum for software development was not designed for this expansion.
In the 20+ years since Scrum’s ascension, several useful methods – Flow, Theory of Constraints, Kanban and Lean Management have come to the forefront. I believe it is past time to create a “New Scrum Game” from The New New Product Development Game along with these new methods to create a team-level framework that can be used for teams doing either product and IT. We should also include double-loop learning and the "sister" article of TNNPDG, Toward Middle-Up-Down Management: Accelerating Information Creation, which lays out the role of management. This modern transposition of the concepts Hirotaka Takeuchi and Ikujiro Nonaka were espousing creates a new, more modern Scrum, and which can be used for teams in small and large organizations.
This "New Scrum" doesn't require immutable roles, rules, artifacts or events but can be tailored for what teams in an organization needs. It can be based on time-boxing or flow. It also can provide guidance when fully cross-functional teams can't readily be achieved
This "New Scrum Game" already exists in the combination of Disciplined Agile Delivery and FLEX's Team-Agility. It’s not a variation of Ken and Jeff’s Scrum, it’s simply a modern version of the Scrum Hirotaka Takeuchi and Ikujiro Nonaka (authors of TNNPDG) had in mind.
Figure 2 illustrates how The New Scrum Game derives from the modern thinking that Flow, Lean, Theory of Constraints and Middle-Up-Down Management provides us. Note that it is not a derivation of the old Scrum for software development but is a new codification on its own.
Figure 2: The New Scrum Game Stands on its own.
Let's take a look at some of the differences:
Systems Thinking. Systems-thinking is one of the foundations of The New Scrum Game. This means that systems inform behavior. We can't just say - use cross-functional teams because that might not fit well with the system within which the team exists. We must also consider the culture of the organization because how they will react to the Agile transformation.
An understanding of how things work. From Jeff Sutherland's Scrum papers, he describes how Scrum treats major portions of systems development as a controlled black box. 20 years ago this was the best we can do. But with modern Flow and Lean thinking we know that we can improve our process by managing work in process to avoid going over our capacity, by managing our queues, eliminating handoffs and lowering delays. We can anticipate many of our challenges and accurately predict many actions that would be improvements by attending to our value stream.
Plan-Do-Study-Act (PDSA) vs Inspect and Adapt. Having empirical process control be the main model for improvement leaves out an understanding of why things work the way they work. Old Scrum works as a black-box model where teams see what happens, inspect and then adjust our understanding of what is happening. With this deeper understanding we act by improving our practices. Then we repeat the cycle.
Flow vs iterations. There are several factors as to whether you should based your product development and/or IT efforts based on a time-boxed or flow approach. Nothing in The New New Product Development Game requires time-boxing (iterations). This addition often makes it easier to explain, and possibly do, Ken/Jeff Scrum, but it is not always the best way to do it. The New Scrum Game allows for either. Understanding Lean and using cadence for coordination makes it so time-boxing is not required.
Roles. The "one wringable neck" of Ken/Jeff Scrum does not sound very respectful. It also takes away from the team owning the product. The product owner as a person separate from the team is the anti-thesis of the lesson being taught in TNNPDG. For teams in IT is may be a useful concept, which is probably why it came into being. A problem with specified roles is that in many organizations, people have become attached to their roles. This is especially true when there is no career path in the organization for the Ken/Jeff Scrum roles.
Management. Management was virtually excluded from the Agile Manifesto, not being mentioned even one time. While this was generally true of the Agile community, it was particularly true for those following Scrum, calling managers "chickens" and not committed to the team. Ironically, Towards Middle Up Down Management was almost a companion piece to The New New Product Development Game but has been ignored by the Scrum community.
Structure. There is no question that cross-functional teams are good. But what if they are either difficult, or even inadvisable to form. The insistence on having them often leave people in question of what they should do if they don't see how to form them or not.
Flow thinking. Being based on Flow includes attending to cost of delay and the use of Minimum Business Increments (MBIs) as a core driver.
The New Scrum is not Professional Scrum with Kanban
Since originally posting this I’ve had comments that this is the same as Scrum with Kanban. If it were I wouldn’t have bothered to write this. There is a big difference between putting Kanban practices into Scrum and having Scrum practices in Lean. In fact, Kanban University Kanban (formerly Lean-Kanban University) is based on Theory of Constraint, not Lean.
Scrum is not a systems-thinking approach. It doesn’t attend to culture either, but requires teams change roles and work with iterations and cross-functional teams. Using Kanban practices inside Scrum is a great step forward, but not the same as a Lean-based approach that has Scrum and Kanban practices in it.
In Summary
Be clear that The New Scrum Game is not a derivation from Ken and Jeff's Scrum. It is a new transposition of the original Scrum described in The New New Product Development Game that incorporates modern thinking.
software engineer specializing in software engineering process, methods, practices, and professionalism
4 年Can you elaborate on two things? First, the statement "Understanding Lean and using cadence for coordination makes it so time-boxing is not required." It's not clear to me what the difference between doing things on a cadence and timeboxing is. It seems to be a nuance since timeboxing leads to a cadence. Second, the cases where a cross-functional team isn't desired or good. I can think of cases where you want to supplement a cross-functional team in one of a few different ways, but I'm struggling to think of a case where a cross-functional team isn't the desired state.
DevOps & Agile Engineering Senior Leader
5 年Looks interesting Al! Some of it reminds me of where Mike Beedle was going with 'Enterprise Scrum' a few year back, I don't suppose you'd care to compare & contrast the two just a bit? Also looking fwd to your new acronym for the equivalent of the SoS meeting -- TNNSoTNNS :-)
Agile Wingman
5 年Thanks for reminding me of reading The New New Product Development Game again, many years ago now. My read-through confirmed my memory of the article, Scrum is mentioned once, in a sub-heading and the authors didn't indicate to have any kind of today's Scrum in mind. Sutherland's and Schwaber's Scrum is inspired by that article, not a further development of it. And Scrum, the framework, has developed vastly since the original presentation at the OOPSLA conference '95. Al, present a reference case with an actual user describing in own words why your framework is superior over something else. That would be evidence talking for itself, rather than you trying to pursue the world why something you got is better than the framework that conquered the world. Until then, my humble suggestion, put your energy with your clients and try to build this fabulous reference case.
This is a nice roundup for a multitude of posts from the recent weeks. Being in the agile game for some 10 years, I can say my scrum practice had undergone many changes and when coaching the are many differences depending on the organization and culture. In all these places things evolved. Scrum has evolved and the actual practice is not the framework but the interpretation of everything useful including DAD, SAFe, LeSS etc.. these indeed have good ideas for scaling up. Your FLEX book provided me with some valuable insights that helped me coach teams to seek flow @ scale and cooperate more. SAFe places good emphasis on handling the delivery pipeline and DevOps. LeSS provides some good inspiration for Product ownership. Since the very beginning I used TOC (a Goldratt fan), Lean, Flow and Kanban to improve the scrum based practice. I guess what I think here is that the New Scrum, is an evolution of the classic scrum that improves on it by adding the experience from other practices into a specific practice that not the framework but is inspired by some. 20 years of evolution! Regarding empiric control, I guess it’s the continuous improvement framework that makes it all work! Thanks for the article!
Author, C-Level Mgmt, R&D, Product Development
5 年Scrum is a command and control overlaid to a waterfall model where rather than primarily managing risk it focuses on client reqts. The post, standup, and task pulls are further details. But no magic. Talented people, good communications, a gifted chief designer with vision and drive, realistic goals, and adequate resources and expectations... The “agile” and “scrum” methods seem overhyped, though many manufacturers and operations thrive on it.