Game Theory in Management

During the May 1916 Battle of Jutland, elements of Great Britain’s Grand Fleet fought to keep the Kaiser’s High Seas Fleet from breaking the naval blockade of Germany. Hundreds of ships were involved, but when the smoke cleared the Germans had failed to significantly change the ratio of capital ships between the two belligerents, and the blockade remained in place. One American journalist summed up the massive naval action thusly: “The convict assaulted his jailer, but he is still in jail.”

Along those lines, I think much of what passes for chapter and verse of lessons learned reports can be boiled down to its essential elements, and documented so. A few examples on the topic of the PM’s technical approach:

• On Earned Value reporting: (a) just because you are spending money on a project does not mean you are accomplishing anything, (b) we can prove it, and (c) if you tell me how much you have spent, and how much progress you have made, I can tell you within ten points how much your project is going to cost when it’s done.
• On Critical Path methodology: (a) if you are “managing” your schedule via an action item or milestone list, you’re doing it wrong, (b) we can prove it, and (c) tell me when you started, and how much progress you’ve made, and I can tell you both when your task will be completed, and how that impacts the other activities in this project.
• On Critical Chain theory: if you get done early, help any other members of the project team who are behind.
• On (modern) Communications Management theory: tell everybody you think of everything that’s going on in your project, and nothing bad will happen to you.
• On (modern) Risk Management Theory: you might want to worry about stuff that could happen to your project.
  • Hey! Risk Management also involves “Opportunity Management!”
  • Says who?
  • SAYS US!!! We’re members of the risk management community, and we’re more than a little tired of you accusing us of engaging in “openly fraudulent management science.”
  • Are you aware that none of the authoritative dictionaries of the English language make any references to “opportunity” (other than in listings of antonyms) in their definitions of “risk”?
  • That doesn’t make any difference, since we’ve arranged to have almost all of the PM organizations include our “upside risk equals opportunity” tripe
  •  in their glossaries of terms.
  • Okay, how about “you might also want to concern yourselves with opportunities that may or may not arise.” Happy now?
  • Yes, you may proceed.

The following elements are extremely common to (truthful) lessons learned reports coming from failed or overrun/delayed projects:

• We should have used Earned Value and Critical Path methodologies far earlier than we did.
• The “bottoms-up” Estimates at Completion (EACs) were always optimistic, and often catastrophically so. Conversely, the calculated EACs were always accurate, and could have notified management of difficulties far earlier than the re-estimated versions.
  • On a related note, that project controls person who told us that, in order for our EAC to be accurate, our cost performance would have to go from a cumulative Cost Performance Index of 0.7 to a To-Complete Performance Index of 1.4, and that that had virtually no possibility of happening, and that our posted EAC was therefore nonsense – well, that person was right.
• The tasks that drove most of the project’s overrun were subject to significant amounts of scope creep, and we should have listened to those PM-types who were harping about better and more detailed descriptions of the original scope.
• An “overly optimistic original estimate” explanation in the Variance Analysis Reports (VARs) of negative cost variances will only work a few times before the customer catches on that it’s really our pitiful performance causing the variance.

As with anything that gets condensed, these lessons learned can be significantly harder to digest than their more verbose equivalents. But if the lessons to be learned are couched in phrases designed to soft-pedal the truth, I have to ask: did anybody really learn anything from the past projects’ failures?

Towards the beginning of the motion picture The Return of the Jedi, Luke Skywalker, having just witnessed his mentor Yoda pass away peacefully, has a chance to confront his original tutor, Obi-Wan Kenobi. Luke had a right to be put off – he had learned at the end of The Empire Strikes Back that his antagonist, Darth Vader, was, in fact, his father, even though Obi-Wan had told Luke in the first movie, Star Wars (later renamed A New Hope), that Vader had “betrayed and murdered” Luke’s father. When Luke confronts the holographically-represented spirit of Obi-Wan with this jarring inconsistency, Obi-Wan begins his defense with the immortal words “”Luke, you’re going to find that many of the truths we cling to depend greatly on our own point of view.”

To which Luke replies, incredulously, “A certain point of view?”, a line delivered by actor Mark Hamill as if to imply “is that really the best defense the writers could come up with for this massive plot reversal?”

Of course, the character of Luke had every right to be incredulous at this turn of the narrative. Yes, the revelation that Vader is Luke’s father is considered to be one of the most astonishing reveals in all of cinema history, but there’s a reason for that: nothing, and I do mean nothing, in the dialogue of the previous two films offered up the remotest hint that Vader and Luke were even related, much less father-and-son, prior to Vader’s infamous line to the contrary. And, as Luke pointed out, Obi-Wan had told Luke a very different story about what happened between Anakin Skywalker and Darth Vader. And now that complete reversal was to be somehow explained using the “point of view” defense?

Riiiiiiight.

Meanwhile, in a galaxy very, very close to ours, and a few short minutes ago,

ProjectManagement.com’s theme for June, Lessons Learned, got me to thinking about how narratives are formed and then documented, and how sometimes these narratives are inconsistent with the facts. A few examples:

• Senator Joe McCarthy never wrongly accused anyone of being a communist or communist sympathizer.
• Kirk never said “Beam me up, Scotty”, and Lassie never saved Timmy from a well.
• Julius Caesar was never emperor of Rome (he was assassinated the morning the Senate was to vote him that title and those powers).
• There is absolutely no evidence that Marie Antoinette ever said of starving Frenchmen, “Let them eat cake.”
• Richard III was only 32 years old when he died at the Battle of Bosworth Field, and never said “A horse! A horse! My kingdom for a horse!”

I could go on (and often do), but you see my point. These narratives take on a life of their own, and any hard data that challenges (or even overturns) them is ignored or disputed.

This being the case, what are the chances that your organization’s lessons learned documents flawlessly reflect the precise nature of the causal elements of its previous project successes and failures?

I would guess the answer to that is “pretty darn low.”

Can “Point of View” factors even be quantified?

 It’s been my experience that the people who prepare such documents can’t even articulate the difference between proximate cause and material cause, let alone in the highly complex/near chaotic environs where Project Management actually happens.  The Big Apple website has a list of the six phases of a project (which were probably originally developed in the early 1970s):

1. Enthusiasm
2. Disillusionment
3. Panic
4. Search for the guilty
5. Punishment of the innocent
6. Praise and honor for the non-participants

Since almost all humor must have at least a little bit of truth in it, and since most PMs I know would find this list at least a little bit funny, ask yourselves:  why are numbers 5 and 6 funny? Could it be that it’s a common foible of human nature to mis-remember what actually happened in a given project’s history to promote a narrative at variance with the facts?

If the answer to the previous question has even a possibility of being “yes,” then my readers should keep something in mind: that moment at which you are in the Project Management-equivalent position of being on a narrow ledge above an abyss, with your ultimate adversary preventing any apparent escape or remedy, is a really bad time to learn that key assertions from your organization’s lessons learned documents may be 180 degrees out-of-phase due to “a certain point of view.”

During my last semester in High School I narrowly missed out on a drama scholarship to the University I was going to attend anyway.  Of course, to even be considered for such a prize, one had to spend a lot of time on-stage, in costume, pretending to be a different person, and speaking that person’s words as written down by the playwright. In order to perform well, the people on-stage had to have a good idea about what their character was about, why they thought the way they did, what made them tick … in essence, we needed to know the characters’ motivation for the role in the action being portrayed.

Flash forward to my time spent behind a podium, presenting papers or track courses on the topic of Project Management. As with being on-stage, a lot of people would pay good money to hear what I and the other conference/seminar presenters had to say. Unlike the stage, they weren’t there to be entertained, and we weren’t speaking other writers’ words. The peeps were there to hear about recent or overlooked research or advances in PM, and we selected presenters were there to give that to them.

Well, maybe.

Why are we here? (The non-existential version)

I found myself thinking about those radio advertisements where a person offers to teach you how to make a killing in the stock market. These “experts” claim to have discovered a highly formulaic approach to trading stocks and, for a fee, will share this approach with potential investors. Why would they do this? I mean, if you know a secret to making a lot of money in the stock market, why don’t you just continue to make a lot of money in the stock market? And, depending on the formula, one gets the sense that, if everyone was employing the identical approach, this approach would quickly cease to be lucrative. Stock markets, after all, are not money printing machines. Not everyone can buy low and sell high. So, why would these guys offer to sell their secret, when doing so is almost guaranteed to make it less effective the more the secret is disseminated?

Similarly, if I come across an idea that makes my organization far more effective in bringing in projects on-time, on-budget, why would I want to share this idea, particularly if my organization’s competitors might have employees in the room at the time I share the insight? If that’s not sufficiently suspect, consider this: I’m paying for the opportunity to help enlighten perfect strangers. Does this seem right?

This is the kind of effect that occurs when management science swerves away from genuine research that can establish causality. Consider the opposite alternative, that of a seminar presenter who does not have a genuine insight, but wants the audience to believe that he does. He will point to a particular process as representing an enlightened approach to PM, but will (predictably) be light in proving its connection to actual project success. Clearly there’s some other motivation in play here, and I think I know what it might be.

Always suspect free-lancing researchers

I think that there’s this idea out there that there are roving bands of powerful executives, each of whom desires to acquire a PM expert, and will pay handsomely for the right one. Such an idea is anathema to performers, but it acts like catnip on processors. I believe that the processors think that (a) if their pet process becomes widely accepted, and (b) they have become strongly associated as the subject matter expert in that particular process, they will (c) reap substantial benefits. Hence the seemingly endless paper presentations on some sort of process that can’t seem to firmly establish itself as the proximate (or even material) cause of project success.

Of course, it’s always dangerous to reverse-engineer motive based on outward behaviors, and I could easily be wrong in my speculations.  But ask yourselves – if you had a formulaic approach to consistently out-perform the competition in anything, and being known as the expert for a particular process was not your motive, how eager would you be to blab about it?

Aristotle’s “ingredients for persuasion,” or “appeals,” the basis for creating an argument, were as follows:

• Pathos, from where we get the modern term “pathetic,” is an appeal to feeling, or emotion. The tip-off: use of the term “I feel,” particularly “I feel very strongly about this…” Pathos is not considered a valid argument.
• Ethos is an appeal to a person’s level of expertise or authority, either the speaker or some other individual. It’s not necessarily a poor type of argument; but, as we will see, it does have its pratfalls. And it is most certainly not…
• Logos, where we get the current term “logic,” is an appeal to verifiable facts and structured arguments, and is the foundation for the scientific method. The verifiable facts are observable and repeatable in an experimental setting (or, at the very least, agreed to by all participants in the debate), and become premises. Conclusions follow from the premises based on structures that can be verified as valid, such as:
  • All As are Bs.
  • No Bs are Cs.
  • Therefore, no As can be Cs.

Valid conclusions can then become premises in more advanced or complex logical arguments.

Where have we heard this before?

My regular readers are well aware by now that one of my recurring themes is the hazards being introduced into the project management community by what I perceive is a move away from management science based upon, well, actual science, and towards hypotheses predicated more on traditional business theories, on the opinions of those either in academia or PM practitioners. Such ideas are based on ethos, and not pathos. These practitioners may or may not have a valid understanding of basic causality, much less the precise reasons behind their projects’ successes or failures. Look at the presentation schedule of a typical Earned Value Management System seminar, and you will see a cavalcade of talks and paper presentations that are little more than kabuki theater versions of ethos-derived stories of how such-and-such a project was successful due to the team’s observance of traditional PM business theory, or else presentations with a content that could have been written 30 years ago. There’s rarely anything new in the way of datum or research proffered.

I was once involved in an effort to create a guidance document on a certain aspect of cost performance measurement, and one of the other contributors tried to insist that his ideas should be considered preeminent because he had done some PM-related work outside of North America, and should, therefore, be considered an “internationally recognized expert.” I swear I am not making this up. No new research he had conducted, no possible technical challenge to existing theory, none of that. He had his passport stamped, so the rest of us were supposed to shut up and put him in charge, I suppose. It represented the most blatant example of self-importance-driven PM hypotheses advancement that I had ever observed; sadly, it wouldn’t be the last.

About that superfluous complexity…

Two weeks ago I blogged about invalid theories being advanced from traditional, academic sources, predicated on the asset managers’ take on the world of business. Last week I complained about invalid theories being advocated by people calling themselves project management experts. What do these drivers of needless complexity have in common? Their appeal, not to facts and logic, but to their authors’ own level of expertise, however artificially derived that level may be.

When David Christensen and Scott R. Heise published their findings about Cost Performance Index stability,[i] they drew their conclusions from data gleaned from hundreds of projects. Now, that’s real management science. Quick – when was the last time you heard a communications expert, while insisting on the “engagement of stakeholders,” appeal to any real data? Or even a citation of the number of successful projects that did so, as compared to unsuccessful ones that did not, while even making a token attempt at controlling for the dependent variables? (Can a level or degree of “stakeholder engagement” even be objectively measured?) I never have. Not once. Nor have I ever heard or read any similar attempt by risk management experts as they insist that all PMs include a robust risk management program, or those who demand “bottoms-up” Estimates at Completion, or of a dozen other PM-flavored hypotheses being pushed by these so-called experts. They never seem to base these ideas on facts or logic; rather, they assert these appeal-to-authority ideas, and then pretend that anyone who opposes them or their ideas is some kind of rube. Since these hypotheses are not logically derived, the opportunities for them to be inconsistent (or even contradictory) lead directly to added levels of superfluous complexity within the commonly-held body of PM knowledge.

And, just to be clear, this is not how any science (particularly management science) is advanced.

It’s how snake oil is sold.

[i] Christensen, David, and Heise, Scott R., Cost Performance Index Stability,  National Contract Management Journal, 1992.

In last week’s blog I discussed the phenomenon of the codex of commonly-accepted management theories being, well, wrong in several of their assertions, and of the outright impossibility of getting those theories’ adherents to see their error. Frustrating, isn’t it? I mean, these so-called experts in business and management theory can be shown instance after instance of real-world contradictions to their theories, and they just won’t budge. If you happen to be in one of these “expert’s” classes, and you, as a PM-type, provide such evidence, they won’t listen to you – they’ll just flunk you. If this “expert” is a colleague, and you happen to be sitting across from him in the board room when a particular point of managerial technical approach comes up that absolutely should not be performed their way, and you say so, they will attack your intelligence, credibility – heck, even your character.

Wouldn’t it be so much better if those calcified hacks had just a modicum of the flexibility that we PM-types exhibit on a regular basis?

Hey! Who’s That In The Mirror?

Well, brace yourselves dear readers – the other reason (from last week’s blog) that PM is so difficult is because of what we do to ourselves in accepting uncritically (or even perpetuating) our own version of flawed management science hypotheses.

Probably the most pernicious recent addition to the cavalcade of these PM hypotheses is the push to align the basis of estimate’s (BoE’s) line items with their equivalent entries in the general ledger. It’s a commonly-known axiom among real Earned Value specialists that comparing budgets to actual costs is meaningless, and doing so at a greater level of detail is not only just as meaningless, it takes a lot more time and effort to accomplish.

Can I prove it’s meaningless? Absotutely.

Consider a project that was originally estimated at $100,000 (USD), with the split of $75K in labor, $25K in equipment. When the project is completed, it actually spent $25K in labor, and $70K in equipment. A properly-functioning Earned Value system would have indicated the project’s mild positive performance throughout, but the “management information system” that tracked the original estimate against the actual costs on a line-item by line-item basis would have been throwing up red flags all over the place, despite the fact that the project came in under budget!

Comparing budgets to actuals doesn’t generate a Cost Variance – that’s done only by comparing Earned Value to Actual Costs. It creates a spend variance, which is truly irrelevant information. But the guidance-issuing entities that insist on this kind of “analysis,” which results in no usable information whatsoever, have also pushed the notion that too many start-to-start relationships in a Critical Path schedule network will allegedly hide true performance issues. I swear I am not making this up. These guidance-issuing entities will insist on the generation of provably invalid data streams, and then turn right around and impugn the integrity of a valid one if it shows a singularly irrelevant characteristic. Does this seem right?

Irrelevancies! Irrelevancies Everywhere!
And don’t even get me started on the time-phased estimate to complete (ETC).

Or the bottoms-up estimate at completion (EAC).

Or the so-called 80% confidence interval on cost or schedule baselines.

But it must be noted that we PM practitioners are doing these things to ourselves. These guidance-generating organizations are churning out this stuff under the auspices of advancing project management, and they are doing so through means that would never pass muster with true peer-reviewed PM evaluation venues, such as The Project Management Journal.

However, in order to assure that I don’t turn in to one of these calcified hacks, I would like to invite my readers to send me any instance of where comparing budgets to actuals was (a) relevant, and (b) more so due to being done on a line-item basis. Fair warning – please include specific assertions, and assemble them into a valid, logical argument.

Otherwise, my frustrations with the abandonment of valid management science scholarly research will soon turn me into a calcified hack.