Espionage and Statistical Inference


by ASQ Member Jay Armstrong

We don’t usually confuse statisticians with spies, and yet, statistics played a significant role in the defeat of Germany in the Second World War.  By the 1930s, Germany’s pre-war industrial prowess and engineering capabilities were both mature and formidable.  This was particularly true for her various weapon systems.  With D-Day looming, of greatest concern were the newest Mark V “Panther” series tanks. Unfortunately, these were superior to the allies’ tanks across all performance dimensions – armor, cannon size, speed, and mobility. Beyond this painful reality, was the critical need to understand the actual number of tanks Germany was manufacturing in order to develop effective tactics and counter-measures.  This is where statistics – or more precisely – the point estimation formula came in.  The Germans, notorious as they are for order and precision, thoughtfully applied serial numbers to their equipment, which exactly matched the production date for that item.  This compulsion would provide the allies with a solution to their problem.

The idea that serial numbers could provide insights into material production output first arose in 1943 and was immediately used by the Economic warfare Division to gauge German tire yields. More creative statisticians soon realized that such thinking could be leveraged against all military output and began asking for the serial numbers from captured and destroyed tanks.  The front-line soldiers, although annoyed at what seemed to be a trifle, complied.  Let’s review the statistical underpinnings of this thinking with an example.  Suppose that I have the serial numbers for 20 destroyed tanks in Table 1.


What is a reasonable estimate of the total number of tanks that have been produced, to date?

Or statistically speaking, I want to perform an: “estimation of the maximum point of a discrete uniform distribution using sampling without replacement.”

Our useful formula, the “Maximum Likelihood Estimator” is:


Where:   = the estimated total number of tanks

n-in-formula= the estimated total number of tanks

m = the highest obtained serial number

n =  the sample size (my destroyed tanks)

I note from my sample that “669” is the highest serial number in the series, and thus becomes “m.” My sample size is “20” and is therefore, “n.” Substituting, I determine that  equals 701.45, or approximately 700 tanks.

After the war, the statistical calculations proved to be far more accurate and reliable than did the traditional methods of SWAG and intelligence gathering at German factories (Table 2).  The actual monthly production number – 245 tanks/month, matched the statistical prediction of 246 tanks/month.  Depressingly, initial Allied intelligence had suggested an astounding 1400 tanks were being produced monthly.  In fact, the Germans had cleverly contributed to this numerical confusion by painting and repainting higher numbers on their tank turrets to confuse spies.


New Location Beginning this Month’s Leadership Team Meeting

Come join the leadership team this month at DeVry University, Fort Washington campus, this Wednesday, October 12, 2016 at 6:00 PM.

The address (for those who haven’t been there before) is 1140 Virginia Drive, Fort Washington, PA. There is construction being done at the University entrance so you will need to enter the complex from Ditech (the first light up the street) and drive around the buildings until you reach the front entrance of DeVry. Once you enter the building please check in with the guard at the door. Then proceed through the Atrium past the library and turn left to room 124.

A complimentary light dinner is served.

Indiana Jones in the Temple of Portfolio Management

By Jay Armstrong

Imagine that you are Indiana Jones – a series of ancient doors before you – behind one – unimaginable success, profitability and reward. Behind the others – defeat, loss and failure. You have but a single key and too many possible doors – which to choose? Does this sound like your organization? This is the constant dilemma facing those inindiana portfolio management. With exploding R&D costs, falling productivity, high failure rates and persistent competition from generics, the need for effective portfolio management in the pharmaceuticals is a forgone conclusion. While every drug company utilizes some management process to ensure the success of their portfolio, certain approaches can be substantially more effective than others.

The Nature of the Beast

Historically, organizations have prioritized and applied resources against what were the major (or most politically connected), projects in their portfolio – sound familiar? Although this is the easiest and seemingly most obvious solution, it is not always the best or most efficient one. Projects evolve, expectations change, results and requirements conspire over time, all derailing the best laid initial plans and portfolios. By returning to the process that got you there and then reapplying that thinking, you only create more issues and deeper frustrations – the cycle of performance despair.

Resource directed portfolio management has enabled organizations to extract a competitive advantage from the application of valuable, unique and usually, scarce resources. Traditionally, resources are your human assets, financial strengths, processes or knowledge brought to bear by your organization to fulfill its strategic initiatives. This definition illuminates an underlying problem faced by pharmaceutical companies – various factors must intelligently be brought to bear in a complex and dynamic reality to produce a profitable outcome within an acceptable time-frame, under regulatory scrutiny. Unfortunately, all of this usually occurs in a vacuum of understanding around the organization’s resources potential, leading to poorly supported or weakly delivered project objectives. So, which approach(s) make the most sense and can consistently provide the highest ROI and prospects for success ?

The Process

Obviously, your initial strategic and organizational assessments must be robust – therapeutic areas, finances, markets, production and R&D capability and internal and external variables – all must be understood, evaluated and tracked well before portfolio planning begins. Only then, can the heavy lifting of portfolio resource management begin. Your first step should be to create a benchmarking strategy to obtain insights into the best practices in portfolio resource management. By reviewing how those in other industries and companies operate, you can compress your learning curve, establish solid performance metrics and save valuable time. It’s best not to restrict yourself to the pharmaceutical industry for ideas – the financial, manufacturing and software industries can all provide ah-ha moments and creative solutions to your planning problems. In addition to providing your portfolio teams with a solid lift-off, you can gain a deeper sense into how good portfolio planning should perform and execute.

A good second step is to identify all of the available or potential resource options that you have which match your anticipated portfolio plan. Are there resources in other sites, countries or departments that could be applied in the portfolio? Should you hire-in, or train your people, develop in-house, or purchase new capabilities and technologies? Can existing resources, tools or processes in one area be re-purposed to support another, new one? Do you need full resource support, or can people be shared, or rolled on/rolled off for peak demand periods? Many organizations fail to adequately understand or gauge the full breadth and depth of their potential resources and thereby artificially limit what they could accomplish. When you begin with assumptions or suppositions, you have already significantly undermined the possible capability of your alternatives. Use a broad based portfolio team from different sites, departments or countries to ensure that all options have an equal probability for consideration.

The third step requires an evaluation of how those resource options can best be leveraged across the portfolio mix. Which resources and in what combinations provide the greatest ROI with the lowest cost?Will they change over time? How can I measure effectiveness and impact? This is where software and tools such as decision trees, greedy algorithms, resource matrices or scenario modelling may be particularly impactful. Singly, or together, these instruments will allow you to ask various what-if questions, without the burden of having to learn from failure. Because they execute brute-force quantitative analyses quite well, these approaches are seductive, but be careful to evaluate the outcomes against your internal experts for a reality and fitness check. Once you have created an optimal resource mix and launched, it is critical that you are able to measure the performance and outcomes of your portfolio.

An important overarching continuous responsibility is having the capability to carry out a real time analysis of your portfolio performance. This provides an appropriate answer to the How are we doing right now? question. It enables quick responses to unexpected outcomes or evolving conditions and requirements that are inherent in any portfolio. In fact, it may be wise to develop short-term and long-term forecasts to optimize staffing over time as conditions and demands change. Additionally, this philosophy allows you to compare, balance and re-balance the portfolio resource demands against financial costs, outcomes and skills capability.

Finally, any reasonably good portfolio resource management plan must have an internal stage-gate review process. This can be as simple as a quarterly meeting of the principals crafting a go/no go continuation decision based on the current performance metrics. Nothing is more financially draining or demotivating to an organization than an unbalanced, weak portfolio, or initiatives that hang around long past their expired by date. It drains and misapplies critical resources, while wasting time and money that could be better applied elsewhere.

Effective portfolio resource management is a dynamic beast, it requires insight, flexibility and responsiveness to be able to balance and extract the full value from your portfolio. Solid portfolio planning beforehand, helps to minimize failures and creates a culture of excellence that is both sustainable and highly profitable.

Jay Armstrong holds certification as an ASQ Six Sigma Black Belt and advanced degrees from Stanford, University of Pennsylvania, Johns Hopkins in fields such as Biotechnology, Management, Biomedical Technology. He holds certificates in Project Management, and Decision/Risk Management. He has led workshops in employee coaching, designing and fundamentals with led to eight Kaizen events, and five discovery events. He was the leader in launching of the process workshops for clinical trials, and developed and led a clinical trials E-2-E work-stream which identified 10 clinical process area improvements. He is responsible for leading of five R&D process-improvement teams in delivering major cycle time, cost, and resource reductions of up to 80%, leading to $11M savings.