Game Theory in Management

This story appears in a beloved book I read as a child, Reader’s Digest Treasury for Young Readers:

Yet even Dr. Bell (the person after whom Arthur Conan Doyle modeled Sherlock Holmes) sometimes made mistakes. Luckily, he also had a sense of humor when people asked him to give examples of his skill as a detective, he liked to tell this story: One day he and his pupils were examining a patient in a hospital bed. “Aren’t you a musician?” Dr. Bell asked him.

“Aye,” admitted the sick man.

“You see, gentlemen, it is quite simple. This man has a disease of the cheek muscles, from too much blowing on wind instruments. We need only ask him, and he will admit it. What musical instrument do you play, my man?”

The man got up on his elbows. “The big drum, doctor!”

I like to remember this story whenever I explore the epistemology of management information streams, since it is so easy to equate perceived project success with factors that may have been entirely incidental to the remembered project’s actual success. I once was working to set up the cost and schedule performance systems on a major project that was headed by a manager who insisted that my team provide a report he termed a “swim lane chart.” What he actually wanted was a PERT chart, sorted by performing organization. I could see the utility – the various performing teams were in a column to the left, and the activity boxes that appeared to their right represented the scope for which they were responsible.

“Okay” I offered, “we can do it, but we’ll need to start with a Work Breakdown Structure. Then, we’ll need the Organizational Breakdown Structure, so that we can cross-reference them into a Responsibility/Accountability Matrix, or RAM. Once we have that, we can load the information into the critical path software, and generate your report.”

“I don’t want to do any of that stuff. I just want a swim lane chart.”

“I know you want the chart, but we can’t get there without the RAM.”

So he had me removed from the project.

This guy just knew that his so-called swim lane chart was the key to managing the project to a successful outcome, and wasn’t going to listen to any of that project controls nonsense about how to get there.

I’m sure my readers have many similar stories. Some manager or exec has a particular project management artifact that serves as their security blanket, and will brook no challenge to its efficacy. Their attachments to these talismans can reach a zeal rarely seen outside of religious institutions or sports bars. And, when these superfluous information streams become institutionalized, the amount of managerial folly they generate can become extraordinary. Just look at the number of U.S. Government agencies who insist that projects have a risk management system.

It doesn’t stop with the risk management crowd, either – they’re just the more irksome of the bunch. Virtually every attempt at quantitative analysis in business or management requires the use of some subjective variable, variables that really cannot be known to within the boundary triggers for making decisions. For example, in comparing the Return on Investment (ROI) among competing prospective projects, the anticipated rate of return can only rarely be estimated to within double-digit accuracies. However, the decision on which projects to pursue are almost always made by single-digit margins. It’s as if the organization is making a decision to select a subcontractor based on the estimation of who has the fewer Capricorns on staff, and attacks anyone who doesn’t acknowledge that as an appropriate basis for making the decision as hopelessly ignorant.

Of course, there are ways of objectively determining the validity (or lack thereof) of the competing management information streams, but how to do so would take an entire book, which, fortunately, I happened to write.

I remember an English class I took when I was doing my undergraduate work, where the professor quipped that there is no record of a written language prior to around 3600 B.C., but at around that time there was a veritable explosion of languages recorded in various media.

“So, theoretically, the story of the Tower of Babel could be true” I suggested.

He chortled condescendingly as only professors of the humanities from state universities can, but essentially agreed with my point.

Some years later I was watching a documentary on ancient Babylon, where some archaeologists had unearthed a rudimentary battery, and dated it to around 3500 B.C.  The analyst being interviewed speculated that, had mankind pursued that technology back then, we would today be hopping galaxies in starships. Of course, he had no way of knowing that, but the speculation was fascinating. Could it have been another hint that the story of the Tower of Babel was not completely allegorical?

For those who are not familiar with Genesis 11, all of the people had moved to the valley of Shinar, and decided to launch a very big project, indeed. They were going to build a city for themselves that included at least one tower that would reach to the heavens. God came down, saw it, and said “If as one people speaking the same language they have begun to do this, then nothing they plan to do will be impossible for them. Come, let us go down and confuse their language so they will not understand each other.” (NIV). Well, that did the trick: they stopped building the city, and scattered themselves all over the Earth.

What we have here is an example of the Almighty derailing a project. Of course, we humans are pretty good at doing the same thing, but I think it’s instructive to observe how He did it. He confused their language. Why would this spell automatic doom for the Tower project? While the designers, engineers, and laborers no doubt spent a lot of time using their common language to make crude jokes and compare the statistics of ancient sports teams, it’s obvious that critical information could no longer be understood by the project team, making coordinated work or progress against desired scope impossible.

Don’t misunderstand – I do not agree with the so-called communications experts who insist that, if we could just perfect our methods of communication, the management world would be all glitter and unicorns. But I do believe that, should pathologies in the business model creep into the performing organization, the earliest place it will manifest will be the avenues of communication.

I define “office politics” as those instances where members of the organization behave in a manner that is inconsistent with (or even contrary to) the stated goals of that organization, but benefit them personally. Since that definition would also include theft, let me hone it down a little further: they do so by manipulating the information streams on which the organization depends.

Take the favorite tactic of the Maccoby archetypical Jungle Fighter: they magnify their rivals’ (virtually everybody around them) failings, while minimizing their own. They amplify their accomplishments, while trivializing their rivals’. If the organization does not depend on the information shared informally among team members, this is fairly insignificant . However, if this information stream is acted upon by management, then the poorest decision-makers advance at the expense of the truly talented members of the project team.

Consider also those who tamper directly with the management information stream. Our friends, the accountants, and risk managers leap directly to mind. By vastly overstating their techniques’ efficacy, and advancing them, they elbow aside more legitimate sources of valid management information, significantly increasing the chances that poor management decisions will be made.

Relevant, timely, and accurate information is the life-blood of any organization, and its reliable conveyance is often easily manipulated by the project-killers in your ranks. And such ones will rarely self-identify by wearing hockey goalies’ masks.

As I’m sure most (if not all) of my regular readers know, many psychologists believe that, when we dream, we are experiencing our subconscious selves attempting to communicate with our conscious selves, but in a visual language which is difficult to understand in our waking existence. The images can be primordial, powerful in connotation, but irrational in delivery. And, when we experience recurring dreams – or, painfully enough, recurring nightmares – there’s a natural tendency to try to unravel the message and, potentially, act on it.

So, what are the most common project management nightmares? Well, clearly, there’s the over-budget, late-schedule, but not knowing about it until it’s too late one. Then there’s the customer-pushing-extra-scope variety, leading to the most common project performance killer, the dreaded scope creep. There are many others, and yet I must ask: what are these nightmares trying to tell us?

Consider our information streams. They are the real-world data feeds, what we see on the surface, and the information that leads us to act, to make decisions. As managers, the information we seek is analogous to what we see, hear, smell, and touch in our waking existences. When we seek irrelevant information, and get it, and then make decisions based on it, then the nightmares start to become realities. Why? Because there’s a project management reality out there, that’s not being perceived, and threatens to become an inescapable outcome … and management is oblivious to the imminent disaster looming. The information streams an organization chooses is a clear indicator of its true character, as well as its chances for success.

Take the organization that eschews earned value analysis. This is most often the case where our friends, the accountants, have successfully pushed the fraudulent idea that the general ledger can provide all relevant cost information to decision makers. But there are many instances where the organization elects to nix any EV system because they simply do not want to know how their projects are performing in cost performance space. Why would any organization not  want to know this? Because it’s directly tied to responsibility. Organizations that seek to obscure who or what was responsible for successes or failures, and, instead, substitute narratives bereft of facts, cannot abide even the most basic of earned value management systems. Organizations seeking to deflect responsibility will resist information streams that accurately reflect cost and schedule performance, and will repeatedly encounter the way-over-cost, way-past-schedule nightmare.

Risk management is another conceit, that the real world of project performance will attempt to penetrate with frightening feedback. Bad things happen on projects – it’s a fact of management life. How the project team responds is the difference between success and failure. So, what if a risk analyst predicted that a given bad thing might hit the project? Did it alter the team’s response at all? If yes, then, yeah, it was a good idea to spend that money on the risk analyst. However, in my experience, the risk people rarely – if ever – proscribe a course of action that wasn’t already on the project team’s roster. In other words, they accurately worried for you, Mr. /Ms. PM. And it bought you exactly nothing. The anti-PM nightmare talisman that you bought from the risk management types is literally worthless.  Sorry.

And what about those who attempt to “manage” their schedule via milestone lists? These invariably list project objectives (again, don’t tell me, right? In a spreadsheet?), and pull “status” by asking their principals if they anticipate that they will accomplish their milestones on-time, a little late, or very late/miss altogether. The answer is, invariably, that they will certainly meet their schedule dates … until they don’t. This is also reflective of an organization that has no intention of respecting responsibility for project performance and is, instead, situating itself for convenient scapegoating.

Are these a few of your PM recurring nightmares? The interpretation lies within identifying the business model pathologies that have wormed their way into your organization. And that exercise, dear readers, is never without pain.

Probably most people who took a challenging course in college have the recurring dream of becoming suddenly aware of having signed up for a course that should have been attended, but wasn’t, and finals week is upon them. I’ve had this dream fairly often – at least four times per year since finishing my undergraduate work – so you would think that whatever message my subconscious was trying to tell me would have been transmitted by now. But, no, I guess my id is just having too much fun giving my superego fits.

But project disasters also recur, and, often, for very basic reasons. One of the most basic has to do with the development of the Work Breakdown Structure, or WBS. Anybody who knows the first thing about project management has to know what a WBS is, which makes it so abused. The novice executive simply can’t admit to not knowing how to develop one, so they will take a stab at it, even if they have no idea how it ought to be developed. They’ll put in organizational breakdown structure elements, or functional breakdown structure elements, and double-dog dare the project team to tell them nay.

What’s the consequence? Since OBS or FBS elements in the WBS cannot be objectively measured as to percent complete, they negate any attempts at capturing cost or schedule performance through either earned value or critical path methodologies. The core of these methodologies relies on the estimate of percent complete as of the end of a given reporting period. Quick! What’s your percent complete on the design team? On the reporting team? Can’t offer it up? It’s because there are non-scope-based elements in your WBS, so you are pretty much helpless when it comes to assessing cost and schedule performance. And, when you are flying blind in cost/schedule performance space, that’s the stuff of managerial nightmares.

Consider also the almost aristocratic trend away from “doing” the aforementioned earned value or critical path analysis. Oh, sure, our friends, the accountants, have convinced you that they can provide cost performance information, based on analysis derived from the general ledger. Such assertions are complete equine droppings, but, for whatever reason, seem to gain credibility in boardrooms. With the supine whisperings of the asset managers lulling project leaders into a false sense of security, projects leap into nightmare space, with chilling regularity. 

And what of the milestone watchers? They eschew actual activities, with schedule logic, in favor of putting project objectives – don’t tell me, let me guess: they’re in a spreadsheet, right? – into some report, where the milestones are assessed with one of three status colors:

·         Green, we’re gonna hit this on-time.

·         Yellow, not so confident.

·         Red, yeah, we’re gonna blow it.

Problem here is, everything is green, until it isn’t. Absolutely no ability to forewarn execs in time for them to correct what would have been easily fixable, had they just been notified in a timely manner.

Can risk managers fulfill the role of Sigmund Freud, and alleviate your project nightmares? Um, no, and for one simple reason: the future cannot be quantified. The RM-types can push all the statistical – what was my term before? Oh, yeah, equine droppings – they want, but they can’t eliminate the recurring PM nightmares.

What can provide relief to such nightmares? It’s simple – return to the basics.

Almost everyone on Earth knows the fate of the passenger liner Titanic: she struck an iceberg on her maiden voyage and sank with heavy loss of life in the icy northern Atlantic on April 15, 1912. I think part of the continuing fascination with this otherwise open-and-shut case lies in the forensic analysis of the contributing causal factors. Whenever there’s a disaster, it seems, there’s always dozens, hundreds, thousands of narratives that attempt to connect the dots of the factors that led to it.

Take the aforementioned Titanic. The proximate cause of her sinking was the hole torn open in her hull from colliding with the iceberg. The material causes, however, are myriad (for those less familiar with causal analysis, in order for something to qualify as a material cause , it must pass the if-not-for test, as in “if not for Titanic’s rudder being too small, it would have missed the iceberg, and, therefore, not have sunk.”). For example, let’s start asking “why” a few times to see if we can’t isolate a material cause or two. Why did the ship hit the ‘berg? Because her crew recognized the danger too late to turn the ship away. Why did they recognize the danger so late? One possible reason is that the lookouts were not equipped with binoculars. Why not? Because the ship’s purser had been relieved of his duties back in Liverpool, prior to the ship sailing, and took his locker keys with him. The binoculars were in the lockers, but nobody had the keys, so that’s where they remained (and may very well be there to this day). Also contributing to the ship’s inability to navigate around the iceberg was the fact that the sea was extremely calm, and an iceberg is far easier to see at night if there are waves breaking against its base. Additionally, the Titanic was – informally, at least – attempting to break a speed record in her Atlantic crossing, and ships turn more slowly the faster they are going.

Given these facts, can the following be reasonably asserted?

·         If not for the purser being fired prior to the voyage, the lookouts would have had access to binoculars, warned the bridge in time to turn the ship, and the Titanic would not have sunk.

·         If not for the new purser being reluctant to break open the lockers holding the binoculars, the Titanic would not have sunk.

·         If not for the lookouts watching Leonardo DiCaprio and Kate Winslet kissing on the foredeck, the Titanic would not have sunk (I just made that up, but there seemed to be at least a hint of this in the movie).

·         If the sea had not been so calm, the Titanic would not have sunk.

·         If the ship could turn sharper at higher speeds, the Titanic would not have sunk.

·         If the ship was not involved in an attempt at record-setting, the Titanic would not have sunk.

This list could go on and on (e.g., let’s say the purser was fired for a trivial reason. If not for the purser’s boss having cold coffee at breakfast, he would have been in a better mood, not fired the purser for trivial reasons, meaning the lookouts would have had access to binoculars, etc., etc. The assertion then becomes, if not for a cook giving the purser’s boss cold coffee, the Titanic would not have sunk.).

When we conduct a forensic analysis of what happened in a project nightmare (finally! Michael’s turning to the October theme!) there’s a distinct danger of swerving away from a legitimate causal analysis, and towards competing narratives. Whenever you hear the expression “historians disagree” about anything, then you are seeing just such an instance of competing narratives, and I have to ask, “why?” Aren’t the facts known? Haven’t the assertions that don’t meet the criterion for being evaluated as facts been tossed out? What are those guys doing hanging around the grassy knoll with their silly-assed pamphlets, then?

Don’t think the exact same thing doesn’t happen whenever a project goes south. People with only the barest grasp of causality analysis, the rules of evidence, or the rules of logic will leap to the forefront with their particular narratives, narratives that serve their own particular interests. The whole “lessons learned” effort then becomes a bitter competition among narratives, dramatically reducing the odds that the true causal factors of any given project disaster become known, much less avoided in the future.

Which is why Jack Dawson and Rose DeWitt Bukater, single-handedly, sank the Titanic.