Tuesday, October 25, 2011

The CHAOS Report (Managing Impossible Projects, Part 3)





Photo: Baggage system, Denver International Airport


The following series is adapted from a keynote I delivered at the Washington, DC, chapter of the Project Management Institute back in August. Parts also come from my book Creative Project Management (with Ted Leemann), published by McGraw-Hill. 



The annual Standish Group’s CHAOS Report, which tracks software project performance reported these abysmal numbers for 2009:

  • 32% on time, on budget, to spec
  • 44% finished late, over budget, or partially completed
  • 24% failed, cancelled, or abandoned

Why do so many projects fail, either in part or in whole?

Certainly, there are badly managed projects — even some headed by PMPs. But it’s hard to blame a 68% defect rate on poor practitioners alone.

Sometimes the problem is the definition of success. By the logic of project management, the Leaning Tower of Pisa was an abysmal failure: late, over budget, and…well, aren’t buildings supposed to be straight? The Standish Group would clearly label the tower as a failed project, yet if the tower didn’t lean, who today would bother to visit?

The third reason, of course, is that the project was problematic to begin with — arguably or actually impossible. This can happen for a variety of reasons. In the case of Apollo 13, the constraints of time, resources, and performance were established by the situation, not by the will or desire of the project managers or owners. They were what they were — whether they were achievable was a separate issue.

In the Battle of the Bulge, the German Ardennes offensive had already begun. We would be able to get troops to the front to relieve the pressure on beleaguered First Army units, or we wouldn’t. Neither project managers nor project owners necessarily control the constraints that drive their projects.

Sometimes we’re just making a guess when we establish a project’s constraints. The original budget for Denver International Airport was $1.2 billion, and the original deadline was October 1993. The final price came in at $4.8 billion and the actual opening date was February 1995. (The infamous automated baggage system, budgeted at $186 million, wasn’t cancelled until 2005, by which time its construction costs were increasing at a staggering $1 million a day!)

Denver wasn’t an example of engineering or technical failure — no, not even the baggage system. The failure was driven by political considerations, including a nonstop war among several key stakeholders. When customer conflict generates mutually exclusive requirements, “impossible” becomes just another word for nothing left to lose.

In any event, we project managers are hired hands. Sometimes we may do double duty as customers or sponsors of our own projects, but when we put on our PMP hats, we’re here not to decide, but to execute. We are bound by the decisions and choices of others, and sometimes we start the project on the precipice of failure. After all, how many of you get to decide your own timelines, set your own budgets, and establish your own performance requirement?

I didn’t think so.

Still, you don’t want to be too quick to pull the trigger. Let’s imagine that you have a project and your experience tells you it can’t be done. Isn’t delaying the inevitable bad news just going to make the problem worse?

That depends on what you do in between receiving the assignment and giving the answer. Even if your project’s impossible, or at least compromised, there’s still a customer problem needing to be solved. Telling people what they can do and what they can have tends to get a better reception than telling people what they can’t do and what they can’t have.

That’s why the first step in managing a potentially impossible project is analysis. When you analyze an apparently impossible or potentially impossible project, you may learn different things.


  1. You confirm that the project is in fact impossible, and can provide evidence to the customer. You and the customer can begin to figure out what alternatives may exist or how to deal with the consequences of an unsuccessful project.
  2. You confirm that the project as originally proposed is in fact impossible, but are able to find potential changes that will make the project possible, which you can present to the customer.
  3. You confirm that the project as stated is in fact impossible, but are able to offer alternatives and compromises that might satisfy at least some of the customer’s requirements and needs, or close part of the gap.
  4. You can’t confirm that the project is in fact impossible, but you can identify at least some of the risks and challenges you face, which you and the customer can then assess.
  5. You find a creative way around the barrier that made the project impossible, and achieve the original goal.

Even if your analysis leads to the first outcome (it’s just flat impossible), however, your situation is still improved by your ability to give a thoughtful reply with supporting evidence, and your attitude in making a good-faith attempt to solve the problem.

Partial successes (outcomes two, three, and four) are a marked improvement. Even if the project is impossible — or highly risky — as stated, the customer may be able to get a significant portion of what he or she wants. Plus, it’s well known that the first approximation of available constraints may not be the final word. There may be more to draw on. And again, people tend to react better to hearing what they can have, and less well to hearing what they can’t have.

We renovated our house last summer, and the project was on time, exceeded our expectations — and cost about twice as much as we’d planned. We still call it a rousing success, because we never really expected to meet the budget anyway. It was a hope, not a realistic assessment. The Standish Group would call it a failure, but we don’t — and the customer is always right.

The fifth outcome (solve the problem with creativity) is ideal, but often challenging and not always successful.  The best direction to find the creative answer is, paradoxically, to focus on the barriers in the first place.



Next Week: The Power of Negative Thinking!

Tuesday, October 18, 2011

Potentially Impossible (Managing Impossible Projects, Part 2)

The following series is adapted from a keynote I delivered at the Washington, DC, chapter of the Project Management Institute back in August. Parts also come from my book Creative Project Management (with Ted Leemann), published by McGraw-Hill. 


The Most Dangerous Word

At the outset of the project, you may not always know whether the project is possible or not. That’s why the process of managing an impossible — or potentially impossible — project begins at the outset, during the very first stages of project initiation.

Here’s one of Dobson’s Laws of Project Management: The most dangerous word in project management is a premature “yes.”

Premature certainty, whether it’s positive or negative, can backfire. Saying “yes” before you really know what you’ve said “yes” to can result in a world of trouble.

You can also get in a world of trouble by being too quick to say “no.” Even if your experience and wisdom tell you the project’s impossible on the face of it, saying so too quickly will produce a negative reaction. And, frankly, sometimes “no” is just not going to be an acceptable answer.

When you say, “Sorry, that’s impossible,” they think, “You didn’t even try!”

Failure, Alas, Is an Option 


Let’s set the Way-Bac machine for April 14, 1970, just about 56 hours into the flight of Apollo 13. With the spacecraft about 200,000 miles away from Earth, Mission Control asked the crew to turn on the hydrogen and oxygen tank stirring fans. About 93 seconds later there was a loud “bang.” Oxygen tank #2 had exploded.


If American Movie Classics ever ran a series of “Great Project Management Movies,” surely Ron Howard’s Apollo 13 would be a natural candidate. Faced with a potentially disastrous accident, project teams overcome one potentially fatal barrier after another to bring the crew safely back to earth, guided by mission director Gene Kranz’s mantra, “Failure is not an option.”

But of course failure is an option, and in the case of the Apollo 13 mission, the odds were heavily stacked against a happy outcome— and everybody (including Gene Kranz) had to be well aware of that fact. Within the overall project “get the astronauts home safely,” there were numerous subprojects, including:

  • Develop a power-up sequence that draws fewer than 20 amps
  • Calculate a burn rate to get the reentry angle within tolerance using the sight of the Earth in the capsule window as the sole reference point
  • Design a way to fit the square command module carbon dioxide scrubber filter into the round Lunar Excursion Module (LEM) filter socket.

That last subproject was vital, because the LEM’s carbon dioxide scrubbers were designed to take care of he needs of two people for a day and a half, not three people for three days. And nobody ever imagined that the command module scrubbers would need to be used in the LEM, so they weren’t designed to be compatible. They’re square, and the necessary holes are round. Meanwhile, the carbon dioxide levels are already past 8, and at 15 things become dangerous, and eventually deadly. As the engineers gather in a conference room, boxes of miscellaneous junk — basically everything that’s loose on board the spacecraft — are being dumped on tables.

Managing a Crisis Project

Let’s look at the project management problem.

It’s easy to build a carbon dioxide filter on Earth; there’s a standard specification, a deadline measured in weeks, if not months, and all the resources you need are easy to acquire. In a crisis situation, such as existed aboard Apollo 13, the project looks a little different.

At the beginning of the project, the engineers involved could not know whether the project would turn out to be ultimately impossible. Impossibility could exist in any of the three fundamental constraints.

  • Does the time available to accomplish the project equal or exceed the time necessary? 

In developing a replacement for the Apollo 13 mission’s overloaded carbon dioxide filter, engineers were constrained by the amount of time until the astronauts became too impaired to build what they designed. If the deadline turns out to be too short, then the project is impossible.

  • Are the resources needed to accomplish the project less than or equal to the resources available? 

The project was constrained by what was actually available on the spacecraft. If their resources are short by even one critical component, no matter how small — a 20¢ screw — the project is impossible.

  • Are the performance criteria achievable within the outer boundaries of the other constraints? 

If the improvised filter can’t be made to work long enough for the astronauts to reach Earth orbit when they can return to the command module, then the project is impossible.

As we all know, the Apollo 13 engineers did come up with a workable solution, but that was hardly guaranteed. Had the constraints been slightly different — less time, fewer resources, more challenging performance standards — the outcome would likely have been failure.

But that doesn't mean you shouldn't try.


Next Week: The CHAOS Report!

Tuesday, October 11, 2011

If I Had a Lever (Managing Impossible Projects, Part 1)

The following series is adapted from a keynote I delivered at the Washington, DC, chapter of the Project Management Institute back in August. Parts also come from my book Creative Project Management (with Ted Leemann), published by McGraw-Hill. 



If I Had a Lever...

When we say “nothing’s impossible,” we usually mean that given unlimited time, unlimited resources, and really flexible performance standards, we can do anything. “Give me a lever long enough and a platform to rest it on, and I will move the world,” said Archimedes, but he was obviously not a project manager. Our projects are constrained: the iron triangle of resources, time, and mandatory scope are only three of the dimensions that restrict our options.

There are many types of impossibility: legal impossibility, scientific impossibility, metaphysical impossibility, and logical impossibility, to name a few. Each has its own definition and its own specific context. Our question is more focused: what does “impossible” mean in the context of project management — and more importantly, in the context of your project?

We define a project as “operationally impossible” if it cannot be accomplished within the boundaries of its mandatory constraints. Of course, seemingly “impossible” projects succeed all the time, and there are a number of proven strategies that work — at least part of the time.


  • Sometimes a team discovers a brilliant critical insight, or is simply smart enough and good enough to achieve what lesser mortals cannot.
  • Sometimes the team gets it done by sheer Herculean effort, working harder and longer than anyone expects. The project succeeds, but sometimes the organization pays a long-term price.
  • Sometimes the team gets it done, but the outcome is compromised. Maybe the project cost more, or took longer, or did less. Sometimes we can point to the corpses of the projects we sacrificed in order to make the current project succeed
  • And sometimes the team fakes it, slaps a coat of paint on it, and hopes nobody notices that the wheels have fallen off. (Make sure your résumé is up to date first.)


There are, fortunately, some better answers, and in this and the next few blog posts we’ll journey through history to see what lessons we can pick up.

Patton and the Bulge

Let’s set the Way-Bac machine for December 19, 1944, where the Allied High Command is meeting in Verdun to plan its response to the German offensive Wacht am Rhein, popularly known as the Battle of the Bulge. 

Elements of the United States First Army, supposedly in a quiet sector of the front, are pinned down in Bastogne. Eisenhower asks the assembled generals how long it will be before Allied forces will be able to relieve the beleaguered Americans at Bastogne. British Field Marshal Bernard Montgomery says it will take weeks.

George Patton, commander of the United States Third Army, jumps in. . (Patton, by the way, sounded a lot less like George C. Scott than he did like Ross Perot.)

“I can attack with three divisions in 48 hours,” he says.

The response from the other generals was not polite.

Patton’s boss, General Omar Bradley, was not amused. “Ike wants a realistic estimate, George. You’re in the middle of a fight now. It’s over a hundred miles to Bastogne.”

They were, of course, right to be skeptical. Extricating three divisions from a tough fight and moving them 100 miles in 48 hours?  That’s not just difficult, that’s downright impossible. 

Let’s see why.

A division is an Army unit consisting of approximately 15,000 soldiers, along with everything they need to do their job. Imagine picking up a town of 45,000, along with all the services needed to keep them going, and moving 100 miles in 48 hours…and forget the interstates; there aren’t any. Just for starters, if you don't have a detailed movement plan, you'll end up with the world's biggest traffic jam. 

Armored vehicles are gas-guzzlers, people have to eat, and soldiers need ammunition. That means you'll have to pre-position gas, food and supplies along the route. 

A moving division is more vulnerable than a division on defense. That means you need fighting units to protect moving units, and they need more gas and food and ammunition.

A move of this nature requires a planning staff in the hundreds. In World War II, without cell phones, laptops, and GPS units, orders were typed on mimeograph stencils, duplicated, and hand-carried to unit commanders stretched out over an immense area. Today's technology is far superior, but so are the demands involved.

It takes weeks to pull off an operation like this. It really can’t be done in 48 hours. It’s an impossible project — flat out impossible.

And yet it was done.

But wait a minute. If it was done, then wasn’t it by definition possible?

Like with any good magic trick, the key often lies in challenging your assumptions. Eisenhower’s headquarters learned about the German Ardennes offensive late in the game, and that’s why Patton needed to move within 48 hours. But Patton, alone among senior Allied commanders, had anticipated the possibility, deployed his own intelligence resources, identified the threat, and bought himself the extra time he needed. He didn’t do it in 48 hours. He changed the constraints.

There’s a scene earlier in the movie in which he instructs his staff to begin planning the move northward. His staff had several weeks to prepare three different contingency plans. All Patton had to do when the meeting broke up was walk downstairs to his jeep, call his headquarters on the radio, say “Nickle,” and the forces were on their way. (His driver, Sergeant Mims, reportedly said, “I don’t know why they need all them other generals. You and me can run this whole war out of your jeep.”)

If a project can’t be done within its constraints, one obvious approach is to follow Patton’s example, and alter the constraints, but of course, that’s not always an option. Sometimes, the project is inherently likely to fail — but that doesn’t mean you don’t still have to manage it.

Next week — Failure IS An Option!

Tuesday, October 4, 2011

Riddikulus! (Part 14 of Fallacies)

More red herrings, argumentative fallacies that distract from the argument rather than address it directly. This week, the final three appeals to emotion: the appeal to ridicule, the appeal to spite, and wishful thinking.

Reductio Ad Ridiculum

Riddikulus! As all Harry Potter fans know, the way to defeat a boggart is to convert it from an object of terror to an object of mockery. While the spell clearly works, in real life, the appeal to ridicule is a type of red herring fallacy in which the opponent presents the original argument in a way that turns it into a mockery of itself, either by emphasizing the counter-intuitive aspects of the original argument, or by creating a straw man to debunk it.

An example of the first approach is the argument, “If Einstein's theory of relativity is right, that would mean that when I drive my car it gets shorter and more massive the faster I go. That's crazy!” It’s also true. The problem is that the effects are not easily measured at automobile speeds, but only become significant as the object nears the speed of light.

The second approach misrepresents the argument in order to ridicule it. “If evolution were true, that would mean that all the apes wouldn't be here any more, since they all would have evolved into humans!” That’s ridiculous indeed — but it’s not actually implied or stated in the Theory of Evolution.

Argumentum Ad Odium

The appeal to spite exploits existing bitterness or dislike in its attack. The various attacks on union benefits (such as retirement), particularly in government workers, relies on the negative emotions aimed at the target group as the primary justification for cutting back or cancelling previously agreed-upon benefits. “Why should people enjoy a comfortable retirement with my tax dollars?”

Wishful Thinking

Wishful thinking is based on the premise “I wish P were true/false, therefore P is true/false.” You see this in a lot of superstitious behavior, from chain letters to the belief in UFOs. Personally, I think it would be really cool if aliens did in fact visit Earth — but that doesn’t make it true.

Tuesday, September 27, 2011

Making the Kessel Run in Under 12 Parsecs

From xkcd, by Randall Munroe.

On the OPERA neutrino experiment, the conservative war on relativity, and the operation of science (with a nod to the biggest science boo-boo in Star Wars).

One can understand, if not necessarily agree with, the conservative war on the Theory of Evolution. If you believe that a specific book must be taken literally, and that book states that the Earth was created in seven days, then it is impossible for Darwinian evolution to have taken place. Either the book in question is not literally true, or the theory must be wrong. You can’t have both.

I must admit, though, to being perplexed by the conservative opposition to the Theory of Relativity. I first encountered this when one of my White House speechwriter friends asked me to review an unpublished manuscript ostensibly debunking Einstein. (It was written by a political science major.) The book was filled with bold assertions, dismissal of contrary evidence, and an ongoing hint that relativity was enthroned not because of its merits as a theory, but because of an ongoing conspiracy by what xkcd refers to as the “Science Thought Police.” In other words, it was just like the anti-evolution arguments.

The objection to relativity seems to rest on a semi-religious foundation: the idea that reality is in some sense objective, without uncertainty or variability. There’s an existential threat in the idea that some parts of reality are subjective, variable, or…well, relative. More importantly, the objection goes to the heart of the idea of science itself, the idea that the experimental method, peer review, and testable hypotheses can separate the valid from the invalid. For those whose authority rests on a foundation of “truth,” testability can be quite inconvenient.

Which brings us to the OPERA neutrino experiment, in which the underground Gran Sasso Laboratory measured the velocity of neutrinos from CERN at moving slightly faster than the speed of light (c). The difference is not great — 60 nanoseconds faster over a distance of 730 kilometers (with an expected error of ±10 nanoseconds) — but any evidence that a particle can exceed the speed of light forms a huge challenge to the foundations of modern theoretical physics. (For the curious, the official paper can be found here.)

Not so fast. The researchers themselves have not made any claims so bold. Before publishing, they looked for experimental errors. They re-ran the experiment. They looked at a variety of potential explanations and controlled for as many variables as they could. By publishing the results, they aren’t making the claim that they’ve refuted Einstein, but rather quite the opposite: they are appealing to the scientific community to repeat the experiment and see if they can discover why the OPERA results are incorrect.

That’s as it should be. Given the overwhelming body of evidence post-Einstein that confirms the principles of relativity, an anomalous result deserves to be treated with skepticism. Extraordinary claims, as they say, require extraordinary proof.

In fact, there’s already a strong experimental case that undercuts the OPERA results. On February 23, 1987, light from a supernova in the Large Magellanic Cloud, one of the two dwarf galaxies that orbit our own Milky Way, reached Earth after traveling approximately 168,000 light years. It was the closest supernova to Earth since Kepler’s Supernova of 1604, and the first since then to be visible to the naked eye.

About three hours before the visible light from SN1987A reached Earth, neutrino bursts were observed at three different neutrino observatories. This doesn’t mean those neutrinos traveled faster than light; the visible light produced by a supernova’s collapsing core has to travel upward to the stellar surface, while neutrinos zip right through the intervening material. If those neutrinos were traveling as fast as the OPERA observations indicate (60 ns per 730 km), the neutrino bursts should have arrived nearly four years before the visible light.

What, then, do the OPERA results mean? One possibility — the likeliest, according to most blogging scientists I’ve been reading in the past few days — is that a measurement error did in fact occur. If there’s not a measurement error, and it turns out that the neutrinos do move slightly faster than our current measurement of c, it still won’t “refute” Einstein -- our GPS units (which rely on relativity equations) still work. Einstein, after all, didn’t “refute” Isaac Newton. Newton’s laws work quite well in the majority of cases. It’s only at extreme speeds and under unusual conditions that relativistic differences matter.

What it will mean, if the results hold up under scrutiny, is that adjustments will be made. The models we use to understand the universe will require modification, becoming more accurate. There’s a small chance we’ll discover some important breakthrough that changes our understanding of the universe in materially significant ways.

In that case, the better parallel will be the 1887 experiments by Albert Michelson and Edward Morley involving the measurement of the speed of light in different directions, with the goal of measuring the motion of the Earth with respect to the luminiferous aether. They also didn’t get the results their theory predicted. The difference here was also small enough to be attributed to experimental error, but over time, as the results became more and more solid, it finally became clear that there was something wrong with the aether hypothesis. 

Although Einstein’s theory of relativity did not rest on “science’s most famous failed experiment,” Michelson-Morley did serve as evidence that helped the new theory of a constant speed of light gain widespread acceptance.

Tuesday, September 20, 2011

Dives and Lazarus (Part 13 of Fallacies)

More red herrings, argumentative fallacies that distract from the argument rather than address it directly.

Argumentum ad crumenam

If you’re so damn smart, why ain’t you rich?

The argumentum ad crumenam, or argument to the purse, suggests that the truth of the proposition can be supported by the wealth of the speaker. If you’re so smart, why ain’t you rich? In other words, if you’re rich, you must be smart.

The rebuttal to this argumentative red herring can be made in only two words:

Donald Trump.

Argumentum ad lazarum

The reverse is known as the appeal to poverty. It takes its name from the parable of the rich man and Lazarus (Luke 16:19-31), in which the rich man suffers the torments of Hades while the beggar Lazarus enjoys the delights of heaven.

While there’s significant Biblical support for the comparative virtue of poor versus rich (see Matthew 19:24, “And again I say to you, It is easier for a camel to go through the eye of a needle, than for a rich man to enter into the kingdom of God.”), virtue and logical argument don’t necessarily correlate. If it’s not necessarily true because a rich person says it, it’s no more true if a poor one does.

Tuesday, September 13, 2011

Known by the Company We Keep (Part 12 of Fallacies)

In the next part of our continuing survey of red herrings (responses to arguments that don’t address the actual argument but merely distract from it, we’ll look at the two types of association fallacies: guilt by association and honor by association. Depending on your point of view, they can be one and the same.

Guilt by Association

Association fallacies take the following form: (1) A is a B. (2) A is also a C. (3) Therefore, all Bs are also Cs. (More formally, (∃x ∈ S : φ(x)) → (∀x ∈ S : φ(x)), which means “if there exists any x in the set S so that a property φ is true for x, then for all x in S the property φ must be true.”

Of course, that’s not at all a necessary condition. The classic rebuttal goes like this:

(1) All dogs have four legs.
(2) My cat has four legs.
(3) Therefore, my cat is a dog.

The PolitiFact Truth-O-Meter recently gave a “Pants On Fire” rating to an August 17 blog post by Texas radio host Dan Cofall, which read in part, “The magic number ‘70’ is the number of members of the 111th Congress who are members of the Democratic Socialists of America (DSA). These are not just politicians who vote left of center; these are card-carrying members of ‘The Democratic Socialists of America.’” (The "70" to which Cofall refers is the membership of the Congressional Progressive Caucus — though I do not believe they actually issue membership cards.)

There are two guilt-by-association attacks here, one direct and one indirect. The indirect attack is in the term “card-carrying,” an echo of McCarthy-era HUAC anti-communist campaigning. The implication goes like this:

(1) Dues-paying members of the Communist Party carry cards.
(2) Members of Group X (Democratic Socialists, ACLU, etc.) carry cards.
(3) Therefore, members of Group X are Communists.

The direct attack takes this form:

(1) The  Democratic Socialists of America have a platform with a number of ideas.
(2) Some members of the Congressional Progressive Congress have ideas that overlap with some items on the DSA agenda.
(3) Therefore, all members of the Congressional Progressive Congress are "card carrying" members of the Democratic Socialists of America.

Of course, liberal Democrats and Tea Party members also share some specific ideas (they both like the idea of voting, for example), but it hardly follows that all liberal Democrats are Tea Party members, or vice versa.

The Democratic Socialists are somewhat chagrined. “If we had formal political relationships with 70-odd members [of Congress], we would be making a lot more money” from dues. And as far as they’re concerned, the problem with the Congressional Progressive Congress is that the members aren’t nearly socialist enough — they prefer a third party movement.

Whether it’s “guilt by association” or “honor by association” may depend on your point of view. When Bill O’Reilly said on his January 19, 2005, broadcast, “Hitler would be a card-carrying ACLU member. So would Stalin. Castro probably is. And so would Mao Zedong,” I decided to look at it this way:

(1) Bill O’Reilly and George Bush say bad things about “card carrying” ACLU members.
(2) I think O’Reilly and his fellow-travelers are jackasses.
(3) Therefore, I joined the ACLU…just so I can carry my card.