Yesterday, Tom Fisher (PAXsims and Imaginetics) and Matt Stevens (Lessons Learned Simulation and Training) spoke about their work on serious game for humanitarian training. If you missed it, the Georgetown University Wargaming Society has posted the video of the event to their YouTube channel.
Pete Pellegrino is a retired USN commander and former Naval Flight Officer, currently employed by Valiant Integrated Services supporting the US Naval War College’s War Gaming Department as lead for game design and adjudication and lecturing on game related topics for the department’s war gaming courses. In addition to his work at the college since 2004, Pete has also conducted business games for Fortune 500 companies and consulted for major toy and game companies.
Pete kindly provided PAXsims with permission to share this video. The views expressed are those of the author and do not represent the official policy or position of any agency, organization, employer or company.
The following was written for PAXsims by Dr. James Sterrett, Directorate of Simulation Education (DSE), U. S. Army University.
The Directorate of Simulation Education (DSE) at the Command and General Staff College (CGSC), U.S. Army University spent mid-March through early June 2020 to prepare for, and then to conduct or support, three elective courses online using commercial wargames. This article outlines our key lessons learned, and then discusses some details of what we did.
In total, the class events we ran totaled 10 different games, each running from 2.5 to 8 hours, each preceded by at least one 3 hour preparation session. In addition, many of these involved numerous internal trainup sessions with each game, plus many trial runs of many games to assess their suitability for use, or in testing VASSAL modules we built for some of these games. For around 9 weeks, from 30 March through 2 June, we averaged one 3-hour online wargame session a day, for testing, preparation, or classes.
We ran wargames for 3 different courses:
Bitter Woods for the Art of War Scholars Program (2x 4 hour classes)
Aftershock for the Defense Support to Civil Authorities elective (1x 3 hour class)
Eight games for History in Action, which we teach in collaboration with the Department of Military History. (8x 3 hour classes)
Top lesson 1: Online wargaming works, but it’s harder than live. Compared to running wargames live, it requires more manpower, time, effort, and technology from both students and faculty.
Top lesson 2: Success requires scaffolding. Don’t assume students are ready with their technology or that they understand the online engine. Plan for on-call tech support during every class. Plan to explicitly teach both the online engine, and the game itself in that engine.
This is the most surprising outcome to us. Several of us had prior experience with VASSAL and were not very fond of it; we are now converts. VASSAL proved to be simple, reliable, effective, and made lower demands on computing horsepower and networks – and it is free. In addition, it was an easier and more powerful tool to make new game modules for.
(Read the detailed section for a more nuanced view of some of the other options.)
Test your tools in online classroom settings before committing to them.
Our initial impressions of tools were frequently overturned after gaining more extensive experience with them in testing.
Ease of use beats flashy presentation.
The more you can minimize the friction of using the online game tool, the more effort you can put elsewhere. This is why VASSAL became, unexpectedly, our favorite application.
Running a wargame online needs more manpower than running the same game live.
Running wargames live, a skilled facilitator can sometimes run 2 or 3 games. Online, you must have one facilitator per game. When teaching the game, you must have one person doing the instruction while another monitors a chat window for questions and puts them to the instructor at appropriate moments.
In addition, we found we needed to have a separate person as dedicated on-call tech support, every time. Although a few classes did not turn out to need tech support, most did, and dedicated tech support meant that the game facilitators could keep the games running while the students with tech problems got helped.
Running a wargame online requires a higher level of skill across the facilitators than running the same games live in one room.
Running wargames live in one room, one person can be the expert whom the others can rapidly turn to for help. Running online, everyone is necessarily in separate rooms, and even with side-channel communications, the question and answer interchange is much slower. Each facilitator needs to be an expert.
Keeping the game moving is harder online due to the limited communications.
Live, you can see what students are doing. You usually know who is thinking, who is confused and needs help, who is done making a move. Online, you usually have no idea. Is the student silent because they are thinking? Confused and lost? Conferring with their partner? Done but forgot to announce it? Done, and announced it, but failed to activate their microphone or had some technical issue? When do you break in to ask, possibly breaking their concentration and creating more friction?
Everything takes longer online.
Your game is hostage to hardware issues beyond your control.
A bad internet day makes for a bad class day. Students come with widely varying degrees of computer savvy. They also come with widely varying quality of equipment. We had one student whose computer was a low-powered laptop around a decade old, which created frequent technical issues. Another used a Surface tablet, which had no direct technical issues, but the small screen caused usability problems.
Ideally, each participant should have least 2 large monitors.
A reasonably modern computer, preferably with at least one large monitor, and, ideally, with two or more large monitors, definitely worked best. Multiple monitors enabled placing documentation and chat windows on one screen while placing the main game map on the other.
Those with only one monitor, especially if on a small screen, found themselves constantly paging between windows and struggling to manage limited screen space.
Some students and faculty took to using a high definition TV as a second monitor, which worked well.
Technology in More Detail
Ideally, we would have done extensive R&D into both a wargame engine and into a communications solution. However, we rapidly determined that Blackboard, which the Army already had on contract, provided a communications system that was both sufficient for our purposes and that students already knew how to use. While not perfect (the interface for splitting students into small groups can be a pain to use), Blackboard worked well for us. Specific features we came to rely on:
The ability to break students into breakout groups, and to have instructors move easily between breakout groups. Each breakout group was one game. Also, we could easily recall all the breakout groups into one room when it came time to return to group discussion.
Screen sharing to assist in teaching the games. While the shared screens were sometimes very fuzzy (which we worked around by zooming in when details were important), the shared screen allowed us to direct people’s attention to the item currently under discussion. In a perfect world, the game engine itself would provide a means of directing attention.
Multiple chat lines: Direct 1 to 1 chat, alongside breakout room chat, alongside group discussion chat, all at the same time. The major feature we wanted, and did not have, was a direct chat line between any subset of people without creating a new breakout room – so that 3 or 4 people on the same side could coordinate their strategy and tactics, for example. We worked around this by having students use their cell phones.
We spent several weeks testing online game engines, both for running games and our ability to modify or create new games.
As noted above, several of us had prior experience with VASSAL and did not have a high opinion of it. However, those opinions were based on the state of VASSAL in the later 1990s, when it was relatively new. VASSAL has improved a lot in the last 20 years, and those improvements are a great credit to its volunteer coding team.
VASSAL is not the prettiest or slickest engine out there. However, it had several decisive advantages:
Highly reliable, it worked on all the equipment students brought into the classes.
Free, while every other solution required either the instructors, or everyone, to buy software.
Easier for students to learn than other systems.
It was significantly easier for our team to make new or modified modules in VASSAL than in other systems.
Presented the widest variety of ready-to-go games relevant to our courses.
Because it is built from the ground up to support wargames, VASSAL’s standard interaction set is tailored to supporting wargames. The other engines seemed, to us, to have standard interactions best suited to running Euro games or role-playing games (which those other engines chase because those are much larger markets!)
VASSAL doesn’t enforce the rules. We thought this would be a weakness, but when the computer enforces the rules, it prevents the facilitator from fixing mistakes – and with first-time players, it’s very handy to let the facilitator see and do anything they want.
Two key workarounds we used with VASSAL:
Normally only one player can join a specific role. However, if everyone who is going to join that role does so simultaneously, you can pack many players into one role, permitting a small team of students to play the same side while maintaining fog of war. Note that this feature is not officially supported.
Most modules that had fog of war also included a “Solo” player who could see everything, so we used this as a facilitator role. We modified the Triumph & Tragedy module to include this as well. Without the ability to see through the fog of war, the facilitator cannot effectively answer questions and solve problems.
Tabletopia was our initial favorite, with a slick interface and great presentation. Our favorite feature is the ability to see the “hands” of the other players, which makes it really easy to direct attention – “Look at the Blue Hand”. Tabletopia is browser-driven and thus is platform independent, which is a great plus. It is also the only way to play 1944: Race to the Rhine online, which we very much wanted to include in our history course.
However, Tabletopia also had some problems. Running a multiplayer game requires that at least one player has a paid account ($9.99/month), and the Terms of Service for game creation included language that we were wary of. In testing, it was much more difficult to make a new game in Tabletopia than in VASSAL, and essentially impossible to modify an existing game we had not made. We could not figure out how to enforce fog of war in a blocks game in Tabletopia.
The great surprise came when we used it in class. We expected students would find the interface simple. However, students found Tabletopia confusing to use and said they preferred VASSAL. Students with weaker computer hardware or slower internet connections found Tabletopia crashed or refused to start.
While we may use Tabletopia again in order to use the excellent Race to the Rhine, we also know we need to figure out how to work through its issues first.
Tabletop Simulator (TTS) has a very large following, but we wound up bouncing off it. The large number of possible interactions means it also has a large number of controls and possible customizations. We found it confusing, and the physics model got in the way of ease of use as pieces bumped into each other. A friend who likes it admitted it takes at least 10 hours to get comfortable with TTS, which is longer than we can afford to spend for classes. In addition to these issues, TTS is a $20 purchase.
Roll20 is built to support role-playing games. Unlike the other options mentioned here, Roll20 includes fairly robust voice and chat communications. It’s reasonably simple to set up a new game in Roll20 as well.
Roll20 fared well in initial testing, and thus became a strong candidate for running Matrix games. However, in full testing, its communications fell apart under the load of around a dozen people. In addition, we ran into significant issues with allocating permissions to move pieces; as far as we could tell, players needed to join so they were known to the game room, then leave, so the GM could make permissions changes, then rejoin, which seemed like an overly complex dance to go through under time pressure in a class with students.
We suspect that our inexperience with the tool is key in some of these problems and intend to retest Roll20 in the summer. In addition, we know of others who have used Microsoft Teams and Google Sheets to run Matrix games.
No Computer Games – Why?
We avoided computer games for several reasons:
Students would need to buy them, and potentially need to buy many games for one class.
Many games of interest run on only a subset of student computers (only Windows, or only high-end Windows computers, for example).
Each computer game has its own interface to learn, on top of learning the game system, increasing the training overhead needed to get to the learning for the class; this is particularly an issue for our history class.
In many cases, understanding the games’ models is an essential component to learning the wider lessons of the class. In our experience, this is harder to do with computer games, whose models are obscured in comparison to manual games. (This is the price paid for the computer doing the heavy lifting of the model; the payoff of the computer is that it does that work.)
We are not adamantly opposed to computer wargames; we use them in our Simulations Lab during live instruction, and are investigating using them in some courses this fall in DL. However, in the short timeframe we had, the above complications were sufficient to rule them out.
Teaching the Games
In all cases, we learned that it works best to:
Provide a 15 minute introduction to the game at the end of the prior class. Students won’t learn the game from this but the overview helps them learn better from the rules and videos in step 2.
Provide the rules and tutorials as homework. YouTube tutorials were very popular with students, when they existed. Students will not learn the game from these but they will come armed to steps 3 and 4 with a better framework.
Provide a practice session. We routinely ran a practice session the afternoon before class. These lasted 3 hours (the same duration as the class) and included the full teaching script plus playing the game. We warned students that this was partly internal trainup, so they knew to be patient with periodic digressions as we worked out unexpected wrinkles. Because they actually play the game, students learn the game in these. If you control the groups, distribute the students who came to the Practice session across the class day student groups. As time went on, we learned to have internal trainup sessions before the official Practice session, so that our people were ready to run a game on their own in the Practice session.
Teach the game at the beginning of class. We find it always helps to begin by identifying the sides and their victory conditions, because you can tie all the game mechanics in the game back to them.
We establish up front that we will not teach all the details of the game, and thus many of these will pop up as they become relevant. We try to warn people if they are going to hit a special case, and if somebody winds up in a bad position because of a rule not previously explained, we will try to come to a reasonably fair adjustment so they are not unfairly punished by an unknown rule.
Doing all this requires facilitators who are experts on the game, as noted earlier.
We find that putting students into pairs on a given side works well in most cases. Two will tend to plan together, each can compensate for the places where the other finds things confusing, and provide moral support where one sometimes feels confused and alone. Three on a team, however, sometimes means one gets left out.
Teaching the Courses
Bitter Woods for the Art of War Scholars Program
The Art of War Scholars Program is a highly select group of CGSC students who engage in a wider-ranging and academically more rigorous course of study, focused on studying the art of warfighting through a combination of seminars and research focused on the operational and strategic military history of the past century. Each student must write a thesis in the CGSC Master’s of Military Art and Science program.
Dr. Dean A. Nowowiejski, the instructor for the Art of War Scholars Program, wanted the wargame to do three things: introduce the students to wargaming, introduce the terrain of the Battle of the Bulge to students for a follow-on virtual staff ride, and to examine the dilemmas facing the Allied forces in reducing the Bulge.
To support this, we need a game simple enough for new wargamers to play effectively, that covered the Bulge in enough detail to gain an appreciation for the terrain and forces involved, and that could be made to start later in the battle in order to cover the reduction of the Bulge.
We selected Bitter Woods for having the best balance of both a simple system (using only the basic rules) and the ability to run the Battle of the Bulge into January 1945. The runners-up were GMT’s Ardennes ‘44 and MMP’s Ardennes. Ardennes ’44 is more complex and Ardennes is out of print, the latter being a key criterion when we made the selection in January 2020 and expected to run the event live.
In order to highlight the dilemmas in reducing the Bulge, we created a scenario that began on 27 December 1944, and also modified the existing Bitter Woods 22 December ’44 start point to cover the entire map, both accomplished with assistance from LTC William Nance, PhD, of the CGSC Department of Military History. After testing both of these, we concluded that the dilemmas showed up best on 22 December, as Patton’s forces begin to arrive. This start point also made a better set of dilemmas for the Germans, as their offensive is not out of steam on 22 December, leaving them with difficult choices about how to protect their flanks while aiming for victory. We divided the twelve students into three separate game groups that executed simultaneously. We had teams of 2 on each side in each game, and each team was split between a northern and a southern command.
Dr. Nowowiejski told us that the Art of War Scholars students would be prepared, and he proved correct. This group of top-flight students, all very comfortable with technology, had no technical issues. In addition, while we ran the game, LTC William Nance moved through the 3 game rooms, offering both historical commentary and acting as the high command for both sides to ping students with questions about their plans in order to ground those in the wider concerns of their historical counterparts. This left Dr. Nowowiejski free to circulate through the groups, observe the students, and discuss wider points with them.
Dr. Nowowiejski had students discuss their plans and operational assessments with the entire class at the end of each of the two 4 hour classes, for a mid-point and final AAR. As the students in the various Allied and German teams uncorked radically different plans, this provided a chance to compare possible courses of action and outcomes for both sides. Students did find they had more units than they could easily control, but this produced useful discussions on the difficulty of integrating tactics into operations. Overall, Dr. Nowowiejski judged the event “very successful” and hopes to have us run it, live or on VASSAL, next year.
Aftershock for the Homeland Security Planner’s Elective
We have run Aftershock in person several times in the past for Clay Easterling and Joseph Krebs’ Department of Joint & Multinational Operations Homeland Security Planner elective course. Much of the course examines higher level legal and policy issues. Playing Aftershock in the middle breaks this up, and also serves as a reminder of the practical impact of the plans and policies they are discussing. Students regularly name it their favorite part of the course. Now we needed to run it electronically…!
No computer version of Aftershock existed. The designers, Dr. Rex Brynen and Thomas Fisher, readily granted us permission to create a version in VASSAL, and Curt Pangracs of DSE spent around two weeks creating and testing the module in time for the course.
There were 33 students in this elective, divided into pairs for each of the 4 teams in the game, making a total of 4 games run in parallel. Four of us from DSE ran the games, while a fifth stood by for technical support, ensuring the two instructors could circulate between the three sessions to observe and discuss.
We knew that this course tended to have a solid proportion of officers with low levels of experience with computers. Because of this, we set the Aftershock module up with two participant roles: the Facilitator, who controlled everything on the board; and the Observers, who could not change anything, but could see everything and call up the supporting documentation. This matched the way we often run the game in person, where the facilitator can keep the game moving by running the board and presenting the players with the next decision. We figured that with some of the students being less technical, making the students Observers would allow them to concentrate on making decisions instead of trying to puzzle out how to make the game execute their intended course of action.
We had far more technical issues than we expected, possibly because the larger number of students – nearly three times the number in any of our other groups – meant there were more opportunities for problems. As a result, in each of the four games, the facilitators wound up using the backup plan of streaming their VASSAL screen of Aftershock out to some of the students who could not otherwise see the VASSAL screen. This is far from ideal, as those students reliant on the stream could not control the view, and the Blackboard shared screen is often fuzzy, but it was better than not seeing the screen at all.
Despite the technical issues, students found the exercise very useful, and the instructors named it “a highlight of the course”. As one student wrote in their AAR, “Finally a time at CGSC where we are truly talking with one another to get something done and seeing the results of our decision”.
However, a key lesson here is that the event would have gone a lot more smoothly if we had conducted a readiness check at the end of the prior class session, just to make sure that everybody had VASSAL installed, could load the Aftershock module, and could join the online session – and then to help those who could not, so their troubles were fixed before the main event.
History in Action is a joint elective taught by DSE and the Department of Military History, run with the aim of teaching military history through wargaming, and also teaching a better understanding of wargaming through learning the history. Knowledge of history should inform both playing and assessing the game. Equally, playing the game should help better understand the history; while wargaming can’t let you walk a mile in someone’s shoes, it can let you walk ten feet in their socks. In prior years, DSE’s partner in this class was Dr. Greg Hospodor, but he moved away and we now partner with Dr. Jonathan Abel, and were also assisted by LTC William Nance, PhD, when he was available.
To be selected for this course, a game has to pass all of these tests:
It has to be a good game – fun is the hook, though it isn’t the point.
It has to be available for our use (some that pass the other criteria are out of print, or, for online, have no online implementation).
We have to be able to teach and run it within the 3 hour class time while leaving time for discussion.
It must be dripping with history. It has to highlight unique aspects of the historical event it covers, so it both helps teach that history directly, and further helps teach when compared to the other games in the course. This tends to rule out many less complex games because they wind up being functionally generic. For example, if the game system doesn’t help drive home the difference between commanding World War 2 armor divisions and Napoleonic cavalry divisions, or treats the employment of the Roman manipular legion as little different from that of the Macedonian phalanx, then it doesn’t drive the learning we are looking for.
While in past years we tried to sequence the games according to a theme or timeline or the scale of the actions, our test sessions in early April convinced us that we should sequence the games in order of probable complexity to students. While we began with the list of games we use when teaching the course live, but some of them were not available online, while others we would like to use were. We used, in order:
Battle for Moscow (The 1941 drive on Moscow)
Napoleon 1806 (The Jena/Auerstadt campaign)
1944: Race to the Rhine (The Allied drive across France, with a logistics focus)
Drive on Paris (Schlieffen Plan and Plan XVII in 1914)
Strike of the Eagle (1920 Soviet-Polish War)
Triumph & Tragedy (The struggle for Europe, 1936-1945)
Fire in the Lake (Vietnam War)
Nevsky (Teutonic Knights vs Novogord Rus in 1240-1242)
In each 3 hour class, we began by teaching the game, then we ran it in parallel student groups until there were 45 minutes remaining. The next 30 minutes or so were spent in discussions, and the final 15 minutes or so were spent introducing the next game in the class. Between classes, students were assigned material on the history behind the next game, rulebooks and tutorials to learn the next game, and a graded AAR sheet to fill out on the game just played. The AAR sheet asks for paragraph-length answers to these questions:
What was your plan/COA going into the game?
How did your plan/COA work?
How did the game illustrate the specific contextual elements of the period?
Was the game effective in conveying these contextual elements? How or how not?
What did you learn about warfare in the game’s time period? What surprised you?
What specific lessons can you draw from this game to apply to the future?
We were very pleased with student learning in the class. Student AAR papers were full of observations on things they had learned about history, about wargaming, and that they could carry forward to future assignments. As one student wrote in their end-of-course feedback, “more than anything the course provided context and examples that I can use in the future when explaining the challenges at the operational level of warfare”. Success! However, we did have to overcome various issues along the way.
We intentionally began with Battle for Moscow, the simplest game, to ensure we could also teach VASSAL in the same class. This generally paid off, as subsequent games utilized, at most, a few more features of VASSAL each time, and thus the learning curve was well controlled and students seemed comfortable with VASSAL most of the time. This process worked poorly when we jumped to Tabletopia for Race to the Rhine in class session 3, and then back to VASSAL for session 4 and beyond. Some of our issues with Tabletopia likely stem from our assumption that its interface was easy enough to need little direct training, and to the ways in which it is different from VASSAL. Equally, we had a slight uptick in trouble with the VASSAL interface in class session 4, perhaps because the students had been out of touch with it for a time.
We began inviting students to our internal prep sessions once we realized they might be able to attend. Students who had the time to attend these were normally much better versed in the game than their peers. We, in turn, had to recall that those unable to attend the optional prep session should be assumed to have a good reason! We also learned to spread the students who attended the prep sessions across the student groups. Arranging the student teams ahead of time, and publishing them for students, also helped, as some student teams would strategize ahead of the class.
This course charted the middle ground in the level of technical issues. All the students were comfortable with technology, but some had poor internet connections or weak computers, including the roughly ten year old laptop mentioned earlier. This led to those students losing connection to VASSAL or Blackboard. When using Tabletopia, weaker internet connections and weaker computers completely failed. Just as we all have learned that internet meetings go better when everybody turns off their video feed, opting for systems, such as VASSAL, that made less intensive use of network and computing power proved better in practice.
Online wargaming works, but it is more effort than live, because:
Test your technology thoroughly and ensure you have support on hand to run it.
Running wargames online will require a higher level of expertise from all of your facilitators, of technology and the games.
Running wargames online will require more preparation from students, both in learning the game and ensuring their technology is ready.
BoardGameGeek (description) and VASSAL (module) links for all the games mentioned:
William Owen recently offered some thoughts at PAXsims on “what is wrong with professional wargaming.” Jeremy Sepinsky (Lead Wargame Designer at CNA) then replied with some comments—which I have reposted below for greater visibility. The views expressed are those of the author and do not represent the official policy or position of any agency, organization, employer or company.
I think the challenge here comes with equating “wargaming” with an analytic discipline rather than an analytic tool. Wargaming looks completely different in various context, but criticizing the rigor of the discipline is like criticizing the p-test, the Fourier transform, an MRI, or anonymous surveys: there very valuable if done well, and damaging if done poorly. The trick is in educating sponsors and potential sponsors as to what “bad” looks like. Even peer-reviewed journals have gone for years without identifying the “p-hacking” that has been taking place in quantitative analysis. And wargaming is a lot more diverse a toolset with a smaller number of skilled practitioners (and no peer-reviewed journals, as you point out) than quantitative methods, which makes it even harder to call out the bad actors.
To respond to Owen’s question of “so what, is it obvious?”: When a person running a professional wargame cannot effectively translate real-world decision making into relevant impacts in the conduct of the game, then either a) the person needs to be able to fully articulate why decisions at that level are beyond the scope of the mechanics, or b) it is a poorly run wargame. But many of the situations he discusses are “game-time” decisions. And it would be impossible/impractical (though probably beneficial) to include Matt Caffrey’s “Grey team” concept in all games. In that concept, there is an entire cell whose job it is to evaluate the wargame itself. Not the outcomes, or the research, but instead to critique whether the wargame was an appropriate model of reality for the purpose defined. Though, to support the other points in Owen’s article, I have not been able to find any published article discussing the concept.
But this leads into another point: wargames are more than combat modeling. Many of Owen’s examples and statements about the model seem to imply that the wargames he discuses are those that are interested in modelling and evaluating force-on-force conflict—and that the side that understands the underlying wargame mechanics of the conflict will succeed. To that end, those games do not seem to be played manually for just the very reason that you’re discussing. However, they are instead reproduced as “campaign analysis“. Models like STORM and JICM are trusted, I would argue, overly much. It takes away the requirement for the player knowing the rules, because it pits computer v. computer where both sides know all the rules.
When a given conflict can be reduced to pure combat, campaign analytics are a good tool for calculation. But when conflict is more than combat, the human element comes to the fore and wargames have an opportunity to expose new insights. In these cases, the specifics of the combat models should play less of a role in the outcomes. They are more highly abstracted to allow time and attention of the more humanistic elements of war: the move-counter-move in the cognitive domain of the players. Wargames structured properly to emphasize that cognitive domain should overcome the requirement of memorizing volumes of highly detailed rules by simply not having that many rules. Players only have so much mental currency to spend during the play of a single game, and where that currency is placed should be chosen (by the designer) wisely.
Finally, I’ll concluded with a response to Owen’s final statement: “The right wargame applied in the right way clearly does have immense value. It merely suggests we need to get better at understanding what has value and what doesn’t.” Who is it that defines the value of the wargame? Is it the sponsor? The designer? The players? I guarantee you that each come out with some value, and that they all may not agree on what that value was. Most US Department of Defense wargames that I am familiar with are one-off events. Understanding the implications of each wargame rule on every wargame action or decision is beyond the scope of most wargames and beyond the interest of wargame sponsors. Instead, we wargamers can do a better job explaining the limits of our knowledge. When we design a game, there is a delicate balance between fidelity and abstraction. Some aspect of the game are highly faithful to reality, while others are highly abstract. Where you place the fidelity and what you abstract has a tremendous outcome on the conclusions that you can make at the end of a wargame. Wargame designers, facilitators, and analysts owe it to their sponsors to make it clear what insights and conclusions are backed by a high degree of fidelity and which are not. Complex wargame models always run the risk of inputs being identified as insights, and our due diligence is important here. But that diligence extends beyond the numerical combat modelling into the facilitation, scenario, and non-kinetic aspects of the wargame as well.
The following piece was written by William F. Owen, editor of the Infinity Journal. In addition to this he consults for industry, government agencies and armed forces on a range of command, doctrine and capability issues. He started wargaming as a teenager.
Professional wargaming should aim to provide insights that can inform decisions, based on a degree of evidence. In essence professional Wargaming should equal the test of theory, in that it should explain extant phenomena and enable a degree of prediction. Both the phenomena and prediction would have expression as insights or issues requiring further investigation. If professional Wargaming cannot do this, what use is it.
In other words professional wargaming should tell you why, when, where and how combat occurs, thus give the practitioner a sense of what would occur in reality. Well-executed professional Wargaming of the right type has immense value, though the actual empirical basis for this, while extant, may not be as comprehensive or as rigorous as popularly imagined and there is almost no body of peer reviewed body of unclassified academic research. Most importantly, the problem is that what makes a good wargame seems to be poorly understood, particularly by many who advocate Wargaming professionally. This paper takes the view that the best insights are derived from multiple iterations of truly adversarial wargames, using a number of different valid models and methods.
A very small number of books written by an equally small number of professional wargamers and/or analysts do exist, but little of it seems concerned with the validity of wargaming as a professional tool as concerns how different models produce differing insights. Books about how to wargame and the history of Wargaming does not a body of academic or professional literature make!
The consequences for being wrong or using a bad wargame are both expensive and extremely serious. Firstly people can die based on bad advice/practice emanating from bad wargames and secondly it seems logical to suggest that poor choices based on wargame evidence can easily waste as much money as Wargaming might supposedly try to save. Wargaming may not always be a cost saving measure, if done badly. Wargaming can be used to give the appearance of validity to bad and very expensive ideas. It needs to be explicitly stated that Wargaming can be both very good and very bad. The problem is no one ever seems to talk about the bad or very bad, and how closely the professional community seems to flirt with it or even knowingly ignores such an issue. Thus evidence derived from Wargames is extremely unsafe unless the modelling and processes used have been subjected to a high degree of rigour.
The heart of a professional wargame relies on the consequence of decisions as explained by a model. That model should be a useful approximation of reality. It doesn’t have to be highly detailed or even complex, but it must be able to produce outcomes that are by and large valid in the real world. This is what separate hobby wargamers from professional wargamers. Professionals need it to make sense, or someone gets hurt.
Given the centrality of the model, what drives that model is clearly critical, yet there seems to be very little, if any operational, rigorous or academic literature based on the validity of the models apparently professional wargames employ. Indeed even professional models may well pander to popular perceptions of outcomes as the mechanics are often modified from hobby games. For example the idea that infantry derive in an increase in effectiveness if defending in wooded terrain is highly context specific, so not the absolute given most models assume. There is a body of operational/historical analysis literature that suggests infantry attempting to defend within wooded terrain usually loose and loose badly. That may well mean that some professional wargames rely on very poor models and thus produce unsafe insights.
This may seem contentious but I would like to propose some simple observations that may strike to the heart of professional wargaming.
For example, the participant or participants in a wargame that best understand the rules and how the model or models work have a disproportionate advantage. What his means is that a highly experienced and capable military commander will probably lose when playing a wargame against a civilian who happens to be a more experienced war gamer, all things being equal. This would be because the military man will fight and operate as per his real life understanding and the teenager will merely do what he knows works in the game. So for a military user of wargames the more he is exposed to the game model of combat the more comfortable and able he will become in terms of its employment. If that model his not strongly based on reality, then he will be learning all the wrong lessons and that will or could have real life consequences. The same applies to wargames used for military education, force development and/or doctrine development.
This extends across all wargames not just the professional domain. The Dungeons and Dragons players who are experts in the rules, books and the combat resolution model should and most probably do make far better decisions than people entirely new to the game who may actually be better decision makers, but lack the knowledge to inform them. The immense challenge produced by computer game AI models is not usually a product of complex tactical algorithms. Most computer game AI is tactically simplistic, but if it plays exactly by the rules, which it has to, it will simply compensate to an incredible degree for its lack of tactical acumen compared to the human player, who is unfamiliar with the detail of how the model works.
If you want to excel at any wargame, play it a lot, learn, test and investigate how the rules predict the outcomes of engagements. You will leverage yourself a measurable advantage over someone who knows less. However this will not make you a skilled commander in the real world or give you operational insights that are safe on which to base expertise or professional judgement. You will merely become an expert war gamer.
However given real world experience the same would apply in the real world. The commander who knows most about his force, in terms of how it does what it does and it strengths and limitations will be the man best able to employ that force in combat. Knowledge of your own force is literally a combat multiplier.
OK, so what? Is this not all obvious?
If it were obvious where is the discussion? Who has addressed these issues in an open forum? Why is there not more written on bad wargames and why is professional Wargaming so variable in terms of output? It seems unlikely that research agencies and armed forces that have had bad experiences with wargaming experimentation would be open about what when wrong, even if they were prepared to admit the error; which strongly suggests this subject is avoided even no classification issues exist.
Almost all wargame literature unquestioningly champions and advocates wargaming for the sake of wargaming with almost no professional rigour. The validity of the model and the rule set is simply terra incognita to the vast majority of wargamers as well as a lot who employ it professionally. The reason why many military professionals have been historically and contemporarily dismissive or agnostic to wargaming is that they simply don’t trust the models used. It would seem logical to suggest that if they thought that the combat modelling was accurate they would engage more than they do with the process. Why would they not?
So there are actually two distinct but closely related problems here. The first is that it is entirely right to be sceptical as to the validity of most wargame models or rule sets. The second is that a high level of familiarity with any rule set, in any game confers an advantage which will not translate into safe insights or professional development, unless the model used can be shown to have a high degree of real world validity.
The issue is thus the models and the rule sets. The view that all wargames have a degree of professional merit is toxic to the validity of Wargaming as both a tool for professional military development or indeed any practical military application. It is entirely valid to note that a model is an approximation of reality, not an exact replication of it. Thus the problems occur when those approximations generate false lessons that would not aid understanding or experience in the real world. That said, real world combat is so infinitely variable and subject to friction that any model will struggle yet the very nature of Wargaming seeks to address this specific issue as in to model warfare. By using models were axiomatically accept both their utility but also their limitations. Wargaming is far more the dim candle that lights the path, than the night vision goggles it is often advertised as.
So the challenge offered to those advocating the professional application of Wargaming is, why should any professional have any faith in the validity of your modelling and does their having deep knowledge of your model gain them an advantage that would not be present in the real world? If playing the wargame does not make them better at their job in reality, what is the use of doing it? To be deliberately contentious, if the good civilian wargamers, experienced in Wargaming alone, can beat experienced military commanders what does that tell you? What would that suggest?
Now the premise of this paper fully concedes that Wargaming can be and has been shown to be an extremely valuable tool, but there needs to be an evidence-based understanding of why and how we know that. For example, why would anyone use a hex-turn-based wargame instead of 1/285th scale micro-amour to address a particular point of force structure design? If the answer doesn’t lie in the validity of the model, but in the human organisational, time, budget and playability needs of the organisation conducting the work, then there maybe something very wrong. Likewise how safe are insights generated by one methodology, if they do not concur with the insights generated by a different method, approach or model, especially when examining the same problem?
One interesting aspect of comparing the validity of wargame models via comparison is both the suggestion and assertion that manual as opposed to computer based Wargaming allows for a greater understanding, and visibility of the underlying model. The counter argument is that limited time and budget means that only relatively simple manual models can be used because they simply cannot process or account for the wide variety of parameters inherent to most computer models. Manual games are forced to use inherently simple models. It seems to be a very reasonable conjecture that computer based game are actually played in a very different way to manual ones, to the extent that given broadly the same problem different behaviour and decision making would be required dependant on whether you were playing a computer based game or a manual one. This is largely to do with the number of parameters the model can process. Hex-based computer wargames actually allow this is to be investigated, as do computer games based on Dungeons and Dragons Rule Sets. If the behaviours, thus outcomes are different then this phenomena lies within the model. The issue is not computer versus manual. The issue is the veracity, thus usefulness of the model. Computer models can actually be investigated to a very high degree, although they cannot often be altered. Given simple tools like scenario editors you can investigate the behaviour of combat resolution models and/or the AI underlying the adversary decision-making or AI behaviours. Games that allow third party mods can allow even deeper levels of investigation and understanding. It could be suggested that agencies and organisations that employ Wargaming are/should be well aware his, though perhaps reluctant to engage in a conversation about. If this isn’t the case then it seems fair to ask why after 15-20 years of such games, does this condition persist? The use of computer simulations for operational analysis is a well-trodden path in defence circles.
To conclude, there seems to be little informed discussion or scientific and academically rigorously writing on what makes a good or bad wargame fit for professional use. In fact there seems to be little beyond opinion and faith based assertions that x or y models are valid and safe to employ and that professional wargames are of value regardless of the model. This is not to say professional Wargaming has no value. The right wargame applied in the right way clearly does have immense value. It merely suggests we need to get better at understanding what has value and what doesn’t.
It’s out! Ellie Bartel’s long-awaited PhD dissertation on Building better games for national security policy analysis is now available on the RAND website.
This dissertation proposes an approach to game design grounded in logics of inquiry from the social sciences. National security gaming practitioners and sponsors have long been concerned that the quality of games and sponsors’ ability to leverage them effectively to shape decision making is highly uneven. This research leverages literature reviews, semi-structured interviews, and archival research to develop a framework that describes ideal types of games based on the type of information they generate. This framework offers a link between existing treatments of philosophy of science and the types of tradeoffs that a designer is likely to make under each type of game. While such an approach only constitutes necessary, but not sufficient, conditions for games to inform research and policy analysis, this work aims to offer pragmatic advice to designers, sponsors and consumers about how design choices can impact what is learned from a game.
Table of Contents
Introduction: Games for National Security Policy Analysis and How to Improve Them
Towards a Social Science of Policy Games
Four Archetypes of Games to Support National Security Policy Analysis
Designing Games for System Exploration
Designing Games for Alternative Conditions
Designing Games for Innovation
Designing Games for Evaluation
Trends in RAND Corporation National Security Policy Analysis Gaming: 1948 to 2019
Conclusions, Policy Recommendations, and Next Steps
Appendix ASample Template for Documenting Game Designs
Tim Price has been kind enough to pass on this report from a recent play of the Flattening the Curve matrix game.
Last night I managed to get 11 volunteers together to play a distributed version of the Flattening the Curve matrix game over Zoom. It was an interesting and frustrating experience, but I thought it might be worthwhile sharing it with you.
We used Zoom for the video chat. We felt it was very important to be able to speak and see each other and Zoom has a simple and intuitive mosaic screen setup that is particularly useful for the Facilitator. The surround to the image is highlighted to show the current speaker, interrupters are shown with a highlighted line under them, and their names appear under their faces (really very useful indeed). Of particular interest for running a Matrix Game, it is possible to sent private messages to named individuals using the chat function in the application. It was also stable for the 3hrs we played.
We used Google Slides for the game map (see here). With the map itself as the background image and a number of counters imported as images onto the map (and left outside the slide boundary), so everyone could see and collaboratively move the counters if necessary. It is useful to duplicate the last slide for every turn, so you have a record of the map after each turn, and that also allows a run through at the end as an After Action Review.
Finally, we used Mentimeter to be able to carry out the “Estimative Probability” method of adjudication.
When using Estimative Probability players or teams are asked to assess the chances of success of an argument, and these are aggregated to reveal the “Crowd Sourced” chance of success. In analytical games, this provides potentially valuable insight into how participants rate the chances of a particular course of action. Following discussion, players select the option on the Mentimeter slide which, in their view, best represents the probability of the argument’s success. These are displayed immediately to the Facilitator, but not to the players, so it is using hidden voting. It is generally felt that this is a more accurate method to leverage the work on Crowd Sourcing, as well as making the resulting probability more accessible and acceptable to the participants. The terms on the slide also reflected those commonly used in the intelligence community.
The advantage with Mentimeter over other poll and voting systems is that it is free, feedback is instant, and you can use a single slide for all the Matrix Arguments, because you can re-set the results each time. Of course, if you want to have a record of the results, you will have to buy the upgraded version, or save a screenshot each turn (which is a pain).
Running the Game
As is normally the case with video conferences, we had the usual difficulties getting everyone onto the Zoom, with sensible names displayed instead of “Owner’s iPad”, so the start was a little delayed. I had put out a Loom video with a short introduction about Matrix Games, but inevitably a few of the players hadn’t been able to view it, so we were delayed starting as I had to explain how the game would play.
As the game went on, I modified the map (based on some helpful collaboration with TNO in the Netherlands), to make it easier to follow. The revised map is here:
The game played perfectly well, but at a slower pace that if it had been face to face, and it was certainly more tiring for me as the Facilitator. The inter-turn negotiation between team members and other teams was carried out using Whatsapp: and Whatsapp Web so was private to the other players.
We were time limited and were only able to have 11 participants in the end – but it was mainly a trial to see if running a Matrix Game remotely is at all possible. We got a few insights from the game, one of which I will share – as we all go into working from home full-time and are switching to remote working, we end up downloading all sorts of software and applications that we would never have normally dealt with. This increases the threat surface for cyber-attacks by an order of magnitude, so correct digital hygiene is going to be as important as washing your hands.
Following the game, we quickly did a couple of polls, hopefully better informed by the experience of the game:
Each participant was asked to give me their MOST IMPORTANT thing that would happen over the next month (please note the definition of “thing” was left deliberately vague so the players could decide for themselves what it meant).
They were then asked to vote on which of these was the MOST LIKELY thing to happen.
Next, each participant was asked to give me their MOST IMPORTANT long-term consequence of Coronavirus.
They were then asked to vote on which of these was the MOST LIKELY thing to happen.
It is possible to run a Matrix Game remotely, but it is very tiring for the Facilitator and takes much longer than you thought it would.
The right choice of technology can make a real difference – so mandated standards and corporate choices may well have an impact on the experience. This means that practicing, as I was, while waiting for the corporate roll out of their platform of choice might end up especially frustrating, when I am unable to do something that I know a free app on the internet will let me. But downloading all those free apps and trying them out could be dangerous, because the bad guys are definitely out to get you…
War has become increasingly digital, manifest in the development and deployment of new capabilities in cyber, uncrewed and remote systems, automation, robotics, sensors, communications, data collection and processing, and artificial intelligence. The wargames used to explore such technologies, however, have seen a renaissance of manual and analogue techniques. This article explores this apparent paradox, suggesting that analogue methods have often proven to be more flexible, creative, and responsive than their digital counterparts in addressing emerging modes of warfare.
Warfare has become increasingly digital. Militaries around the world are developing, deploying, and employing new capabilities in cyber, uncrewed and remote systems, automation, robotics, sensors, communications, data collection and processing, and even artificial intelligence. The wargames used by governments to explore such technologies, however, have seen a renaissance of manual and analogue techniques. What explains this apparent paradox?
This article will explore three reasons why analogue gaming techniques have proven useful for exploring digital war: timeliness, transparency, and creativity. It will then examine how the field of professional wargaming might develop in the years ahead. To contextualize all of that, however, it is useful to discuss wargaming itself. How and why militaries use games to understand the deadly business of warfare?
You can read the full thing at the link above. For more on the journal, see the Digital War website.
As readers of PAXsims will know, over the past few years we have run several full day emergency response megagames in Montreal and Ottawa: APOCALYPSE NORTH (simulating a zombie pandemic threat to Quebec and Ontario from south of the border) and ATLANTIC RIM (giant creatures attack Atlantic Canada):
None of these games was meant to be serious, of course—as the after action reports above make clear, we play them for fun. However, the underlying game models can certainly be modified for more serious purposes.
If you would like a copy of my ATLANTIC RIM Design Notes to inspire you in your own megagame design, I’m happy to send them to you in exchange for a donation of any amount to the World Health Organization COVID-19 Solidarity Response Fund. Just make a donation, then email me with the receipt to receive the design notes (pdf). I’m happy to provide tips on adapting the game approach for your needs too.
Please note that the Design Notes were not written for an external audience. Instead, this was our internal reference document. As a result, they do not include all game mechanics nor game materials (such as the maps, science quests, or hospital displays) you require to run a game. They probably still contain a few typos too! Still, at 51 pages long there is quite a bit there to inspire you.
For other inspiration, check out the Jim Wallman’s games at Stone Paper Scissors. The APOCALYPSE NORTH series were modifications of his original URBAN NIGHTMARE megagame, which he has since updated. His GREEN AND PLEASANT LAND national resilience megagame (which he ran at Connections UK 2018) is also very relevant.
We at PAXsims believe that serious games are a very useful tool in the analytical or educational toolbox—if we didn’t, we wouldn’t put so much effort into this website and all of our other game-related activities. However, I often find myself warning about the limits of games too. They aren’t magic bullets. In some cases, moreover, they’re not even an especially useful tools.
**Research Funding Opportunity** King’s has an important opportunity & responsibility to use our research expertise in support of the international response to the COVID-19 outbreak. This goes beyond efforts to find treatments. How can Wargaming help? Details attached. pic.twitter.com/RytFgvIGEf
As we collectively grapple with the unfolding global crisis, however, I thought it prudent to also highlight some the risks of serious pandemic gaming. As I will argue below, while serious games have a great deal of utility, they can also be counterproductive. We thus all have a moral responsibility to make sure (as they say in the humanitarian aid community) that we DO NO HARMwith our work.
First of all, there’s the modelling problem. We have to be very humble in assessing our ability to examine some issues when so little is known about key dynamics. Related to this is the “garbage in, garbage out” problem. Our data is often weak. The excellent epidemiological projections published by the Imperial College COVID-19 Response Team have been very useful in spurring states to action, but in the interests of avoiding confirmation bias we also need to recognize that some epidemiologists are raising concerns about the adequacy of the data used in such models. We need to make the robustness of our game assumptions to clear to clients and partners. Be humble, avoid hubris, make assumptions and models explicit, caveat findings, and don’t over-sell.
Second, playing games with subject matter experts (SMEs) can pull them away from doing other, more important things. I’ve done a lot of work on interagency coordination, where there is a similar problem: coordination meetings are great, but when you add up the time that goes into them they can actually weaken capacity if you aren’t careful. Of course, you can run games with non-SME’s, but then the GIGO problem is exacerbated.
Any gaming generally needs to be client-driven. Do the end-users of the game actually find it worthwhile? What questions do they want answered? This isn’t a universal rule—it may be that gaming alerts them to something that they hadn’t considered. But do keep in mind the demands on their time, institutional resources, and analytical capacity.
We also have to recognized that the much-maligned BOGSAT (“bunch of guys/gals sitting around a table”) is sometimes preferable to a game, when the former is run well. For a game to be worth designing and running it has to be demonstrably superior to other methods, and worth the time and effort put into it. There is a reason, after all, why the CIA’s Tradecraft Primer: Structured Analytic Techniques for Improving Intelligence Analysis warns that gaming techniques “usually require substantial commitments of analyst time and corporate resources.”
We need to debrief and analyze games carefully. The DIRE STRAITS experiment at Connections UK (2017) highlighted that the analytical conclusions from games are often far from self-evident, and that different people can walk away from the same game with very different conclusions.
Messaging for these games matter. The public is on edge. Some are dangerously complacent. Some are on the verge of panic. One wrong word, and suddenly there’s no toilet paper in the shops. If you don’t consider communication issues, reports from a game could feed either a “don’t worry it’s not that bad” view or a “my god we’re all going to die” response in the media and general public.
We need to be careful of both uncritical game evangelism and rent seeking—that is the “it would be cool to a game/games solve everything” over-enthusiasm, or “here’s a pot of money, let’s apply for it.”
In short, in a time of international crisis, we need to do this well if we do it. In my view it generally needs to respond to an identified need by those currently dealing with the crisis—or, if it doesn’t, there needs to be a good reason for that. They’re busy folks at the moment, after all.
UPDATE: I did a short presentation on this for the recent King’s Wargaming Network online symposium. My slides can be found here: DoNoHarm.
The following piece was written for PAXsims by Patrick Dresch. Patrick is based in Salisbury (UK), and is interested in the application of board games as training tools for emergency and disaster response. In 2019 he completed an MSc in crisis and disaster management at the University of Portsmouth, supported by a dissertation investigating the potential for cooperative board games to be used to train emergency responders in interoperability. He has also had the opportunity to test the integration of game mechanisms with table top and live simulation exercises by designing and delivering exercises as a volunteer with the humanitarian response charity Serve On.
I am a great believer in the potential for board games to be used as tools to supplement training and exercising for those working in emergency response and disaster relief. My interest in this field has mostly focused on using cooperative board games to practice interpersonal skills which can improve interoperability, including the potential to improve coordination and joint decision making. More recently, however, I have also been considering how this platform could be used to prompt emotional reactions which may be at odds with what might be called a rational solution.
In an abstract game it is often easy to focus on a game as a puzzle which needs to be solved. A player may have a personal aesthetic preference for the red tiles in Azul (2017), for instance, but this is unlikely to determine their strategy when playing the game. Other popular games use art and aesthetics to reinforce the theme of the game, and provide narrative structure to what could otherwise be an abstract puzzle. One example of this is the choice of illustrations on the adventure cards for The Lost Expedition (Osprey Games, 2017) (Figure 1).
Figure 1: Examples of adventure cards form The Lost Expedition.
Here we can see that the illustration choices not only reinforce the jungle survival theme, but also help players construct a narrative framework by showing dilemmas which work with the symbols and triggers. It should be recognised thatThe Lost Expedition was developed not as a serious game for training purposes, but as a popular game for general entertainment. Other popular games also use story telling and aesthetic choices to challenge players with moral choices, be it through the crossroads cards in the Dead of Winter (Plaid Hat Games, 2014) games, or asking players how far they would go to survive in This War of Mine (Awaken Realms, 2017) which is based on the Siege of Sarajevo. Other games are less explicit in this aspect of design choices, but may still choose to humanise what could otherwise be non-descript pawns to add extra weight to the implications of decisions. Days of Ire: Budapest 1956 (2016) for example, which is based on Hungarian revolution of the same year, includes historic names on each of the revolutionary markers, as well as historic background on the events cards in the manual. These elements add another layer of depth to a game which could otherwise simply be a strategic puzzle, and encourage players to consider what the human cost of their decisions would be.
In addition to using moral dilemmas as a way to encourage players buy into the universe of the game, designers also make aesthetic choices to prompt emotional reactions. This may range from using cute and cuddly imagery to encourage players to smile and laugh, or even quite the opposite. This is certainly the case in Raxxon (2017) which is set in the Dead of Winter universe during the early stages of a zombie outbreak, requiring players to manage a quarantine and separate the healthy population from the infected. Here, players are presented with cards which not only depict ravenous zombies, but also healthy individuals and various other groups such as uncooperative but healthy, violent individuals, and carriers who could spread the infection. Each of these different groups presents players with different issues to consider when managing a crowd formed of a mixed population, with the game employing push-your-luck and role specialisation mechanisms. Moreover, the illustration choices used on the cards can prompt a player to revel in calling in an airstrike to remove zombies from the crowd, or give them a moment’s pause when dealing with carriers who look like they may just have a bad cold. The design choice to use black and white images which focus on the characters’ facial expressions against a coloured background (Figure 2) starkly portray individuals at a moment of personal crisis as they await to find out if they will be taken to safety or left with the zombies. By doing so, this choice puts players in the role of a frontline responder who must deal directly with the public, once again adding a layer of depth to a problem-solving puzzle.
Figure 2: Examples of Raxxon crowd cards.
This is all very well for popular games focusing on entertainment, but is there also an opportunity for serious games to use similar design choices to create discussion points and teachable moments? It is arguable that the more limited market for serious games means that there may not be as much of a financial incentive to develop their aesthetics in the way that commercial entertainment games do. Many serious games also choose to focus on systems where emotional considerations do not have to be included in training, and a print and play approach is aesthetically acceptable. Some sectors, however, may find that including an emotional element is of great benefit to frontline staff who have to deal with the public. In the disaster response sector Thomas Fisher has commented that no matter how well players do in AFTERSHOCK: A Humanitarian Crisis Game (2015), thousands of people will die in the game. This provides an opportunity for those who are new to the sector to reflect on their own feelings to this simulated loss of life and consider whether a career doing this sort of work is really for them. It is also worth noting that Fisher has made the point that when considering design choices for AFTERSHOCK, a conscious decision was made to avoid gratuitous images. Nonetheless, it can be seen that there are some similarities between the illustrations used for Raxxon and some of the Images used in the “at risk” deck for AFTERSHOCK (Figure 3). Unsurprisingly perhaps, the image of children in distress could be considered an effective shorthand for provoking emotional turmoil among players.
Figure 3: Examples of images used in the AFTERSHOCK “at risk” deck.
If we agree that the design and story choices used in games can provoke emotional reactions and moral dilemmas, how can we develop these ideas as effective teaching tools? One possibility would be to use emotional triggers in games to help players become more aware of their own decision-making processes. With practice, this could also help them become more confident in their intuitive decision-making when there is limited time or opportunity for planning and analytical-decision making. In a game this might be done by using art and story to prompt an emotional or moral reaction which if acted upon would be considered irrational play in an athematic puzzle or even an abstract game. This might mean putting triggers on cards which are comparatively high risk and low reward in a game, and observing if they are acted on more frequently than low risk and high reward cards which have neutral imagery. As always, one should consider the learning objectives one is working towards when designing a game, and how different mechanisms can be used to foster different behaviours. The approach described here may be useful for addressing humanitarian principles, for instance and one could discuss the choice of helping an individual in obvious distress while ignoring a card with a higher value which could represent faceless masses. Furthermore, emotional triggers should not be simply limited to images of crying children but could instead be more subtle and nuanced. An example of this might be addressing the humanitarian principle of impartiality by depicting a diverse population and seeing if there is an expression of personal bias in the players’ choices.
In conclusion I think that use of design choices and story should be carefully considered as a game-based learning tool. Not only should aesthetics be considered as a way of making a product appealing to potential buyers, but careful choices have the potential to provide effective learning outcomes. I certainly hope that this will prompt further discussion and study to establish if these ideas can be developed further. Many of these ideas are already put into practice in live simulation disaster response exercises, for instance by using actors, moulage and prosthetics to provide responders with distressed casualties who may not be cooperative. I certainly think that incorporating story and push-your luck elements into exercises could also benefit them, for instance providing a threat to team safety which may influence deployment decisions. The social, face to face nature of board games also makes them an ideal platform in which to practice skills with a social element in a simulated dynamic and developing situation at relatively low cost and with potentially high engagement among participants.
The Contested Narratives Wargame builds on the assertions from Peter Perla and Ed McGrady that wargames “embod[y] two types of narrative: the presented narrative, which is what we call the written or given narrative, created by the game’s designers; and the constructed narrative, which is developed through the actions, statements, and decisions of the game’s participants.” Over the course of the game, select participants shared presented narratives (pre-scripted stories) to amplify or dampen adversary and friendly narratives. Participants then moved between tables developing constructed narratives (revised scripts) amidst the various contested narratives. Using the World Café method, a professionally and nationally diverse group of participants took turns sharing stories of national resilience against malign influence wherein the pre-scripted presented narratives contest for resonance.
The World Café is an exploratory method, designed by Juanita Brown and David Isaacs, that elicits communication patterns. Set in a café-like environment with multiple tables, participants are invited to sit in small groups with participants from other nations. A facilitator initiates the conversation with a narrative prompt to the entire room—“share a story about national resilience,” for example. Then the participants engage in multiple rounds of storytelling. Paper tablecloths and colored pens allow participants to scribble and take notes creating artifacts for later review. As participants move around the room, narratives begin to circulate. Contestation emerges as designated players introduce stories scripted prior to the wargame from an adversary’s perspective. At the end of several rounds, Dr. John DeRosa—game designer, lead facilitator, and one of the authors—led discussions with the participants to find the l’entre deux, the between place, of presented and constructed narratives circulating within the room. In this sense, the process seeks to reveal if elements of the pre-scripted narratives (like those representing the adversary) appear in the revised scripts developed within the wargame.
Two key insights emerged. First, stories coupled with symbols construct powerfully resonant narratives. Second, unlike the linear action-counteraction-reaction model of traditional wargames, methods like the World Café can effectively mimic the complexity of the human dimension.
At War on the Rocks today, James “Pigeon” fielder discusses how to teach wargame design, drawing on his experience at the U.S. Air Force Academy.
I founded my course on three pillars: defining wargames, objective-based design, and learning outcomes over winning. First, I took a blend of James Dunnigan, John Curry and Peter Perla, Phil Sabin, and my own caffeinated madness to define wargaming as “a synthetic decision making test under conditions of uncertainty against thinking opponents, which generates insights but not proven outcomes, engages multiple learning types, and builds team cohesion in a risk-free environment.” Second, I enshrined the primacy of the objective. Put bluntly, without objectives you don’t have a professional game. Although we briefly discussed creating sandbox environments for generating ideas in the absence of objectives, sandbox design at best strays into teaching group facilitation (albeit game refereeing itself is a form of facilitation), and at worst enshrining poorly structured and long-winded BOGSATs as legitimate analysis tools. Finally, neither the U.S. Strategic Command wargame nor the National Reconnaissance wargame included absolute and predetermined winners. Both U.S. Strategic Command and the National Reconnaissance Office faced unmitigated disaster every time they bellied up to the table. The best learning comes from understanding failure, correcting mistakes, and revising strategies, not from sponsors patting themselves on the back. Summoning Millennium Challenge 2002’s chained and howling ghost, gaming with the sole intent to win, prove, and prop up ideas is an exercise in false future bargaining with real lives and materiel.
He cleverly had his cadets design games for real sponsors:
I divided the class into two eight-cadet teams respectively for U.S. Strategic Command and the National Reconnaissance Office. The sponsors and I initiated dialogue, but from that point the games were entirely cadet driven. The teams interviewed the sponsors for objectives, determined how to measure the objectives, prototyped and play-tested their games, and ultimately delivered effective tools for addressing sponsor requirements. Meaning, of course, the games generated more questions than answers: better to ask the questions at the table before bargaining with a real opponent or launching a new military service.
There’s a lot more besides that, including a discussion of the wargame design literature, as well as material on psychological roots and sociological narratives of gaming. James also discusses the importance of learning-through-play.
Go read the entire piece at the link at the top of the page.
RAND has released a new report by Stacie L. Pettyjohn and Becca Wasser on Competing in the Gray Zone: Russian Tactics and Western Responses. This addresses two major sets of research questions: first, “How are gray zone activities defined? What are different types of gray zone tactics?” and second “Where are vulnerabilities to gray zone tactics in Europe? What are those vulnerabilities?”
Recent events in Crimea and the Donbass in eastern Ukraine have upended relations between Russia and the West, specifically the North Atlantic Treaty Organization (NATO) and the European Union (EU). Although Russia’s actions in Ukraine were, for the most part, acts of outright aggression, Russia has been aiming to destabilize both its “near abroad” — the former Soviet states except for the Baltics — and wider Europe through the use of ambiguous “gray zone” tactics. These tactics include everything from propaganda and disinformation to election interference and the incitement of violence.
To better understand where there are vulnerabilities to Russian gray zone tactics in Europe and how to effectively counter them, the RAND Corporation ran a series of war games. These games comprised a Russian (Red) team, which was tasked with expanding its influence and undermining NATO unity, competing against a European (Green) team and a U.S. (Blue) team, which were aiming to defend their allies from Red’s gray zone activities without provoking an outright war. In these games, the authors of this report observed patterns of behavior from the three teams that are broadly consistent with what has been observed in the real world. This report presents key insights from these games and from the research effort that informed them.
Can a game model gray zone competition in a empirically ground sound yet playable way?
What is the game design process for developing a structured strategic game for a complex political-military issue that simultaneously operates in two different time horizons?
How can structured strategic gaming help researchers gain an understanding of adversary gray zone tactics and tools?
To explore how Russia could use gray zone tactics and to what effect, the authors of this report developed a strategic-level structured card game examining a gray zone competition between Russia and the West in the Balkans. In these games, the Russian player seeks to expand its influence and undermine NATO unity while competing against a European team and a U.S. team seeking to defend their allies from Russia’s gray zone activities without provoking an outright war. This report details the authors’ development of this game, including key design decisions, elements of the game, how the game is played, and the undergirding research approach. The authors conclude with recommendations for future applications of the game design.
The Balkans gray zone game demonstrated that structured strategy games are useful exploratory tools and this model could be adapted for other contexts and adversaries.
While the gray zone remains a murky topic, this game demonstrated that it was feasible to break the gray zone down into concrete parts, to conduct research on each of these parts, and to link these components to create a playable strategic game that yielded useful insights.
The scoped and structured approach to this game allowed for enough structure to keep discussions on track and provided links between inputs and outputs while still allowing for creativity, flexibility, and transparency.
This gray zone game can be adapted to focus on different regions or adversaries, could include additional allies, or could be made into a three-way competition.
The RAND team started with a series of matrix games to scope out the problem, and then progressed to semi-structured game. Finally, they moved on to creating a structured, three-sided (US, Europe, Russia) gray zone board game focused on the Balkans.
Countries were tracked for governance quality and diplomatic-political orientation, as well as economic dependence (on Russia) and media freedom.
Players acted through a deck of action cards, each specific to the actor(s) they represented. Potential Russian (RED) actions are shown above, and sample cards below)
The report discusses the game design approaches taken, assesses their utility, and concludes with some suggestions as to future modifications.
All-in-all, it is a rare and outstanding example of serious game designers fully documenting their game design approach and research methods so as to inform future work on the issue. Kudos to all!
The following piece has been contributed to PAXsims by James Halstead.
Part Two: Applying the Framework
In Part 2, the framework introduced in Part 1 will be used to study debates around a historical battle: the 1917 Third Battle of Gaza. The ‘Gaza School’ counterfactual has been a recurring element of the battle’s historiography since its inception in the immediate aftermath of the battle and was brought to greater prominence in the 1930s with Clive Garsia’s book A Key To Victory which continues to be an influential source for studies on Palestine. The Gaza School therefore remains an intriguing counterfactual possibility amidst continuing debate within the historiography
The Third Battle of Gaza
The ‘Gaza School’ debate revolves around the strategy employed by Edmund Allenby to eject Ottoman forces from their defensive line between the towns of Gaza and Beersheba in southern Palestine through October and November, 1917. Historically Allenby launched attacks on either flank of the Ottoman line between Gaza and Beersheba, drawing Ottoman reserves to both flanks before breaking through the weakly held centre. The inland flank was attacked first with the Desert Mounted Corps (DMC) and XX Corps outflanking, surrounding and capturing Beersheba. Meanwhile XXI Corps diverted Ottoman reserves with a holding attack on Gaza while the formations in Beersheba built up water stockpiles then broke through the Ottoman centre, forcing a full-scale Ottoman retreat.
Garsia champions the ‘Gaza School’ counterfactual in his book, A Key To Victory, which posits Allenby should have eschewed the attack on Beersheba and focussed all resources upon breaking through at Gaza then exploiting with cavalry rather than outflanking the Ottoman line on the more logistically precarious inland flank. This article will use the wargaming research framework laid out in the first part to explore the feasibility of Garsia’s alternative plan. Indeed, the suggestion to use a wargame to model this came as early as the early 1930s in Cyril Falls Official History.
A study of the terrain reveals the difficulty of attacking Gaza with several hills, traditional fieldworks and thick cactus hedges all significant obstacles and made the town difficult to take. Two attacks at the beginning of 1917 had already failed while XXI Corp’s holding attack during Third Gaza did poorly, failing to achieve the modest objectives set.While Garsia argues Gaza could have been masked by XXI Corps while the DMC broke through along the beach even this argument is difficult to qualify. High sand dunes near the coast made the ground unsuitable for wheeled vehicles and would make the movement of three cavalry divisions burdensome. Force to space ratios are also often forgotten and a study of a map reveals the beach route offered a frontage less than a mile wide, through which three cavalry divisions would have to ride. This would necessitate a limited, single Brigade front to overcome the Ottoman positions codenamed Lion, Tiger and Dog positions and then further redeployments and fighting across the Wadi Hesi before the cavalry could cut Gaza’s supply, while a long spread-out column of cavalry might prove vulnerable to artillery fire and Ottoman counterattacks regaining the beach defences.
Adherents to the Gaza School maintain the coast road would have made movement and supply much easier, there are a number of factors which discount this. This road moves directly through Gaza; where the heaviest held part of the entire Ottoman line was located so use of the road would have necessitated decisively shattering the heaviest part of the Ottoman defences, before pushing three cavalry divisions across heavily fortified ground, through a major urban area, across the heavily held Wadi Hesi and all along a single-track road.
Exploitation along the coast would also be harder than supposed with XXI Corps advance following Gaza’s evacuation requiring tractors to move supplies along the coast even with road access. Heavier sand also exhausted the cavalry’s horses and bogged down wheeled transport making rapid movement difficult. There were therefore significant obstacles to cavalry exploitation as a serious study of the terrain demonstrates.
Order of Battle and Generic Capabilities of Formations
Study of the order of battle reveals several insights. Firstly, that while the strength of Ottoman formations was highly variable, and the specifics of the numbers employed still remain unknown, they appear to have concentrated their best divisions on the coast behind Gaza. The historical attack on Beersheba pulled these troops away from the coast, to reinforce the inland flank, although even then there were still sufficient reserves to reinforce Gaza against the holding attack by XXI Corps. To focus the offensive on Gaza would, very likely, have meant that Ottoman forces could have concentrated upon holding Gaza and the terrain behind it even more rather than being split between two axis of advance as they were historically.
Third Gaza also provides an example of how order of battle research can reveal sources ignored by military historians in the 15th Imperial Service Cavalry (ISC) Brigade. Cyril Falls omits the Brigade from the Official History’s Order of Battle, a mistake which future historians have copied and while the absence of a lone Brigade may not seem especially significant the existence of 15 ISC is significant because the brigade’s performance during the battle provides direct evidence of how effectively larger bodies of cavalry would have operated on the coastal flank.  Garsia argues it would have been sufficient to simply mask Gaza with XXI Corps and then slip the cavalry along the beach to cut Ottoman communications. 15 ISC’s war diary, however, makes it clear that opening up a gap, and breaking through, would be no simple matter. The Brigade did actually form up behind the XXI Corps infantry assault but were unable to exploit through as Ottoman counterattacks recaptured the beach defences. Additionally, as discovered in the survey of the terrain the heavier sand on the coast would have exhausted the cavalry. Cavalry tactics also heavily relied upon infantry, artillery and air support. Any unsupported cavalry penetration behind Gaza would struggle against renewed Ottoman defences and counterattacks as shown by EEF cavalry actions at Huj on November 8, Beit Hanun and in the (attempted) crossing of the Nahr el Auja. In all of these cases unsupported cavalry on the advance struggled to overcome what were often weakly held defensive positions and indicates that the cavalry might not even have been able to achieve their objectives even had they broken through.
The EEF’s Decision Making Environment
While the creation of the physical model demonstrates the difficulties with the Gaza Camp approach further analysis of the decision-making environment in the EEF in autumn 1917 further supports a case that the Gaza School approach simply did not align with EEF strategic priorities. Philip Chetwode wrote in October: ‘it is desired to get the enemy on the move from his strongly entrenched positions with as few casualties as possible, relying on our preponderance in cavalry to do the execution.’ It is also worth bearing in mind the directive given to Allenby before the battle to capture Jerusalem and ‘occupy the Jaffa-Jerusalem line’ as cheaply as possible. Preponderance in cavalry, and the advantage this gave, was a clear motivation for seeking the open, inland flank. While the EEF had three cavalry divisions, and three independent cavalry brigades the Ottoman cavalry only consisted of one division, barely stronger than a British cavalry brigade. Turning a weakly held flank would also likely be much cheaper than a head-on assault against the most strongly held part of the Ottoman line. The more indirect inland route via Beersheba was chosen because it maximised the EEF’s advantage in cavalry while helping to keep casualties as low as possible. XXI Corps losses in just their holding attack on Gaza were double those of the assault on Beersheba; and for little tangible gain with even the single Brigade of cavalry present unable to exploit.
Allenby’s decision to risk the inland attack on Beersheba therefore is as much to do with wider strategic priorities as it is to do with the practicalities of the terrain and force composition.
Integrating wargaming within military historical research, not just within the context of counterfactuals, offers a number of important tools that military historians continue to underutilise. By creating an analytical model of events that aims to conform with the course of historical events military historians can analyse individual factors based on under-utilised (but commonly available) evidence while the successful creation of an accurate model encourages historians to explore the full range of evidence. If the model doesn’t work for whatever reason, then this simply encourages further research to understand why the model doesn’t conform. Extra playtesting and refining of the model is something that can introduce previously unknown or unconsidered factors that suddenly appear more decisive for their effect on the accuracy of the model.
Wargaming military history therefore, while still a tool for support of a wider analytical goal (and as such should be employed appropriately), fills in a number of crucial gaps within a military historian’s toolkit. Design of a wargame encourages rigorous analysis of under-utilised sources in a wider framework and, most importantly, incorporates these into a wider model which must be adapted to fit the historical result. When an initial model doesn’t conform then this just encourages further exploration of why your rigorously researched model hasn’t conformed. Much like wargaming mechanics this creates an important feedback loop, and encourages the researcher to go back and check their sources again: something that the dominant research methodology within history fails to do. Indeed, much of the time in traditional military history contradictory, and inconvenient, sources are often seemingly explained away, ignored or subsumed into wider arguments. Wargaming encourages a more involved research process right from the beginning of a project and, furthermore, relies upon sources that very often can be easily obtained without endless days in the archive. Meanwhile testing the design, especially with a third party, can often lead to fundamental reevaluations of either sides decision space: ‘what constitutes ‘victory’ for either side and what are they willing to risk to attain it?’ are just two questions that applying a gaming approach can encourage. Designing a wargame for a battle at the outset of a project can often produce new priorities on archival research and when new evidence is discovered allows it to be reincorporated into the model: often improving the pursuit of a historically accurate result. While military history is increasingly moving to incorporate more qualitative, and innovative methodologies there are still ways that military historians can integrate more traditionally social science approaches like modelling, and wargaming, to the benefit of their research.
 Clive Garsia, A Key To Victory: A Study in War Planning (London: Eyre and Spottiswoode, 1940)
 Cyril Falls, Military Operations Egypt and Palestine: From June 1917 to the end of the War Part I (London, 1930), p. 32
 SHEA 6/2, The Liddell Hart Centre for Military Studies and JONES, CF, The Liddell Hart Centre for Military Studies
 Cyril Falls, Military Operations Egypt and Palestine: From June 1917 to the end of the War Part I (London, 1930)
 Lieutenant Colonel, The Honorable, R.M.O, Preston, The Desert Mounted Corps: An Account of the Cavalry Operations in Palestine and Syria 1917-1918 (Boston, 1920), p. 6
 Falls, Official History p. 142 and Marquess of Anglesey, A History of the British Cavalry 1816-1919: Volume 5, Egypt, Palestine and Syria (London: 1994) p. 188
 Anon. History of the 15th Imperial Service Cavalry, p. 17
 Wavell, Allenby: Soldier and Statesman p. 178
 John Ericksen, Ottoman Army Effectiveness in World War I: A Comparative Study p. 123
 Phil Sabin, The Future of Wargaming to Innovate and Educate, Public Lecture at Kings College, 22.11.2019
 Jonathan Fennel, Fighting the People’s War (Oxford: Oxford University Press, 2019); Ben Wheatley, A Visual Examination of the Battle of Prokhorovka (Journal of Intelligence History,), Volume 18, 2019
James Halstead is a military historian who is primarily interested in the two world wars of the 20th century. He studied for his Masters at Kings College London (including Professor Phil Sabin’s Conflict Simulation module) and is currently studying for his PhD on Information Management in the British and Commonwealth Armies at Brunel University, London. James has delivered lectures on the Royal Flying Corps and Air Force in the Palestine Campaign at the RAF Museum, Hendon and will do so again at Wolverhampton in 2020. James can be found either on twitter at @JamesTTHalstead or you can read his research blog at: youstupidboy.wordpress.com