Game Theory in Management

It was yet another dark and stormy night. My secretary had just left, complaining yet again over not having been paid exactly on-time, and slamming the frosted glass door with my name on it, ylnatS yrrebpsaR, eyE etavirP behind her, when the phone rang.

“Is this Raspberry?” the older man’s voice at the other end inquired.

“Is that who you called?”

“Yeah, but I was expecting your secretary to answer.”

“She’s upset about something, and left early. Is this Roger Doright?”

Roger had been one of my earliest clients. He was a regular, straight-up guy, something of a mentor to me. As he approached retirement age, he had turned over more and more of his corporation’s decision-making to his kids.

“Yeah, it’s me. How have you been Stanly? I haven’t spoken to you in years.”

“I’m good Roger. How are things with you?”

“Well, I’m having a bit of a problem, which is why I called you. I have a hunch that a couple of the major projects in my company’s portfolio are about to go off the tracks, but their PMs insist that all is well, and offer up reports reflecting that. I can’t argue with the reports, so I need somebody to dig a little bit deeper. Are you available?”

“Sure. I can be out there tomorrow morning.”

“One more thing…” Roger continued, “my youngest son, Irvin, needs to learn the project management ropes, so I would like for him to accompany you.”

“What’s his background?”

“Oh, he’s been through all the relevant classes and coursework to do what you do, the whole investigative PM thing, with his specific role being ‘Clue Analyst.’ But he lacks real experience. He’ll meet you at my corporate headquarters tomorrow morning.”

As I put down the phone, I remember thinking “Great, just great. A newly-minted ‘Clue Analyst.’ But Roger has done so much for me, that I suppose I can grit my way through this.”

*   *   *

The dark and stormy night had given way to a bright, clear morning. An overly enthusiastic young man approached me in the foyer of the office building.

“Stanly Raspberry? I’m Irvin Doright. Let’s get  you badged in – the portfolio reviews are about to start, down in the Hatfield Conference Room.”

“The what conference room?”

“It’s in the basement. It was named after an obscure ProjectManagement.com blogger my father liked to read. Let’s get going.”

As we walked into the conference room, several of the PMs glanced up, recognized me, and stifled grimaces.

“What’s he doing here?” I heard one whisper to another.

I looked over the review documents in the pack that was passed out to the participants as I took a seat in the back. The first PM got up to address the projected slides behind him.

“As you can see, my project is on-time and on-budget. At this rate of performance…”

“Excuse me,” I interrupted, “but I can’t seem to find the Cost Performance Reports.”

“The CPRs are in the back.”

“Oh, okay, but why aren’t they in the front? Why are we talking about … what is that chart, anyway?”

“It’s the bottoms-up Estimate at Completion, showing that we’ll finish on-budget.”

“Well, I did a quick calculated EAC, and your own Earned Value data indicates a $12 million dollar overrun.”

“Mr. Raspberry, I’m not seeing that in the risk register!” Irvin hissed at me.

“Nor is that consistent with our information” the PM added.

“That’s because conventional risk analysis techniques and bottoms-up EACs are irrelevant to project performance and accurate at-completion figures” I replied.

Most of the people in the room gasped, almost as if they were paid actors awaiting some common-sense challenge to their embraced business model.

“Irvin Doright here is the owner’s son” began the PM, “and heads up our risk management division. Irvin, do you agree with Mr. Raspberry’s assertion?”

“Absolutely not – risk management deals with virtually all aspects of project management, and my Monte Carlo analyses have been spot on.”

“Not this time” I answered.

“Bruno, confiscate Mr. Raspberry’s badge, and escort him off the premises” Irvin ordered.

Once I was “escorted” off the site, I placed a call to Roger.

“I think I see your problem…”

We all have our own way of looking at the world, and this worldview is shaped by our experiences. Indeed, B.F. Skinner launched an entire school of psychological thought, known as “Behaviorism,” that essentially said that we are all products of our experiences, to the point of minimizing – or even eliminating – the concept of the autonomous, free-thinking person. These experiences he broke down into a stimulus-response-consequence sequence, a model that came to dominate Behaviorism for its time in the psychological world’s limelight.

It was hard to argue with them. When I did my undergraduate work, the university I attended had a psychology department dominated by Behaviorists.  My three semesters (required) on the subject were taught by them, and they had some pretty impressive stories. One of the ones that stand out for me was the one where Behaviorists were invited to manage a mental hospital, taking over from a group of Freudians. Instead of having the patients lie on couches and describe their dreams, the Behaviorists instituted a system where the inmates would be awarded with a white marble if they got up on-time, made their beds, and performed other basic functions without mishap. Should they help others, or go above and beyond their normal chores, they would receive a black marble. These marbles could be exchanged for extra desserts, or more television time, or extra time in the recreation area, or other desirable things. According to my professors, the Behaviorists would have virtually the entire institution’s population acting fairly normally in a brief period of time, previously unmatched by the adherents of any other school of thought. Skinner actually wrote a novel, “Walden Two,” in which a reporter is assigned to report on a commune that had been founded and maintained on the ideas of Behaviorism. And – wouldn’t you just know it? – the commune’s inhabitants were all so much more happy, prosperous, well-adjusted, and sophisticated than the rubes outside that the reporter ends up staying on in order to become a part of the society.

Behaviorism has fallen into disrepute since my undergrad days. One sure indicator had to be the redefinition of key terms by the pop culture (“negative reinforcement” came to be commonly associated with punishment, when it was actually the exact opposite). Another hint was in Walden Two itself: if you can only advance your theories in a fictional setting, you probably have precious little hard data supporting your position. Basically, the notion that we are little (or no) more than the sum of our experiences came to be viewed as sub-optimal (to hijack yet another psychological term).

So, as we shift gears to November’s theme of business analysis, I caught myself wondering if something similar to what happened to the Behaviorists might occur within the business analysis realm. Does anyone remember the quality craze of the 1980’s? Or, even closer to home within the PM world, the emphasis on “life-cycle cost” management?Or “critical chain scheduling”? In each of these cases an idea that had been around for years and years was assigned a new lexicon, trotted out as a new idea that had never (or only incidentally) been considered previously, and then laden down with story after story about how organizations went from struggling to succeeding wildly once they adopted the new, trendy theory set.  Eventually these organizations would run into difficulties that were outside the trendy theories’ purview, and management science’s buzz would return to its nominal dominance by the general ledger custodians’ ideas. In the case of critical chain management, its supporters went so far as to write a novel, advancing the well-known concept of crashing the schedule, as some kind of special insight.

And yet, to real business analysts, it’s all fake. Hooey. Alchemy. Smoke and mirrors. It’s the marketing of a narrative as a substitute for legitimate management science.  Behaviorism worked as long as its researchers could engineer the experiences of its subjects – hence their success at getting lab rats to run mazes faster via punishments, rewards, or a combination of both. But business analysts know that, with very few exceptions, they cannot engineer (or even quantify) the external parameters that lead to management success (at least not legally). The best they can do is to help prepare the organization to present the most robust response to whatever happens to (in our case) the project team as they pursue their target scope.

Unfortunately, there remains a plethora of “management insights” masquerading as legitimate business analysis, making claims to deliver success to their adherents, and there’s really no way of making them go away. Hey! Maybe we can borrow ideas from the Behaviorists, attach electrodes to some of these management narrative promoters, and…

Hatfield’s Rule of Management Number Ten is that in order for a piece of management information to be of any value, it must be timely, accurate, and relevant. If it is not all of these three things, it’s not only useless, it’s hurting your organization. But where quantifiable criterion for information’s timeliness or accuracy can be developed and used, solid bases for evaluating relevancy is a trickier matter, and yet a business analysis issue that the PMO must get right, or risk the organization’s own relevancy.

Does Relevancy Even Matter In Pick-Up Lines?

Take, for example, one of my favorite targets for the irrelevant label, the comparing of budgets to actual costs. This has value only in the asset management arena, specifically:

• Depreciation, and
• When performing an analysis of how accurate your estimates are

… and that’s it. However, since such a comparison is a critical part of managing a household budget, a natural assumption is that it’s important to evaluating project performance. It’s not, and a simple thought exercise can demonstrate this clearly.

Suppose a $100,000 (USD) construction project. To keep it simple, let’s assume that the original estimate included $75,000 in labor costs, and $25,000 in heavy equipment. At the end of the project, the final costs were $20,000 in labor costs, and $70,000 in heavy equipment. According to the “analysis” of comparing budgets to actual costs by category, this represents a monumental failure, even though the project came in under budget by 10%! Further, for each reporting period that this analysis was performed the resulting information was inaccurate inasmuch as it portrayed a performance issue with the work.

It Actually Gets Worse

This technique can be critically wrong in the opposite direction, as well. Let’s say our $100K project was actually spending at the 75%/25% split of labor-to-equipment at the 2/3rds-spent point of the project (i.e., $49,500 in labor, $16,500 in equipment), but the project is only half-complete. The comparison of the budget to actual costs by category (or even at the line-item level) indicates no problem, when even a simple earned value analysis would readily indicate that the project is on a path to a $32,000 overrun.  A broken clock, as they say, is right twice a day, but comparing budgets to actual costs, even at a detailed level, is right only by coincidence. It’s far more likely to be profoundly wrong, indicating problems where none exist, and portraying smooth sailing when the project is in deep trouble. If there were a toolbox of the exact opposite of valid Business Analysis techniques, then comparing the budgets to actual costs (no matter the level) would have to be near the very top of the set. Yet, I can almost guarantee that the majority of PMOs will engage in this form of analysis, and treat the illusory variances as if they were legitimate cause for concern.

The problem is actually worse than even this. Management Information Systems require time, effort, and money to design, set up, and maintain. Another quick thought exercise: why would any manager pay for information that’s already readily available? Obviously they will only spend money and staff time chasing information streams (a) that are not already available, and (b) that they are convinced are relevant. It follows, then, that should those resources pursue irrelevant information, they do so at the expense of the relevant info streams – a prime example of opportunity costs. For example, your business analysts can be assigned to either set up a basic earned value system for the projects in the portfolio, or else you can ask them to perform whatever skullduggery they need to do to uncover the zodiac signs of the competitor’s project managers. A comically extreme example, to be sure, but irrelevant is irrelevant, and having your business analysts pore over projects’ initial baseline estimates, line item by line item, and compare them to the actual project costs in the general ledger, again, line item by line item, is just as futile (if not more so, since it’s not inevitable that knowing the competition’s zodiac signs will lead to a singularly erroneous conclusion the same way that the budgets-to-actuals comparison will).

You Didn’t End Up Marrying That Person, Did You?

The budgets versus actual costs at the line-item level comparison is not only common, but I know of at least one PM guidance-generating organization that mandates it. If your organization is committed to this silliness, it’s committed, and that’s unfortunate, since there’s no telling what other irrelevancies they have embraced. But, if that’s the game being played, perhaps you can convince them to stop performing that particular analysis by asserting that a Feng Shui expert has established it’s damaging to the organization’s invisible forces.

We humans analyze a lot of things – our preferred modes of transportation, clothing, and the various technical approaches we take to bring our projects in on-time, on-budget. Of course, the way we analyze these things can (and does) have potential difficulties; hence the thankfully short-lived popularity of the AMC Pacer, bell-bottom jeans, and a plethora of business analysis techniques of questionable utility that have somehow crept into the PM codex.

Duct Tape Won’t Fix That

I appreciate the need to perform business analysis well. With the majority of companies in a wide variety of industries operating with as little as a 3% profit margin, any insight that can be gleaned from the available data might be the difference between survival and bankruptcy for quite a lot of managers in general, and Project Managers in particular. The problem is getting at those insights, at knowing what kinds of data to collect, how to process it into usable information, and which decisions are indicated from accurate, timely, and (most of all) relevant information. These are all very fluid parameters, varying from industry to industry, and project to project. The inventory control and reordering advances that made Walmart a giant in retail would have little to no effect on your typical software developing company, much as Carnegie Melon University’s Capability Maturity Model would not be expected to provide insights on getting ahead in retail.  The specific tools that help in one type of project aren’t necessarily useful in another, while, at the other end of the extreme, notions that are so general as to be accurate across industries (e.g., “work hard, and treat customers well”) are rendered cliché. What’s a business analyst to do?

Let’s start by defining which type of management we’re talking about here. As my regular readers know, there are three types of management, Asset, Project, and Strategic, with different objectives and tools used to attain them. Since this is a Project Management blog, I’ll jump right in to PM, with the caveat that it’s absolutely necessary to discard the tools belonging to the other types. As a PM, calculating an asset’s return on investment (ROI) does nothing for me but waste my time (and don’t get me started on those who attempt to compute ROI on things like a project team – the epistemological equivalent of having a 16^th century apothecary fill your prescription for Amoxicillin).

Neither Will Crystal Balls or Tea Leaves

Within the realm of PM, the next crucial question is: what is it, generally speaking, that you want to know? Everyone wants to know the future. Business Analysts get that. But the future cannot be known, or quantified, no matter how fancy the Gaussian Curve-overusers’ (also known as risk managers) formulae appear to be. All management information systems that attempt to capture the future are known as “feed-forward,” and depend on highly subjective data. And just so we’re clear – by “highly subjective data,” we’re talking about somebody’s prejudices, or guesses. Feed-forward systems are notoriously unreliable, again, no matter how much statistical jargon is used to convey their “results.”

For the Project Manager, then, the best available information is based on feed-back systems; and, in the PM world, this information comes from the analysis methods of Earned Value and Critical Path. Each system is based on known facts – objective data – but can deliver a remarkably accurate picture of the future. This is because these methods return the project team’s performance, and in rather stark terms. Using estimate-at-completion formulae (Earned Value) and recalculating the project’s status file (Critical Path), these systems can return the total project cost and completion date, almost always within 10%, and often even more accurately.

Dogs or Cats, Though…

Indeed, Earned Value and Critical Path methods are the dogs of the business analysis world. They’re reliable, relatively inexpensive to obtain and maintain, and provide many benefits, all while being very happy to just be a part of the project team. Risk management, and other feed-forward, subjective data-based systems, on the other hand, are the cats of the business analysis universe. They’re the very picture of unreliability, and yet through a bizarre appeal to the intellect, have been accepted at a level equal to (or even higher than) dogs. They are high-maintenance, and carry an air of being superior to the other elements of the project team, even though they are demonstrably unable to make any real contributions to the end goals of the organization.

So, the overarching business analysis question is much like the preferred pet quandary, with a similar optimal solution. Go with the dog, and ignore (or even get rid of) the cat –it ignores you, after all.

As I’ve laid out in this month’s previous blogs, Strategic Management is the art of acquiring more market share, whereas Project Management is the science of executing project work so as to bring the effort in on-time, on-budget, and meeting all customer expectations concerning scope. And Asset Management is, well, just that (how previous business scholars have sold the management science world that it’s all about them is beyond me). Some so-called experts will assert that Strategic Management is just one more rung up the scalability ladder of PM, as in Project-to-Program-to-Portfolio-to-Strategic, but this is false, to an extent. Let me explain.

A strategic plan is an overarching structure, used to organize the more specific tactics. For example, if your strategy is to gain more market share by attracting more clients while keeping the ones you have satisfied, then the following tactics are called for:

• Create a product/project management office, or PMO (if you don’t have one already)
• Attract talented workers
• …while getting rid of the pretenders (need help identifying the real PM-types from these pretenders? I’ll take that up in subsequent blogs).
• Place a priority on the training budgets of the valid PM types.
• Develop a feedback communication stream from your customers (Customers, not “stakeholders,” no matter what your communications experts say).
• Ignore the accountants and risk managers. They do not help with this strategy.

Conversely, if you are given to taking recreational drugs while executing the duties of a Chief Executive Officer of an unfortunate organization, and have selected a strategy that calls for wasting resources while transferring the costs to your soon-to-be-ex customers under the guise of legitimate project management, then the following tactics are in order:

• Bring in high-level risk managers and analysts, and have them crawl over every aspect of the cost and schedule baselines. Insist on thorough documentation of their analyses, to include a risk management plan, risk register, Monte Carlo analysis, and the complete set of assumptions that gets you to an 80% confidence interval on the baselines.
• Hire communications experts, and have them “engage stakeholders.”
• Don’t forget the quality analysts, and their infernal fishbone diagrams with a “branch” for every activity in the WBS that could have been planned differently.
• Ask the risk managers to perform an analysis on having the quality people do the Hasegawa Diagram on each of the lowest-level items in the WBS index.
• Have the quality experts perform an analysis on the methods employed by the risk managers.
• Assign each of the previous two bullets their own Work Packages, and schedule their activities. Have your project controllers/schedulers ping them at the end of each month for their percent complete estimates.
• Arrange to have the risk managers and quality control specialists present their pre-kickoff findings at the same meeting. Sell tickets to help offset project costs.
• Arrange to have the communications specialists attend as well. Prep them by reminding them that all stakeholders’ opinions have value, and must not be shouted down or categorically rejected,
• …particularly when the other members of the Project Team tell the communications specialists that their ideas have no value, and should be categorically rejected.

The beauty of this approach is that, in those instances where your customer is convinced of the efficacy of risk, quality, and communication management practices, the fact that these members of the Project Team will probably be shouting at each other will only serve to support the notion that the appropriate level of (wasted) energy is being devoted to each of these approaches, and that the project in question is being “properly” managed. But as a way of increasing market share – i.e., Strategic Initiative Management – it’s not going to work.

No matter how much twisted fun it may be to see unfold.