INFORMS Annual Meeting
Foundations of Modern Predictive Analytics
Data Analytics for Action & Impact: Transforming Data to Goal-Driven Insight for the Data-Rich yet Information-Poor
Predictive Analytics and Data Mining Strategic Implementation Course Series: A Comprehensive and Active Experience of the Full Modeling Practice Framework
Special ArticlesStudy: When it comes to sales & marketing analytics, few companies ‘get it right’
Companies across industries admit to the growing importance of data analytics to improve sales and marketing effectiveness and decision-making. However, many struggle to piece together siloed data, properly define the problem or design the solution. As a result, they often fail to realize widespread business impact from their efforts, according to a study conducted by The Economist Intelligence Unit (EIU) and sponsored by global sales and marketing firm ZS.Read More
Special Articles2017 Syngenta Crop Challenge in Analytics offers $5,000 prize
Put your analytics skills to the test and win $5,000 in the 2017 “Syngenta Crop Challenge in Analytics” administered by INFORMS and sponsored by the Analytics Society of INFORMS.Read More
Special ArticlesITIF: Expand computer science education to keep up with demand
Computer skills are in high demand among employers in a wide range of industries, not just tech-related fields, yet despite growing interest in the subject, a new report from the Information Technology and Innovation Foundation (ITIF) finds that too few U.S. students are taking quality computer science classes at the high school and university levels. ITIF makes the case for public action to support and maintain the groundswell of interest in computer science and capture the economic and social benefits that will come from fostering a more highly skilled workforce.Read More
Corporate Strategy & Security
Wargaming for fun and profit
A new tool for competitive analytics.
By Douglas A. Samuelson
Decision-makers trying to understand ill-defined problems, in particular the assessment of likely competitors, are increasingly turning to an unusual tool in the analytics kit: wargaming. Strategic games are expanding from defense applications, where they have been widely used for years, into corporate strategy and security.
Mark Herman, an executive vice president at Booz Allen Hamilton, is a leading authority on wargaming and a long-time leader of the company’s strategic gaming practice. Several people familiar with high-level Department of Defense decision-making called this group “the most influential strategic gaming group in the country.”  In his book , Herman and two of his colleagues described applications from the planning of the two Gulf Wars to advising one major oil company how to position itself to lower its risk and hence to profit during a disruption of global oil supply. More recently Herman has advised a number of large corporations about cyber-security.
The essential idea behind corporate wargaming is to help participants become aware of issues they did not even suspect they had. As Herman puts it, “Wargames are a process where you have an undefined set of data, an undefined set of situations, and you’re trying to leverage collective corporate knowledge to put some structure around the problem. Wargaming is a qualitative technique that makes it easy for the available expertise to drive you toward a structured way of asking the question.”
Harvard Business School case studies embody the same idea: start teams with some basic assumptions, an appreciation of the situation and some rules for interaction, and see what behaviors and outcomes emerge.
Another recent book, Benjamin Gilad’s “Business War Games,”  focuses more on business applications, although it included some of the more intriguing military examples as well. “War games in business are hot,” Gilad writes. Among other examples, he discusses in some detail how Disney/ABC/ESPN could have anticipated the moves that built Fox Sports, at the Disney group’s expense, if they had used wargaming to understand what their competition could do and what they might do to respond. Gilad also describes how IBM’s “blind spot” about likely competition ended up ceding the personal computers market to Dell and some other smaller providers, and how similar inattention to competitors’ best likely moves cost Ticketmaster its dominance in the direct event ticket resale market.
“Identifying a competitor’s blind spots is the highest form of intelligence analysis,” he summarizes. “Identifying competitors’ blinders allows, in the next step of a wargame, to design moves that deliberately take advantage of competitors’ denial and ignorance.”
The Cyber Threat
In a more recent interview, Herman elaborates about cyber-security. “A lot depends on how you think about policy and behaviors of people. For example, look at Wikileaks. The U.S. government gets some ambassador’s report, in not politically correct language, no coloring. A company’s internal e-mails have the same concerns. For instance, you might be having a candid conversation about how to get more margin from a trading partner or more market share from a competitor. You value candor, but you don’t want to be embarrassed by seeing these communications in the New York Times. So just as CNN changed us to 24/7 news and changed news cycles, now we have to rethink the best way to handle internal communications.”
He adds, “Now we also have to think about how we talk about how we’d handle a crisis we haven’t seen before. It’s my personal view that some ‘whistleblowers’ are manipulated by others who have a profit motive, such as selling stock short, and there will be more of this kind of manipulation in the future. So how do I run my business with these issues, too, in mind?”
Herman points out that management usually can’t see in advance second- and third-order effects of either cyber-attacks or the defenses the company employs. “You can build a network and have your hardware people fight it out, testing technologies,” he suggests, “and you can wargame vulnerabilities, then change policies based on understanding where you could be attacked. You can do things like quickly check when someone starts downloading much more than he or she usually does – this might be a person who’s guessing he’s about to be fired, grabbing sensitive data before the company shots off access. Or you might restrict peer-to-peer connections using the company’s computers, since those can create paths into your data, but you don’t want to go too far and interfere with productivity. With good understanding, you can stay both more competitive and more secure.”
“Then,” he adds, “you also need to look at larger policy questions. You can start to plan strategy to ameliorate a leak beforehand – clean up e-mails, use kinder, gentler language, expect to operate in a more transparent world. And then look at how you can gain competitive advantage in this changed environment, such as using digital media to gain in the market. Wargames are ideal for this because there’s no computer program that’s going to tell you the answers to those questions.”
Is There High-End Analytics Inside?
Like computational modeling of organizations, wargaming has deep roots in high-end analytics such as operations research (O.R.), but the two disciplines have diverged. Herman acknowledges, “I’m not a quantitative guy per se, just logic-driven.”
In his view, quantitative methods have declined in influence in the strategy-making community, as simple, qualitative methods, including the low-tech style of wargaming, have proven more effective. However, not everyone in the wargaming community shares quite so pessimistic a perspective. Tom McCormick, senior analyst at Alpha Informatics, also deeply experienced in the analyses that support DoD policy and doctrine, notes, “You’re trying to predict what resources you’ll need, how much damage there will be, how many bodies. To think of anyone trying to do that without math is just scary.”
Still, many other people familiar with the current set of issues and methods do share the view that multi-disciplinary methods, with limited math, work best. Herman started his career as a commercial games designer but now favors manual games over the computer-based ones, as manual games are easier to modify and replay as new ideas emerge. “If you go back to the origins of O.R.,” Herman says, “it was a multi-disciplinary approach that eventually ossified into a math-physics focus. If we went back to our roots, we’d be better off.” Still, he says, “I’ve been working the problem” of convincing clients that more quantitative analyses would be helpful. And, he acknowledges, “after wargaming, we often have to bring other analytic techniques into it to follow up quantitatively.”
A promising, somewhat more quantitative method is the extension of game analysis to games within games. If, for example, some of “Blue’s” possible actions consist of deceiving “Red” about what game they are playing, then we have a hypergame with a number of traditional games as component parts. Peter Bennett first advanced this idea in the late 1970s in two journals, Omega  and The Journal of the Operational Research Society. During the past decade, Russ Vane elaborated the concept in considerable detail. [1, 8] Vane is a former U. S. armored cavalry officer and ranger who now works as an analyst and researcher for IBM, supporting mostly national security arena clients. “One of the most valuable things you can do,” he explains, “is ask: if the other guy knew my plans, what’s the worst thing he could do to me? Now, how could I counter that?”
This approach, in turn, generates some interesting questions about how to validate modeling assumptions, how to decide whether results are reasonable and the very nature of what constitutes evidence. A multi-level games-within-games structure (“hyper-hypergames”)  provides greater realism but requires much more attention to how well both players and game-runners are handling context and interpretation. If, as some physicists argue, all apparent randomness is just a measure of our lack of knowledge, then the role for applied probability in large-scale strategic hypergames is likely to remain quite large. Some commentators, notably David Schum at George Mason University, have proposed that a full reexamination of what evidence is, and how to decide how much weight to give it, would be a useful area of research. Russ Vane concurs.
Wargaming can also interact with models of organizational decision-making, in particular agent-based depictions of decision processes.  These models, in turn, can suggest courses of actions to be tried in games.
How to Do It
Herman, Gilad and Vane all stressed the importance of certain key components for wargaming to be useful:
- • Develop scenarios that push participants to get into the full range of issues, without getting hung up on how realistic some aspects of the scenario are. The idea is to get people to propose courses of action that haven’t been contemplated. Preparing good scenarios is probably the area in which wargaming expertise based on prior experience is most helpful.
- Include a broad range of participants who know enough to come up with new ideas and are outspoken enough to do so and defend their views.
- Put the most senior and knowledgeable people on “red” (adversaries) teams, so that you have the best chance of seeing the adversary played aggressively and well.
- Each team should have one “troublemaker” who will challenge everyone in the game with unconventional approaches.
- Avoid having a game with only senior decision-makers, especially if they share a point of view, again to ensure that alternative ways of looking at the situation get considered.
- Do not allow any idea to be dismissed or suppressed without strong evidence that it really wouldn’t work.
- Concentrate on producing indicators and warnings that would let you know if a threat the game identified seems to be happening or about to happen.
- Go over the lessons learned, preferably revisiting some of the more vexing questions more fully in one or more subsequent games.
When to Do It
Herman strongly recommends that people know when to use wargaming and when not to use it. “We don’t want to go in with a pre-fit approach to a problem,” he explains. “If you know your problem well, wargaming doesn’t help. This is a technique you use when you really don’t know what might be going on. When we do wargaming, we do a lot of research in setting up the issues and the environment, and getting new ideas and viewpoints around the table. Usually a company’s managers have really good information about one competitor they are most concerned about, but not so good about competitors No. 3 and 4.”
What about people thinking they know more than they do? Herman replies, “There’s always that possibility! Something like the real estate bubble played out many times before. Lots of key people just weren’t challenging their preconceptions about what would happen. Wargames take unwritten assumptions, writes them down and then asks, ‘Is this really true?’ Doing that can avert a lot of bad situations.”
Clearly new multi-disciplinary, flexible approaches to improved decision-making are important and valued now, and there is opportunity here for those prepared to pursue it. Wargaming and similar approaches offer a good way to explore issues and raise new questions, but more quantitative support will likely be needed as decision-makers probe more deeply into what evidence supports the teams’ and game-runners’ assessments.
Such support will have to be offered, however, in ways immediately understandable and useful to the decision-makers. How analytical results are incorporated – or not – into decision-making, how intuition competes with analysis and how to improve decision-making to make better use of available information combine to constitute another field of study with great potential for analysts, among others, willing to acquire the additional knowledge and work with the other disciplines that are also essential.
Meanwhile, decision-makers can profit right now from making more use of wargaming, especially the low-tech variety. Just a facilitated discussion of “what might happen and what might be there that we haven’t thought about?” would often be well worth doing. Wargaming appears to be an important, valuable and growing part of business analytics.
Douglas A. Samuelson (email@example.com) is president and chief scientist of InfoLogix, Inc., a research and consulting company in Annandale, Va. He is a frequent contributor to Analytics and a contributing editor of OR/MS Today.
- Michael Bennett and Russell R. Vane, III, 2006, “Using Hypergames for Deception Planning and Counterdeception Analysis,” Defense Intelligence Journal, Vol. 15, No. 2, pp. 117-138.
- Peter Bennett, 1977, “Toward a Theory of Hypergames,” Omega, Vol. 5, pp. 749-751.
- Benjamin Gilad, 2009, “Business War Games,” The Career Press.
- Mark Herman, Mark Frost and Robert Kurz, 2009, “Wargaming for Leaders,” McGraw-Hill.
- Douglas A. Samuelson, 2006, “The Hyper-Hypergame: Issues in Evidence-Based Evaluation of Social Science,” Proceedings of the Capital Science Symposium, Washington Academy of Sciences.
- Douglas A. Samuelson, August 2008, “Understanding Organizational Anarchy: Agent-Based Models Help Explain Well-Known Dysfunctions,” OR/MS Today.
- Douglas A. Samuelson, 2009, “Playing for High Stakes: Wargamers, Cognitive Scientists Seek to Avoid ‘Strategic Surprise,’ ” OR/MS Today, December 2009.
- Russell R. Vane, III and Paul E. Lehner, 2002, ”Using Hypergames to Increase Planned Payoff and Reduce Risk,” Autonomous Agent and Multi-Agent Systems, Vol. 5, No. 4, pp. 365-380.