Game Theory in Management

I was casually watching a Star Wars movie marathon a while back, and was struck by how often the practitioners of the Dark Side of The Force (“Sith”) would articulate a desire to complete the training of the young Jedis, Anakin and Luke Skywalker. What was so obviously missing in their nominal Jedi training? It became clear that they weren’t really interested in conveying advanced techniques, but instead wanted that training to go in a dramatically different, sinister direction.[i]

Meanwhile, Back In The Project Management World…

And now, on a completely different subject, in last week’s blog I discussed a potential paradigm shift in Project Management along the lines of moving away from predictive models and towards adaptive ones. I conceded that a complete abandonment of predictive models was highly unlikely, due to the fact that government-issued contracts tend to rely heavily on projects rarely overrunning their original baselines, and government-funded projects represent a significant part of the PM community that expects, or even demands, advanced proficiency in Project Management.

With this caveat in mind, I began to reminisce about the kind of project controls analyst I was when I could be considered a young Padawan learner “new practitioner” (ProjectManagement.com’s theme for January). When I showed some aptitude for Project Management in the company I worked for straight out of college, I was sent to a couple of week-long courses in Cost/Schedule Control System Criterion (C/SCSC). I absorbed all I was taught with enthusiasm. It was all brand-new to me, and, it appeared, fairly new to the management community at large.

After a while, though, I encountered significant issues that weren’t even hinted-at in the PM codex as-presented. After being effectively stiff-armed by every accountant I encountered while trying to do nothing more than set up a simple project cost performance measurement system, I began to wonder exactly what they thought they knew that I didn’t, and went back to school to get my MBA. A few semesters in, and I had my answer – their whole management philosophy, “maximize shareholder wealth,” was utterly at odds with the PM approach, of meeting the scope, cost, and schedule expectations of the customer. I had never seen this dichotomy articulated in any of the PM literature I had consumed, but I was also fairly certain that I wasn’t the only analyst in the world that had come to this realization. Clearly my indoctrination as a new PM practitioner had been incomplete (“Ben! Why didn’t you tell me?”[ii])

Fast forward to my more experienced years, after I had quite a few paper presentations, keynote speeches, and training track designs under my belt. I began to realize that those whom I considered my peers were becoming frustrated at the propensity of new practitioners to eschew process, only to re-encounter the entirely avoidable pitfalls that they themselves had had to overcome. It seemed to me that a large and growing percentage of the papers presented at major conferences were little more than eat-your-peas-style hectoring, on how everybody needed to acknowledge or accept the virtues of setting up a tightly-regulated cost and schedule baseline, fully integrated, don’t you know, complete with rigorous change control protocols, risk analysis, and reporting formats.

The more I thought about it, the more I realized that many of the seasoned paper-presenters were trending strongly towards the dark predictive model side of things, and that it was the Agile/Scrum crowd that represented a departure from that approach. Indeed, if it weren’t for the well-written structure of the Manifesto for Agile Software Development[iii], I seriously doubt that any if its basic assertions would be accepted today. With this as the backdrop, what can be said of the advantages and disadvantages of new PM practitioners?

Well, for starters, the newer practitioners have the advantage of not having been involved with projects during the era where United States government projects were more likely to be compelled to adopt a specific set of criteria governing how their projects would be run. With the decline of the Cost/Schedule Control System Criterion as management system absolutes, the use of the two main methodologies behind PM – Earned Value and Critical Path – had to stand on their own as information streams critical to improved project performance. New practitioners, then, have the advantage of the perspective that allows better PM information system integration than those who have been convinced that a rigorous, predictive model PM approach is a condition of doing business.

Another advantage that the whippersnappers newer practitioners have is that they naturally trend toward the adaptive models, owing to the underlying causes of the Project Manager role being nicknamed “the accidental profession.” After having successfully conquered some piece of scope that has led to more, similar work, the succeeding technician/engineer/programmer will often attract a request from their organizations to assemble a team and do more of that kind of work. At this point the newly-minted PM will rarely have a concept of the reasons behind rigorous change control procedures, or risk management approaches. They simply want to get the job done, and, I would speculate, will perceive these vestiges of the predictive model to be bureaucratic impediments to that goal, making the new practitioner more inclined to question their need at the very least, if not abandon them altogether.

One distinct sign, though, that the new practitioner is being trained in the wrong direction is if they develop a predilection for within-duel taunts, such as “Your powers have become weak, old man.”[iv] But at least the older masters don’t have to wear their hair in that weird rat-tail fashion.

[i] Lucasfilm Ltd. ; 20th Century Fox. (2013). Star wars original trilogy. [San Francisco] : Beverly Hills, Calif. :Lucasfilm ; Twentieth Century Fox Home Entertainment,

[ii] Ibid.

[iii] Kent Beck; James Grenning; Robert C. Martin; Mike Beedle; Jim Highsmith; Steve Mellor; Arie van Bennekum; Andrew Hunt; Ken Schwaber; Alistair Cockburn; Ron Jeffries; Jeff Sutherland; Ward Cunningham; Jon Kern; Dave Thomas; Martin Fowler; Brian Marick (2001). "Manifesto for Agile Software Development". https://agilemanifesto.org/. Retrieved 6 January 2019.

[iv] Lucasfilm Ltd. ; 20th Century Fox. (2013). Star wars original trilogy. [San Francisco] : Beverly Hills, Calif. :Lucasfilm ; Twentieth Century Fox Home Entertainment,

In previous blogs I have referenced the old saw of “Quality, availability, affordability: pick any two” as it applies to Project Management. This week I would also like to pull in a discussion of Thomas Kuhn’s groundbreaking book The Structure of Scientific Revolution (University of Chicago Press, 1962), and how manifestations of Kuhn’s theories have been exhibited in the introduction and advancement of Agile/Scrum PM strategies in the Information Technology arena.

Briefly, Kuhn pointed out that, while the common perception is that scientific and technological advances appear to occur at a somewhat steady pace, the opposite tends to be the truth. Kuhn actually introduced the term “paradigm shift,” and theorized that scientific advances follow this model:

Phase 1 is the pre-paradigm phase, marked by the absence of a single, widely-held theory that explains the preponderance of occurrences within the given field of study.

In Phase 2 normal science begins, where most problems can be evaluated within the hypothesis-experiment-confirmation/overturning cycle. However, should a series of problems or anomalies occur that the commonly-held theory/theories cannot adequately address, a series of cycles and then epicycles are layered onto the prevailing theory/theories to account for them.

Phase 3 is marked by the occurrence of the commonly-held paradigm proving unable to adequately address an increasing number of problems or anomalies, eventually leading to an entirely new paradigm being introduced that appears to not only explain the previous paradigm’s older problems, but also accounts for the newer issues that the old theories could not adequately explain.

Phase 4 represents the “paradigm shift,” where the new theory begins to displace the previous one(s) while overcoming the inertia inherent in the dramatic modification (or even overturning) of a widely-held belief structure.

Phase 5 is, essentially, a return to Phase 2, just with the new paradigm being used to solve problems, resolve anomalies, and predict future behaviors.[i]

So, how did the introduction of Agile/Scrum for Information Technology PM perform against this model? I believe we have yet to see the final answer to that question, since much of the literature for Agile/Scrum does not appear to present as a direct challenge to more traditional PM practices, but more like cycles and epicycles – distinct but less-than-revolutionary – changes to the current paradigm. For example, Agile/Scrum does not propose to eliminate the change control process, but to radically streamline it.

However, if there were to be a principal associated with Agile/Scrum that did represent a potential paradigm shift within the PM community, I believe that it would have to do with the fourth “value” of The Agile Manifesto[ii], “Responding to change over following a plan.”[iii] “Following a plan,” in traditional PM parlance, is “Freezing the baseline,” and informal changes to an established cost/schedule baseline are, to engage in a massive bit of understatement, strongly discouraged. Conversely, Agile/Scrum represents itself as being on the opposite side of traditional PM on a scale where the extremes are “Adaptive” and “Predictive,” with Agile on the former end, traditional PM on the latter. I have come to the conclusion that the reason Agile/Scrum was so readily embraced by the IT project management culture at large is due to the prohibitively lengthy and formalistic aspects of traditional change control, which typically involve the documented and approved establishment of:

1. a change control board, with their authorities and thresholds spelled out in precise terms,
2. a nominal schedule for reviewing baseline change proposals or requests (BCPs, or BCRs),
3. another nominal timeline for how rejected BCPs/BCRs can be either appealed or amended for re-submission, and
4. yet another timeline for how approved changes are to be implemented and reflected in the baseline documents, as well as how these changes are recorded in such a way as to establish an audit trail.

For projects like the ones that make up the brunt of Information Technology work, this process was simply untenable, which is why, for example, each Sprint Planning Meeting functions as a practical, streamlined substitute for the Baseline Change Control Board meetings. I believe that this aspect alone points unmistakably to a trend in Project Management away from the “predictive” models and towards the “adaptive” strategies, and not just for IT projects.

Of course, in the realm of government contracts, the various sponsor agencies have an obligation to provably spend the taxpayers’ money wisely, so the predictive models and traditional ways of conducting baseline changes will never be abandoned in their entirety. However, my prediction (it is, after all, New Year’s Eve) is that, as those project teams who are using the more adaptive approaches to their Statements of Work consistently out-perform those following the more predictive methods’ strategies, the adaptive methods will become more and more attractive and will gradually displace the more commonly- (and currently-) held paradigms, at which point the shift will have begun.

It will be at that point that we can recognize and confidently announce the innovative trend that changed Project Management.

[i] Wikipedia contributors. (2018, December 8). The Structure of Scientific Revolutions. In Wikipedia, The Free Encyclopedia. Retrieved 04:13, December 23, 2018, from https://en.wikipedia.org/w/index.php?title=The_Structure_of_Scientific_Revolutions&oldid=872736571

[ii] Retrieved from https://agilemanifesto.org/ on December 27, 2018, 16:45 MST.

[iii] Ibid.

I had the distinct honor of taking several classes from the brilliant scholar William C. Dowling, who taught at the University of New Mexico prior to his teaching career at Rutgers. Professor Dowling had a simple test for evaluating any given sentence’s syntax, which he nicknamed “SAS,” an acronym for Simple Analogous Sentence. Confused between “That team is the one to be avoided” versus “That team are the ones to be avoided”? Since “team” is singular, replace it with an analogous singular noun. Which sounds better, “That item is the one to be avoided” or “That item are the ones to be avoided”? The SAS method points to the clear winner.

Meanwhile, Back In The Project Management World…

I’m thinking something, well, analogous should be employed in the world of Project Management. Yes, yes, I’ve been harping at length about the need to return to some level of science in the management science world, but I readily admit that there’s an art to successful Project Management, and the SAS method struck me as an excellent tool to evaluate competing strategies. For my first eval, let’s take a look at our friends, the risk managers.

Consider the following scenario: you are in a casino in Las Vegas, or Atlantic City, and there’s a game where people can bet on the flip of a verifiably fair coin. Since the most competitive casino games (craps and black jack) give around a 0.5% advantage to the house, the coin flip is very attractive. You approach the actual coin flipper but, as you do, you notice a fellow standing next to her, wearing suspenders and round, wire-framed eyeglasses. Clearly he’s not part of the security team, but he’s not actually standing in your way, so you step up to the table to place a bet.

“Fifty dollars” the coin-flipper says pleasantly as you put down a bill with Grant’s portrait on it. She displays a silver dollar, and shows you each side so that everyone is in agreement about which is heads and which is tails. “Call it in the air!”

“Wait!” exclaims the suspender-clad fellow.

“Who are you?”

“I’m the house risk manager. Are you aware that the odds are 50% that you will lose your money?”

“Of course. It’s also 50% odds that I’ll double my money.”

The risk manager sighs, as if he is the one having to endure the statement of the obvious.

“Would you like to know the recent history of this coin’s flip performance?”

“No. That’s irrelevant, since the coin is verifiably fair.”

“But, if this coin were to have been flipped the previous five times with the same result, wouldn’t that fact be of use to you?”

“No. Again, it’s irrelevant. Don’t you have someone else to irk? I want to get on with my coin flip.”

“Call it in the air!” the coin flipper says again, as she flips the coin.

“Tails!”

“Tails it is!” she exclaims, as she places a crisp $50 (USD) bill over yours.

You scoop up your winnings, and push a $5 tip into the coin-flipper’s jar.

“Just then you changed the dynamic!” the suspender-clad fellow exclaims. “Since you probably would not have tipped had you lost, the original wager’s odds have changed, from a 50/50 chance you would either lose $50, or win $50. Now we see that it was a matter of losing $50 or winning $45, meaning that the adjusted odds were NOT in your favor.”

“Yeah, so?”

“So you should not have placed that bet.”

(Long silence.) “Thanks for sharing.”

Now consider the following analogous PM scenario, where you are about to embark on a $5M (USD) information technology project and, breaking with the organization’s previous techniques, have selected an Agile/Scrum PM approach. At the first meeting of the project team, there’s a fellow in attendance wearing suspenders and round, wire-framed eyeglasses whom you don’t recognize.

“Who are you?”

“I’m the company’s risk manager. Are you aware that the odds are 50% that you will overrun this project?”

“Historically that’s the trend. It’s also 50% odds that I’ll come in on-time, on budget, or even underrun.”

The risk manager sighs, as if he is the one having to endure the statement of the obvious.

“Would you like to know the recent history of this organization’s IT project performance?”

“No. That’s irrelevant, since I’m deviating from the usual technical approach.”

“But if this organization were to have overrun the last five IT projects, wouldn’t that fact be of use to you?”

“No. Again, it’s irrelevant. Don’t you have someone else to irk? I want to get on with my project team meeting.”

(At the closeout project team meeting, nine months later…)

“We submitted our last deliverable, and came in just $50,000 over budget. Good job, everyone.”

“How can you say that?” the (uninvited) risk manager asks. “I told you that there was a 50% chance you would overrun, and you did just that!”

“Okay, well, A, it was a 1% overrun, B, the fee more than made up for it, and, C, the follow-on work coming from this organization’s demonstrated expertise in this particular type of project has led to the creation of an entirely new division. That’s how I can say ‘good job everyone.’ And, just by the way, the only one of these outcomes that was contained in your original risk management plan was the possibility of the overrun.”

“Just then you changed the dynamic!” the suspender-clad fellow exclaims. “Since you didn’t include the parameters of those scenarios in the original risk analysis, we had no way of calculating a stochastic range of the probabilities of those outcomes.”

“Yeah, so?”

“So we had no way of quantifying that eventuality.”

(Long silence.) “Thanks for sharing.”

I’m familiar enough with GTIM Nation to know that several of you are eager to post in the comment section “(Long silence.) ‘Thanks for sharing.’” And that’s okay, just as long as we’re clear who came up with that line first.