Computer Assisted Military Wargaming: The SWIFT Wargame Tool

The Standard Wargame Integration Facilitation Toolkit (SWIFT), an Office of the Secretary of Defense, Cost Assessment and Program Evaluation (OSD CAPE) product, provides a computer environment that supports Department of Defense (DoD) wargaming. SWIFT complements, but does not substitute for good wargaming practices. Deputy Secretary of Defense Robert Work has called for the reinvigoration of wargaming in the Department of Defense.1 Wargames require careful attention in their application and execution to maximize their utility. As in all walks of life, computer assistance in the areas of wargame design, visualization, adjudication, and analysis would be useful to facilitate DoD wargaming. Several tools are available to support commercial games, but are inadequate to support the full range of professional wargames within the DoD.

What is a wargame? Peter Perla defines wargaming “as a dynamic representation of conflict or competition in a synthetic environment, in which people make decisions and respond to the consequences of those decisions.”2 Wargames explore the decision process of the players and provide an immersive environment to think about the issues in question. Wargames results are often what the players take with them when they leave. In other cases, wargames are used to support a larger analytic process where the burden on data capture is more significant. The scale of wargames takes the form of small numbers of participants examining political-military issues in a seminar setting with limited adjudication. Other wargames include a large number of participants examining detailed military issues involving rigid, complex adjudication of combat results. The purpose of the game is akin to the learning objectives, such as new insights into a problem, further testing of a concept or hypothesis, or even for socialization of ideas and issues.

As with any method of inquiry, wargames have a number of inherent limitations. Wargames are rarely repeatable, may be resource intensive, and are difficult to design. Wargames with complicated rulesets or system games are often time-consuming to execute and record. Similarly, seminar games with human adjudication face both dynamic visualization and player move recordation challenges as these processes struggle to keep pace with the social interactions. Moreover, as the resolution and scale of the wargames increase, they are hard to record and even harder to analyze. Computer aids should assist in reducing these burdens in DoD games through providing benefits in visualization, recordation, adjudication, sharing, and collaboration.

Figure 1. Development Engines

What does it take to accomplish this? Thankfully, the private sector has been vanquishing the foe of computer inaccessibility. We need look no further than Android that allows incredible customization on our phones and tablets. In software and analytic domains, development environments have flourished providing order of magnitude advances in productivity for analysts and programmers. Cross-platform gaming engines are the norm in the commercial world shifting the burden from technology implementation to artistic expression. Even the much smaller commercial wargaming domain has led to several gaming engines used for playing games online or through email by providing a customizable digital game board and pieces (Figure 1). The OSD CAPE solution to this design inaccessibility problem is the Standard Wargame Integration Facilitation Toolkit (SWIFT). SWIFT is a software environment used to build, play, and analyze turn-based wargames conducted primarily for analytic purposes. SWIFT provides a toolkit to enable integration of visualization, wargame rules, human and computer-based adjudicators into a multi-player turn-based wargame to facilitate testing, execution, and analysis.

Figure 2. SWIFT Meta-Components as a Common Language

SWIFT supports professional turn-based games from several perspectives: sponsor, developer, player, adjudicator, and analyst. For the game sponsor, SWIFT offers resource efficiency by shifting software development dollars from the “medium” to the actual design or particular game-specific features. Additionally, the wargame is preserved for re-use, modification, presentation, and sharing (think SWIFT as Microsoft Excel for wargamers).

For the game designer and developer, SWIFT not only provides a computer medium for wargaming but it, like most computer environments, enforces a design clarity and common language that is always desired on gaming projects. Games are described and designed in terms of their meta-components: participants, actors, resources, actions, game spaces, turns, and adjudicators (See Figure 2 for complete list). All game meta-components have attributes that can be manipulated to suit the requirements of the game design. The time required to instantiate a game in SWIFT depends upon the game design. It takes days, not weeks to build a manual game that utilizes SWIFT’s visualization and recording capabilities. Days to weeks are required to instantiate semi-automated/fully-automated games depending upon level of complexity.

Figure 3. SWIFT Visualization

For the game player, SWIFT supports visualization and efficient game play. SWIFT is an intuitive, appealing mechanism to learn once and rely on that training to play multiple different types of games shifting the training to the game rules rather than the tool navigation. SWIFT overlays situational awareness, actors, and actions on any map/image resource to include Google Earth’s .kml files (See Figure 3). While players have an enhanced common operating picture (COP) that they can filter in any number of ways, game controllers can implement the fog of war and information hiding by limiting player perception. Although game design can rely upon intermediaries (“pucksters”) between players and the software, we found that players intuitively pick up the point and click interface and have frequently done away with the intermediaries within less than 15 minutes of play. Millennials tend to pick it up with ease.

Figure 4. SWIFT Adjudication Approaches

SWIFT’s general concept for the adjudication process is shown in Figure 4. For the game adjudicator, SWIFT captures the suggested outcomes of an unconstrained number of adjudicators (human or computer) and presents those outcomes for ground truth selection. SWIFT supports a wide variety of manual and automated adjudication types. It also permits the use of several different adjudicators for the same phenomenology and permits the White Cell to choose the most appropriate or a combination of the adjudicator results for the game turn. SWIFT supports the game analyst by providing a consistent, transparent data structure and a game engine to support stochastic analysis of model-adjudicated games. Its use of structured input/output data facilitates inductive analysis techniques. SWIFT has a playback and other after-action features that support post game analysis. SWIFT can import and export most data to/from Excel for analysis and game development. Ease of use was a key design consideration.

As previously mentioned, SWIFT supports a vast variety of turn-based games from very structured games with significant numbers of game pieces and rules to turn-based seminars where managing temporal, spatial, and behavioral complexity is a key element of game facilitation. Game play can be a series of sequential or simultaneous player moves depending upon game requirements. Multi-level games can be supported in a single instance of SWIFT as players play in different theaters and/or echelons. SWIFT tracks changes caused by each of the actions and adjudications allowing traceability and understanding during post-wargame reviews and analysis. Additionally, SWIFT games can be played in a local or distributed environment. Using SWIFT in a distributed design creates a dependency on the network supporting the game, but even in local games the quality of the computers, projectors, and room layout are relevant factors. Infrastructure issues should not be underestimated! Regardless, the successful application of the SWIFT environment to a wargame implementation depends upon the wargame design and the specific requirements of the computer medium. SWIFT is not intended to compensate for poor planning and there are many circumstances and designs where it may provide limited to negative value.

We have encountered several questions when discussing SWIFT:

• What are SWIFT’s technical characteristics? Is it easily available for use by a DoD organization? SWIFT is GOTS software written in the Java programing language. All data is stored via XML. SWIFT has been used at all levels of classification to support COCOM, Service, and OSD games.
• How long does it take to set-up a game? It depends … how complex is the game design? SWIFT can be instantiated to support simple visualization within several days. Games with complicated designs characterized by multiple adjudicators or extensive orders of battle can take weeks. A key question when considering automation is what requires automation and what can be left as manual processes. When the question being examined lends itself to modeling and simulation approaches rather than a wargame, use the technique or tool more suited to the problem. SWIFT was designed to interact with other tools, but not replace them. Wargames are best for examining very complex, wicked problems where the conditions change and the relationships among the elements are unclear. Difficult but highly structured problems are best examined using the appropriate modeling tools.
• How long does it take to run through a game turn? It depends… how complex is the game design? Learning a complicated game tends to far exceed the time it takes to learn how to move actors and make actions in SWIFT.
• How scalable is the tool? This question has a complicated answer. SWIFT can support many game objects (we have instantiated orders of battle up to several thousand units). Note that SWIFT tool has some optimization for speed but it can be a challenge..

SWIFT is a powerful tool for designing, executing, and analyzing wargames, but it is not the entire answer for supporting DoD wargames. SWIFT is a wargame support tool that allows DoD professionals to build aspects of their game into a computing environment without a software developer present. It is not a shrink-wrapped wargame, but an engine for wargaming. It is not a model, simulation or artificial intelligence application even though all three have been considered as sources of adjudication or computer-based opponents. SWIFT requires humans are still required to be in the loop. Secondly, it can’t capture data at the speed of thought and there is no voice-to-text. Many so-called wargames involve discussions (many at the same time in the same room). SWIFT or even a handwritten note-taker is not going to have the capacity to record every element of conversation. However, if you have an experienced facilitator, who pauses and emphasizes the key points, you should be able to record the key actions and results in SWIFT.

Third, SWIFT is not a map/GIS application such as Google Earth. SWIFT can import Google Earth data and look just like Google Earth but without the 3D. In fact, it adds the ability to move objects around while capturing their new location and path. It also can capture the player’s intent. Many groups are using Google Earth for their COP; however, they have to capture intent and move data in a spreadsheet, then combine everyone’s data and create a new COP in Google Earth. SWIFT integrates all these steps. Finally, and most importantly, SWIFT is not a substitute for a good game design. You can create a bad game design in SWIFT, just as you can write a bad book using a word processing application. SWIFT simply provides a tool for you to write the Shakespeare.