Game Theory in Management

It was another dark and stormy night. I was staring at the stencil across my frosted glass office door, eyE etavirP, yrrebpsaR .t yelnatS. My secretary, Anne, had just left, saying “Good night, sleep tight, and if you don’t pay me tomorrow I’m going to kick your &^%,” when the phone rang.

“Stanley? It’s Deborah. You’d better get down here.”

Deborah was my inside source from Monolithic Corporation.

“Why, what’s going on?”

“The big wigs are bringing in some politicians, so it has to be hot.”

“How can I get in?”

“Disguise yourself as one of the custodians.”

“Don’t they need badges, too?”

“Yeah, but the guards never look at them. For some reason, Monolithic treats its cleaning staff rather poorly.”

“Most poorly-run organizations do. Okay, I’m on my way.”

With my cap pulled down low and wearing coveralls, I pushed a trash bin cart straight past the guards. Deborah was right – they barely glanced my way. I headed towards the hall where Deborah’s office was. Just a few feet away, I heard a guttural voice behind me.

“Raspberry! Dang it! You come here!”

I froze in place.

“Yeah, I’m talking to you. Come here!”

I slowly turned, still keeping my gaze low.

“How can I help, sir?”

“Somebody spilled some raspberry preserves in this break room, and if you know what’s good for you, you’ll clean it up RIGHT NOW!”

“Just let me fetch my stain remover, and I’ll get right on it, sir.”

The Monolithic manager huffed off just as Deborah opened her office door and beckoned me inside.

“Get to room 54B; it’s in the basement right next to the Hatfield Conference room.”

“The what conference room?”

“It was named after an obscure ProjectManagement.com blogger. Behind the velvet painting of Elvis there’s a peep hole you can use to listen in.”

As I shuffled in to basement room 54B, I could hear an animated discussion already going on. I closed the door, and pushed the velvet Elvis painting aside, and listened.

Voice #1 (guttural, so I assumed it was Monolithic’s president): “So, as you legislators can see, we don’t outsource anything we don’t have to, meaning more dollars stay here in this country.”

Voice #2 (a bit more earnest, so I assume it was one of the legislators): “But is that the most economical approach? If you were to outsource some of your parts or services, wouldn’t that reduce the costs on your government contracts, and then the taxpayer?”

Voice #1: “That’s not what we’re here to talk about. I want you three to introduce and push legislation that grants preferred status to companies like Monolithic that refuse to outsource, no matter the increased costs.”

Voice #3 (not guttural, but more experienced sounding than the first legislator to respond): “And why would we do that?”

Voice #1: “Two reasons: one, by pushing the preferred-no-outsourcing narrative, you can appear to be patriotic, since the funds stay in this country. Two, if you don’t, not only will we stop contributing to your campaigns, we will start supporting your opponents.”

Voice #2: “That sounds like a threat.”

At this point what I heard sounded like someone had told a really funny joke at a Darth Vader imitators’ convention.

Voice #1: “I’m not threatening. That would be un-American!”

Another very funny joke at a Darth Vader imitators’ convention.

Voice #2: “What about the intellectual forces against that meme, people like Stanley Raspberry?”

Voice #4 (guttural, but younger): “Raspberry? He’s been a pain in our…”

Here I heard a sound like a yardstick being slammed onto knuckles.

Voice #1: “Raspberry and his ilk are nothing to us. If you’re smart, you will do as we say.”

Voice #4: “I don’t get it. If your company is so big and successful, why do you need any kind of help from the government?”

Voice #1: “It’s not about our expenses. It’s about making harder for smaller competitors, who, as you know, could easily undercut us by outsourcing. Customers would flock to them in droves, so we just need you to demagogue this issue, so that people will continue to be mislead about outsourcing.”

Just then I heard the doorknob turn. I let the painting down, and began emptying the trash can as a Monolithic executive walked in.

“Get out of here, custodian!” he growled.

Deborah called me once I had returned to my office, and I gave her the lowdown on what Monolithic was up to.

“What are you going to do about it?” she asked.

“What I always do – counsel my clients to make the best decisions for them, which includes outsourcing, and watch them steadily erode Monolithic’s market share – right after I clean my office.”

In 1981, the rock group REO Speedwagon released a single entitled Take It On The Run, a song that appears to have highly contradictory lyrics. The narrative begins in the voice of a boyfriend who has access to fourth-hand evidence that his girlfriend is unfaithful. The song then jumps back and forth from his apparently believing this information (“…then I don’t want you around.”) and not believing it (“I don’t believe it, not for a minute…”), with a particular non-sequitur thrown in: “You’re under the gun so you take it on the run.”

Since my undergraduate degree is in English, I felt free to try and interpret what the “you’re under the gun” line meant. Unless we’re talking about an active shooter situation, “under the gun” generally means that the person is under extreme pressure. So, since this person – presumably the suspected girlfriend  -- is under extreme pressure, what does she do? She “take(s) it on the run.” Takes what on the run? Her love life? How does one do that, exactly? Is she running from place to place? If so, how does that help her avoid being “under the gun”? Since some variant of this line is repeated six times, the listener could probably safely assume the lyricist thought he was being clear – or else had simple given up on creating an internally consistent narrative.

By the way, Take it On the Run charted at #5 on the Billboard’s Hot 100 list.[i]

Which reminds me of our friends, the risk managers. According to Wikipedia, the definition of risk analysis begins “Risk analysis can be defined in many different ways, and much of the definition depends on how risk analysis relates to other concepts.” Really? A discipline that claims to be a profoundly necessary aspect of project management “can be defined in many different ways.”? Other definitions lack, well, definition, and even the PMBOK Guide® has some of its mushiest language in the risk management section. I maintain that if a concept cannot be clearly and precisely articulated as to what it is, and what its outputs are, then it’s probably an invalid concept, and much sound and fury must be created to obfuscate this fact.

Since what it is is difficult to state, (I defy anyone to make sense of the following: “The goal of risk planning is to establish how the overall risk management will be conducted for the project. The time spent, the role and responsibilities, and template formats of the reports should be all established in this process. Once the preliminary work is done, identifying, analyzing, and adjusting for risks can be done.”[ii]), let’s see if we can shed some light on the topic based on what risk management is not.

1. It’s not a performance measurement system. Risk management cannot tell a PM anything about how his project is performing.
2. It is not a way to quantify the future, since the future cannot be quantified.
3. Much of the input for risk analysis is derived from interviewing the Control Account Managers (CAMs) for their estimate (guess) of the odds of something bad happening, and its estimated (guessed) financial impact, it could be said that risk analysis is an attempt to quantify managerial experience. They can attempt away – experience can’t be quantified.
4. Much has been written about how “positive risks,” or opportunities, also fall under RM’s purview, but this is also just so much silliness. No dictionary outside the reach of the risk aficionados has any reference to “positive” or “opportunity” in their definitions of risk at all – in fact, they are universal in the opposite.
5. So, if RM is an information tool, what information do its analysis techniques pass along to PMs that influence their decisions? The estimated (guessed) odds of something bad happening? How does that help, exactly?

So, risk managers are under extreme pressure (“under the gun”) to present to the management world a coherent structure where their analysis techniques provide verifiable value, when they, well, don’t. Therefore they’re forced into a position of ensuring that the exact definitions of their concepts and techniques are sufficiently obscure as to present a moving target to anyone who would attempt to so define them (they “take it on the run”).

Since RM is a multi-billion dollar industry world-wide, I suspect it has pulled in more money than REO Speedwagon made in their careers. But as for its validity? I don’t believe it, not for a minute.

[i] Take It on the Run. (2016, July 4). In Wikipedia, The Free Encyclopedia. Retrieved 02:08, July 10, 2016, from https://en.wikipedia.org/w/index.php?title=Take_It_on_the_Run&oldid=728209354

[ii] Project Management/PMBOK/Risk Management. (2015, February 25). Wikibooks, The Free Textbook Project. Retrieved 01:55, July 10, 2016 from https://en.wikibooks.org/w/index.php?title=Project_Management/PMBOK/Risk_Management&oldid=2770679.

Politicians and other demagogues have given the business act of outsourcing a bad name, due, at least in part, to the infernal human tendency to overlay some sort of moral judgement on economic transactions that almost always do not involve the person making the judgement. If the person who, say, processes the company payroll is herself an employee, then that is somehow seen as more righteous than if she works for a company that performs the payroll processing service for many companies, particularly and especially if the payroll processing service is headquartered in a different nation. In the case of the latter, some ignoramuses will even assert that such a decision – to outsource the payroll function – is tantamount to an act of treachery, and those cats will step on your face, slander your name all over the place.

But anybody with a room-temperature IQ who gives this even a quick analysis will realize that such assertions are absurd. I’m not a master of internet communications and attractor of PM minds and advertising revenue: I’m just a blogger. In a sense, I have “outsourced” those functions to ProjectManagement.com. Pea farmers have no idea what type of package their delivered crops will use upon their delivery to the market – they’re more concerned with creating and maintaining the best environment for growing peas. If they were to attempt to control all aspects of growing, processing, packaging, transporting, and selling their peas, you can safely bet that the cost of those peas will be significantly higher than the farmers who concentrate on growing their crops, and outsource all of those other functions to those who specialize in them.

No, the judge of which functions should or should not be outsourced is not fear-mongering politicians or hopelessly misguided economics columnists (e.g., Paul Krugman). The proper judge is thoroughly committed to the advancement of people’s well-being, but is known for having a harsh side: the judge is success or failure in the free marketplace. If outsourcing a particular function is a good idea, then the outsourcing company is better positioned to deliver its goods or services at a lower price, or to deliver a superior good or service at the same price, or some of each. Consumers benefit. If the contemplated outsourcing is a bad idea, then the company doing so will be in a poorer position to benefit its customers, and will tend to be eliminated from the particular market. Whether or not the outsourced ability is performed overseas, or by an organization others believe to be unpalatable is truly irrelevant, economically. It really is that simple.

That having been said, is there an exception to which functions can or ought to be outsourced? Yes, there is, and that function is, ironically, Project Management.

Think about it: all of the aspects of asset management can be (and commonly are) outsourced. The aforementioned payroll, plus accounts receivable and accounts payable, are easily performed by specialists in their own, separate organization. Even taxes have been outsourced for generations. In the strategic management realm, companies have outsourced this capability since 1841, when the first advertising agency opened its doors. Of course, advertising is not the complete strategic management function – which is to acquire more market share than your competitors – but it’s pretty darn close.

This leaves only Project Management from among the three management types that cannot be outsourced. Why not? Because Project Management is all about delivering specific scope on the customers’ parameters of schedule and cost, meaning that any attempt to transfer this capability would be to introduce an unnecessary third party in-between the company and its customers. Those companies that deal successfully with their customers do so due (in part) to the directness of the relationship. Now, project controls – the function of collecting data and processing it into usable information for PMs – can be outsourced, since this is an information delivery service, not the actual PM-making-decisions process.

So, in the realm of outsourcing, you can do anything but lay off of my blue suede shoes. Put another way, it’s rather interesting (is it not?) that you can outsource virtually any function that an outside organization may be able to do better or cheaper, but you can’t outsource Project Management.

And don’t even think about drinking my liquor from an old fruit jar.

I was recently in a discussion with my older son, who works as a prosecutor and graduated from Notre Dame Law School (Magna Cum Laude … not that I’m proud, or anything), when he brought up a highly relevant point that I had completely missed.

“Objection!” I interjected, “Evidence not entered as fact!”

“That’s gibberish” he replied.

“Not according to the television law shows I’ve seen.”

“Then you need to watch a higher caliber of law shows.”

Meanwhile, back in the PM world, there exists a surfeit of ideas that claim to be within the umbrella of legitimate project management science. Some of these (critical path, earned value) unquestionably belong, while others are at best suspect, at worst plunging whole sectors of project management theory into irrelevance. As my regular readers know, I regularly skewer these ideas, which nominally include:

• Generally Accepted Accounting Principles masquerading as project cost/schedule performance analysis,
• Risk Management or Analysis, in most of its forms,
• Communications Management, which does have its place, but has been marketed far beyond its nominal efficacy,
• Quality Management, which shares with risk management an over-dependence on Gaussian Curves,
• Human Resources management, which belongs with, well, resource management,
• Procurement, which belongs completely under the asset managers’ purview,

…among others. I can rail to my blogger heart’s content against these theories and practices, but there is no final arbiter, a judge who can render a decision that all parties must respect going forward.

Or is there?

Note what the Agile/Scrum developers did. In the crushingly competitive IT market, they adapted just those project management concepts that they believed provided a competitive advantage, and paid less (or no) attention to those practices that were, well, irrelevant. What traditional PM practices are missing from most Agile/Scrum projects? Just:

• GAAP practices pretending to provide cost/schedule performance,
• Risk Management or analysis, in most of its forms,
• Communications Management has been intensified, but only internal to the project team, and serves as a substitute for more formal change control or configuration management. All that business about engaging all stakeholders is notably absent.
• Quality Management is also put into its place, being very important to successful IT project delivery, but not to the point that the QC guys can slam the brakes on the whole project, or dramatically alter the way the software engineers do their jobs,
• Many (most?) IT projects get delivered without any HR influence, and
• Procurement. Including procurement might be a bit unfair, since IT projects are typically labor-intensive, but you see my point.

For those concepts and ideas whose developers/backers desperately want to be considered relevant, there will be no judge, jury, and competing assertions tested for veracity in a finding-of-legitimacy setting. They can only submit articles, columns, and blogs, entreating their readers to use some of those concepts in order to be doing “real” or “legitimate” PM. In the meantime, the ultimate decider of relevancy – the free market – will pick and choose which ideas are indeed helpful in their markets, and which are, well, irrelevant.

The jury’s out on those areas I’ve regularly criticized in this blog. But I’m optimistic, though this optimism may be fueled by the sub-standard legal dramas I’ve been watching.

Several science fiction works have addressed the concept of robots or androids being made so similar to actual humans that they are unaware that they aren’t real, living people. Movies (Blade Runner), and television series (Star Trek, Twilight Zone) have explored the concept, and I believe that the fascination with the premise has to do with the idea that we may not know ourselves very well, and, if we were to be confronted with our “real” selves, it might be a shocking experience. Meanwhile, back here in the real (PM) world, there are many software packages that claim to be able to provide any project manager with the information needed to keep a handle on their projects’ scope, cost, schedule, configuration, risk, communications, portfolio and strategic management…

Hmmm… that list is most of the whole PM enchilada. Some would have us believe that the capacity for computers to dramatically outperform humans in quickly and accurately collecting and processing data is a good thing. Add that to Hatfield’s Axiom #12, that the 80^th percentile best managers with access to only 20% of the information they need to obviate a given decision will be out-performed by the 20% worst managers who have access to 80% of the information so needed, and we have a truly chilling corollary: a given computer only has to be as good as the worst 1/5^th of managers to believably replace us in the workplace.

Personally, I don’t believe it will ever happen, and here’s why: Hatfield’s Axiom #11. It states that, for project management information systems to have any kind of value, they must be (1) accurate, (2) timely, and (3) relevant. Computers have accuracy down pretty well, and they process large amounts of data so much faster than their human counterparts that #2 is covered; but they will never be able to crack #3, determining which information bits are relevant, and therefore worthy of being used for decision-making, and which are not. Quantitative Analysts, known as “quants,” are professional data-scourers, who review large amounts of data in order to tease out some form of pattern, a repeatable observation of correlation that might point to a previously-unknown causation. But they don’t manage projects, they advise financial market brokers. Projects are a whole ‘nother animal. But that doesn’t stop Gaussian-curve lovers (risk analysts) and asset managers (accountants) from asserting that major project decisions can be reduced to an algorithm, based on the availability of knowable facts that produces winning decisions on a consistent basis. They have, in fact, been asserting some variant of this absurd idea since before robots (by definition, portable computers) knew how to run. Many industrial robots in Japan are so life-like that they’ve begun to assume jobs formerly handled exclusively by real people, such as receptionists. Since it’s a sure thing that, should an organization develop a true PM android, they wouldn’t be eager to advertise it excessively, for fear that real people would tend to avoid being led by a machine. The conclusion: there may be androids in the PM realm, indoctrinated in some truly irrelevant project management techniques, who are currently in charge of major projects. If this is the case, how would we know?

As fate would have it, I have a couple of tests that will determine if your PM is a real person, or an android. Test #1 involves whether or not your PM insists on keeping risk analysts on the project well past the creation of the original cost and schedule baselines. If the answer is yes, you should be concerned. I’ve previously defined “risk management” as institutionalized worrying, tripped out in statistical jargon. Here’s another definition: the results of risk analysis is a lazy thinker’s attempt to quantify relevance. Since relevance can’t be quantified, the best a robot PM could do is to have statisticians continue to guess the odds of bad things happening to their projects. So, if your PM is such a one, he may be a machine.

Test #2 is based on the extent that your PM has embraced the idea that the point of all management is to “maximize shareholder wealth.” My regular readers know that I’ve mocked this notion previously, but here’s a new angle of attack. Whether or not a given decision actually maximizes shareholder wealth is usually calculable, with the notorious “return on investment,” or ROI, being the go-to formula. But remarkably little of PM, past cost and schedule performance, can be readily quantified. A real person PM has a bit of decision-making-from-the-gut that inferior (strikethrough) mechanized managers can never duplicate.

Now, if you have come to the realization that you are, in fact, working for an android, don’t panic. All those science fiction movies about how human-like androids really want to kill us all off are probably extreme. Just the same, you might want to re-watch The Terminator series, just to re-familiarize yourself with how to overcome them…