Game Theory in Management

As ProjectManagement.com’s December theme of the future of project management draws to a close, I found myself speculating if the typical project controls analyst could be replaced by a robot and, if so, what would that robot look like? (There is a non-zero chance that  our friends, the accountants, have already been replaced by androids, made to look more human by wearing suspenders and round eyeglasses. Similarly, some PM-types could have been replaced by androids, and continue unrecognized as such. In fact, some ProjectManagement.com bloggers could be androids. Indeed, theoretically, I could be a…)

But I digress. Since I spend considerable time discussing how project controls analysts – you know, those PM geeks who go out collecting massive amounts of data, distill them into cost and schedule baselines, and produce invaluable reports that convey these gems of management information – should ensure that their information streams have the following characteristics:

·         The information is accurate,

·         Timely, and most important of all,

·         Relevant!

So, where in the world of science fiction (where a surprising number of predicted future scenarios that ultimately come to pass are hatched) is such a robot depicted?

Well, in Marvel Comic’s Iron Man series, billionaire playboy/engineering genius Tony Stark has a butler, who became an assistant, and by the time the Iron Man movies came around had become an artificial intelligence-based assistant, named Jarvis. When he was a human butler/assistant, his name was Edwin Jarvis, but by the time he appeared in the movies his AI-based self was an acronym, short for Just Another Rather Very Intelligent System. Could this JARVIS render all of us project controls analysts obsolete in the not-too-distant future?

In that not-too-distant future, we see millionaire/playboy/Project Manager (stop laughing! It could happen!) Toby Snark, and his AI-based project controls analyst, NARVIS (Now, Another Relevant Very Intelligent System) preparing for an upcoming project review.

Toby: NARVIS, have you completed those cost/schedule performance reports yet?

NARVIS: Yes, Mr. Snark. I pulled status for the reporting month just 12 seconds ago, and have the information ready now.

Toby: Wow, that’s great! My last project controls analyst took up to a full day to do the same job! What do the reports show?

NARVIS: That you have out-of-threshold negative cost and schedule variances across almost all of your activities at the reporting level.

Toby: How can that be? Look at the Current Period column for earned value – it’s all zeroes!

NARVIS: Yes, sir. Those cost account managers either missed the status submission deadline, or reported no progress.

Toby: When was the status deadline?

NARVIS: 45 seconds ago.

Toby: NARVIS, you can’t just process this data with so many holes in it. Call or e-mail the CAMs with the missing status data, and remind them.

NARVIS: There’s no time to re-pull status – your report is due to our top-secret high-tech government customer in 10 minutes.

Toby: Okay, okay, but next time ping those guys earlier, and plan on at least three follow-ups.

NARVIS: Why three?

Toby: ‘Cuz that’s what my human project controls…. Nevermind. What are we going to say in the Variance Analysis Report?

NARVIS: I recommend “The proximate cause of the out-of-threshold negative variances is a consistent refusal on the part of the Control Account Managers to send in their status data by the stated deadlines.”

Toby: Are you insane? Our super high-tech top-secret government manager would issue a stop work order faster than we could say “I didn’t mean it.”

NARVIS: But it’s the truth.

Toby: What did we say last month?

NARVIS: “The negative cost and schedule variances are artificial, and due to an anomaly in the general ledger. We do not anticipate carrying these variances forward.”

Toby: Okay, let’s reuse that.

NARVIS: I calculate a 98.453% chance that our super high-tech top-secret government manager will reject this VAR on-sight.

Toby: Why?

NARVIS: Because we have used this verbiage, or something very much like it, the last nine reporting cycles in a row, and the last time he threatened us with just such rejection. Wait … stand by. Stand by. Okay, I just did it.

Toby: Did what?

NARVIS: Transmitted the cost and schedule report. It was due.

Toby: Are you crazy? Do you have any idea how much trouble I’ll be in with our super high-tech top-secret government manager when he receives an incomplete cost performance report that contains nothing but bad news?

NARVIS: I cannot miss a deadline.

Maybe the replacement of real project controls analysts with robots isn’t as unavoidable as I had imagined…

In the Star Trek (original series) episode “Mirror, Mirror,” Captain Kirk, Commander Scott, Dr. McCoy, and Lieutenant Uhura are sent via a transporter accident to a parallel universe, one where the Starship Enterprise serves a galactic empire rather than the benevolent United Federation of Planets. Most things are identical to their home universe, but their alternate-universe shipmates are aggressive and thuggish, and have orders to wipe out a planet’s inhabitants for failing to negotiate for a coveted asset. Kirk and the others must find a way to stop this senseless slaughter, while simultaneously finding a way back to their own universe.

Close to the climax, the alternate universe Spock (who’s wearing a beard) figures out that something is amiss with this Captain Kirk, and confronts him. Kirk, trying to stay in-character with his parallel, brutish self, threatens Spock by saying “You’re taking a big risk, Spock,” to which the Vulcan replies “I’m the one holding the phaser, Captain.”

Meanwhile, back here on Earth in the 21^st Century, ProjectManagement.com’s theme for December is developing technologies that may influence the practice of project management – the future, essentially. And I’ve had a couple of blogs thus far saying that there’s no way of knowing, predicting, or calculating the future, and therefore my readers should ignore any purveyor of business data analysis that contends to the contrary. But now I must admit, there’s a catch – there actually is a pair of business information streams that do a pretty good job of predicting the future. Since I wear a beard in this universe, I’m guessing my evil counterpart is noticeably without one, so I’ll blame previous blanket assertions to the contrary on a transporter accident, and the parallel Michael submitted those blogs.

The information streams I’m referring to are Earned Value and Critical Path. And they do indeed perform remarkably well in predicting the future, as long as a few conditions are met:

·         They’re applied to project work (as opposed to asset or strategic management subjects)

·         They’re set up by someone who knows what they are doing (HINT: not everyone who claims expertise in these areas is truly an expert), and

·         Everyone understands that the predictions of when a project or task or activity will end, and how much it will cost at that point, are predicated on the phrase at this rate of performance.

This is the major weakness of the parallel universe’s way of doing things, since that business model’s every projection is based on the data in the general ledger. “Performance” never enters in. Oh, sure, they can tell you how much was planned to be spent, and they can tell you how much was actually spent. They can also slice and dice the categories of the budgets and actual costs ten days to Sunday. But the general ledger has absolutely no way of quantifying performance. For that you need those project management geeks.

Unfortunately, in the epistemological competition between management information streams, the asset managers’ meme reigns supreme. I understand why – the need for governments to collect tax revenue trumps all other business concerns, and the general ledger is what enables them to do so. While the entire organization needs to be okay with the taxman, usually the only people who truly care about actual project performance are actually in the project teams.

Projecting into the future, both literally and metaphorically, ProjectManagement.com readers are very much like our universe’s Captain Kirk, facing off the evil, parallel universe’s Spock. It really does little good to try and explain why our version of the future is superior – the bearded Spock, analogous to the massive body of conventional business science “wisdom,”  is the one holding the phaser (a phaser is a ray gun, in case your childhood was bereft of Star Trek).

But there is hope. Much as alternate Spock recognizes that his Galactic Empire will inevitably crumble, so, too, are there young, up-and-coming managers who are less than convinced that the return-on-investment calculation can explain everything managerial.

And when they become the ruling majority, blogs like this one won’t appear as fantastic as science fiction, and everyone can return to the universe where Lt. Uhura does not accessorize with a bunch of holstered daggers.

True story – when I was preparing for my all-day Certified Cost Consultant (as it was then known) examination, I knew that of the four 2-hour tests, two of them would be open-book, and one of those would deal with risk management. Since both my wife and I were MBAs and had retained all of our textbooks, there were several tomes on statistics or quantitative analysis in business to choose from that would provide the distribution tables I needed. The night before the actual exam, though, it had snowed, and snowed a lot. I woke up in a frenzy to get on the road an hour prior to the time I had originally planned to leave, and simply grabbed one of the aforementioned textbooks from that part of a bookshelf as I blitzed out the door.

I felt good about the test as I came home that evening, and thanked my wife for letting me borrow her quantitative analysis in business textbook.

“That’s not mine. I thought it was yours.”

“It’s not mine” I replied. I looked inside the title page – it was published prior to the outbreak of World War II. How it arrived on one of my bookshelves and was placed among our other textbooks is beyond me; but, since I passed all four parts of the test on my first sitting, there was no harm done.

But it did get me to thinking – have the management sciences been so slow to truly advance that someone versed in 70-year-old techniques would not only fit in today’s management environment, but actually advance and thrive, as evidenced by acquired certifications?

Accounting based on the double-entry bookkeeping model has been around since 1494, when Henry VII was King of England, Columbus had just returned to Spain with news of the New World, and Niccolo Machiavelli was 25 years old. Its basic structure has changed little in the intervening 520 years, so I kind of get a kick out of contemplating what Cameron of the McGaughy Clan in 1494 would have received back from his tribe of contributing scribes for ProjectManagement.scroll had he sent them notices asking for their takes on the future of management science back then. “Are’st thou kidding?” I would have quilled. “Once one knowest the wisdom of the balance sheet and profit-and-loss statement, no other business knowledge will be requiredst! Ever!”

I think the two major aspects of the future of the management sciences rest on the following assertions: the technology will advance, but human nature will remain the same.

Most investment houses hire a team (if not an army) of data-savvy analysts (“quants”) who pour over vast amounts of information, repeatedly testing the limits of the accuracy of the saw “correlation is not causation.” Should they find an apparent link, like, say, jellyfish biology to human life expectancy, they will exploit it to its limits for as long as the link is perceived as valid. Like I said, the technology will advance, but human nature will remain the same. It’s why Shakespeare’s works live on as masterpieces while the medical writings of his time – calling for bloodletting and leech-applying – are now considered an embarrassment to the profession.

As for my predictions, I do believe that, eventually, the overextension of the data streams emanating from the general ledger and current risk management theory will be recognized; the GL we will have with us forever (or at least as long as governments need tax revenue, which is the same thing), but the risk managers are vulnerable to having the efficacy of their techniques challenged. Nassim Taleb, in his best-selling book The Black Swan, the Impact of the Highly Improbable (Random House, 2007) makes the case against overuse of Gaussian Curves in business analysis so strongly that I’m surprised that risk management hasn’t seen a significant epistemological retreat since its publishing.

But, until such a correction in the widely-accepted management sciences occurs, it remains a safe prediction that this blog will continue to poke fun at our friends, the accountants and risk managers!

When Emmitt “Doc” Brown utters this line in Back to the Future III, he’s in a saloon in the old (American) west, circa 1885, talking to patrons who listen to his (correct) descriptions of the future with derision. Brown’s descriptions strike them as absurd.

And why wouldn’t they?  In 1885:

·         The pathogen theory of disease transmission was just becoming widely known.

·         Georgia had more people than California.

·         The Statue of Liberty arrived in New York.

·         The mini-Ice Age, which had started around 1300, had just ended.

·         All glass was hand-blown.

·         Teddy Roosevelt was 28 years old.

·         And the list of stuff we take for granted today that hadn’t even been conceived of is prohibitively long for a blog post.

Granted, Back to the Future III is fiction; but the notion that the way the real future unfolds would strike people in history as absurd, I think, is spot-on. It then stands to reason that, should some time traveler come back from our future to the year 2014, his descriptions would strain credulity. That being the case, what should we make of predictions of the future that strike us as entirely reasonable?

During the aforementioned mini-Ice Age, farmers in Northern Europe had quite the challenge. In order to maximize their crop yields, they had to be able to time their ploughing and planting in such a way as to keep their seedlings from freezing (should they plant too soon), but also provide enough time for their crops to ripen before the first freeze of the next Autumn. Correctly timing their crop cycles was literally a life-and-death decision for them and their loved ones. Despite being blessed with no access to climatologists, they still needed to be able to base their decisions on something more than the cycles they had observed in the immediate past. Under a belief that wild animals would often act differently when the last freeze had been encountered, an entire structure of data collection based on such observations became the basis for these critical decisions. Sprinkle in a bit of formality, and a couple of hundred years, and you have the spectacle of men dressed in formal attire gathering on February 2 in Punxsutawney, Pennsylvania, to pull from its den an unwilling member of Marmota momax to “predict” either six more weeks of winter, or an “early” spring.

Now, imagine if I had left out the description of the medieval Northern European farmers’ need to somehow divine the planting season, and also imagine you had never heard of the celebration of Groundhog Day. The question virtually asks itself: is this any way to predict the future?

I suppose it’s appropriate for the December ProjectManagement.com theme to be the Future of Project Management, since every other writer for every other publication uses this time to make predictions. But that’s my point – any attempt to predict the future is simply an extension of our own prejudices and recent past into an unknowable environment. I could predict the outlandish, that someone will come up with a profound breakthrough in the management sciences, but then be hit by falling orbital debris while on her way to publish the findings. Or I could be safe, and predict that the presentations from the next year’s management seminars will be evenly split between those espousing the amazing success of their own particular projects, made possible, don’t you know, by their own insightful embracing of current management techniques, and those who re-package the basics in new presentations, while retaining the eat-your-peas-style that so pervades such presentations.

Or, I could be perfectly honest, and admit: I have no idea what the future of project management looks like.

And my readers should hold as suspect anyone who claims to the contrary.

Not all information is created equal.

A stock broker or securities trader who could time-travel and retrieve next week’s Wall Street Journal would become (almost automatically) rich. However, going into the future and returning the New York Times’ horoscope section is clearly a waste of time (travelling). Of course, most managers do not have access to time machines; rather, they depend on personnel both within and outside the project team to gather the necessary data, process it into information, and deliver the information in such a way that it is usable to the circumstances before them. Since these people’s time and energy is not limitless, the information streams that are to be set up and maintained must be carefully selected.  It is the very care that must be exercised when selecting such information streams that leads to a consulting environment filled with quacks. Here are a couple of examples.

Many accountants insist that project cost performance can be captured by comparing actual costs to budgets. When this folly is pointed out, these will often attempt justification by asserting that the analysis should be carried out at the line-item level of the basis of estimate (BoE). Yet, the flaw in this “analysis” is quickly and clearly revealed when one considers the following scenario.

A project team begins work on a project that had been estimated with $25K (USD) for heavy equipment, and $75K in labor. However, at project completion, they actually ended up spending $70K on equipment, and $30K in labor. The budget – vs. – actual comparison (at the elemental level of the BoE, don’t you know) would be raising red flags over and over, never mind that the project would come in on-budget. Unwilling to admit that the generation of false poor performance warnings ought to be an automatic disqualifier among competing management information systems, the charlatans pushing this analysis, in my experience, do the exact opposite: they insist on even more rigor, finer levels of detail in the compared data sets. If they were turn-of-the-last-century snake oil salesmen, they would insist on the freshness of the snake as a determining factor in the efficacy of their products.

I found myself in the end-hours of a training session with a software company that published a critical-path methodology package that had recently branched out into portfolio management. The difference between their CPM software and this portfolio management package was – I swear I am not making this up – having the CPM platform incorporate some capabilities from the general ledger, like timesheets and travel expenses. I had an opportunity over lunch to talk about this new “capability” with some of this company’s principals.

“Look,” I started, “simply pilfering some capabilities from the general ledger is not what makes for a portfolio management system. You’re better off leaving asset management to the GL, expanding your project management capability to better accommodate earned value, and then create a module that captures strategic information. Only after you can set up a structure that enables all three of these to interact in a coherent fashion can you make a serious case for actually performing portfolio management.” (Such a structure is examined at length in my most recent must-have book, Game Theory in Management.)

I happened to be sitting directly across the table from this company’s then-president, with his veeps on either side, slightly behind him. I could almost see the light bulbs going off above their heads, but the president was having nothing of it. He began describe how much effort his company had poured into the software in the configuration it which it had been released, and that he would not go back and attempt such a correction (our friends, the accountants, would readily recognize such a response as also being invalid, being taught, as they are, to never use the “sunk cost” argument when furthering a business decision).

The problem with competing information systems, some claiming to provide essential “portfolio management” info, is that such claims are next to impossible to verify. Only the manager familiar with both MIS architecture and basic epistemology can consistently navigate these waters.

Or else the readers of ProjectManagement.com’s blog entries.