Designing for Cause vs. Designing for Effect
in Historical Games
Rather than argue definitions at length, I’ll briefly state the difference between designing for cause and designing for effect as I discuss it here. The idea is that when you design for cause, you find the factors that caused something to occur, and design those factors into your game so that it’s likely to occur. When you design for effect, you design the game so that the effect, the result, is a recognizable representation of history. Causes may or may not be reflected, but the goal is the effect, what happened, not why it happened.
Some commentators criticize “design for effect” in wargames. My thesis here is that such criticism is undeserved because “design for cause” is possible only at a low tactical level. From my point of view as a person educated to understand military history, virtually all board wargames, certainly all above a very low tactical level, are necessarily designed for Effect, not for Cause. Designing for Cause is a chimera, something that rarely can actually be done, or if it is, the history will appear “skewed” or wrong. Why? Three reasons:
• 1) reality is too complex,
• 2) reality is strongly influenced by chance, and
• 3) game design is subject to the problem of foresight/hindsight.
Games as Models of (a) Reality
Games are models. Reality is too complex to represent completely, so we simplify. In effect, an “idealist” is needed, not a realist, to make a model of something as complex as a battle or war, because the realist will be tied up in knots trying to include too many causes. Some might argue that what is left in the model can reflect causes, but I would argue that, given the simplicity of board and card games, there is very little room for showing causes. Computer games, which allow much greater complexity of calculation, can come closer to incorporating causes into the game.
Let’s take a simple example. My military history prof used to enjoy saying the Romans succeeded because there were "too damn many Romans". You can't simulate the complex reasons why the Romans could have so much more manpower, all you can do is reflect the effect, that there are lots of Romans and their method of raising manpower is different. (Those who want to try to reflect the causes of the great Roman advantage in manpower–which was not because of a much larger population–can start with the 759 page Italian Manpower 225BC-14AD, by Brunt–have fun.)
At some point people will disagree about this without any common ground for discussion–some will say “yes you can”, others (like me) will say, “no you can’t” design a simple game that reflects causes, so I’ll go on to the next, much more compelling, reason why “design for Cause” is a phantom.
Every historical situation is, in effect, a Monte Carlo simulation where only one of the thousands of possible repetitions becomes the “real history”. (A Monte Carlo simulation is “a numerical modeling procedure that makes use of random numbers to simulate processes that involve an element of chance.”) That’s because history involves a great deal of chance.
This chance involves
• 1) the “butterfly effect”,
• 2) the importance of actions of individuals, and
• 3) (once again) the complexity of reality.
The following traditional poem illustrates much of this idea:
For want of a nail the shoe was lost.
For want of a shoe the horse was lost.
For want of a horse the rider was lost.
For want of a rider the battle was lost.
For want of a battle the kingdom was lost.
And all for the want of a horseshoe nail.
In other words, small chances can have great effects. We see this notion of randomness reflected in Chaos Theory and “the butterfly effect”. This is a term for sensitive dependence on initial conditions. A small thing early on can have big effects as time passes. You see this reflected in most stories about time travel, a small change “back when” makes a big difference much later. Do small changes propagate bigger ones over time, or do they “damp out”, and history returns to its “natural” course? I’m of the “propagate” school, because I think history has so much randomness in it, sometimes a small change will result in larger ones, though most of the time it will not. If chance truly plays such a strong part in history, how much can you “design for Cause”? The biggest “cause” is probably randomness itself.
The alternative to this argument is that chance “dampens out” over time, that is, that chance plays next to no part in what happens. This might be so when time travel is involved; it otherwise appears to be a mystical belief rather than something supported by evidence.
The Importance of Individual Actions
Look at the importance of the individual, especially in recent history. Individual men, and the loss or death of such men, have made or broken entire nations. Here again we have competing notions, that “time makes the man” (and consequently any given individual is not important in the long run), as opposed to the “heroic” view of history, that certain individuals make a big difference. Nothing here can be “proved”. However, I strongly believe that individuals make a big difference. Perhaps Alexander the Great would have been replaced by someone else who could have done almost as well, but what would World War II have been without Adolf Hitler–would it ever have occurred?
We don’t know who most of the individuals were in ancient history, but this doesn’t mean they weren’t there or weren’t instrumental in what happened. We do know about Heraclius, Mohammed, and going back further, we know Sargon I and Naram-Sin. But who led the Medes who helped sack Nineveh? How did Assyria fall so quickly after the power and glory of Ashurbanipal, was it only the absence of a strong leader? Could the Persians have prevailed without Cyrus the Great? Individuals, with all their human foibles and irrationalities, can be influential out of all proportion to what one person can ordinarily do.
The “best side” doesn’t always win a battle, owing perhaps to the effect of one individual, and even in a war so much can depend on a single individual that the loss of that person can turn the tide (think Pericles in Athens, think how many empires fail when the emperor who built them up dies (Harsha in India, Timur the Lame in central Asia), think how many nations depended almost entirely on one man’s skill or drive (Hannibal in Carthage, Stilicho and Aetius in late Rome, even Hitler in Germany). What if William the Bastard had died early in the Battle of Hastings? Many battles have been won or lost because a leader died, or his troops thought he had died.
In ancient and medieval times, smart rulers often avoided battle, because they knew how chancy it could be. Charlemagne rarely fought a pitched battle, yet acquired large tracts of territory. The Vikings tried to avoid battle until they became kingly expeditionaries, and lost battles as often as they won them.
Complexity of Reality
Let’s go back to the Monte Carlo Simulation idea. If you include chance factors in a simulation of anything even beginning to approach the complexity of history, you’ll find that there are many, many possible outcomes in your simulation. Some of those outcomes, sometimes the most common outcome, will resemble actual history. But in many other cases, the most common outcome, if you could model causes perfectly, will be far from the actual history, because chance skewed the “real” history in some way. So the actual outcome of the Battle of Midway might be way down near the edge of the bell curve of probable outcomes, a matter of chance that the American carriers were able to attack first, just as the Japanese carriers had planes and fuel littering their decks.
Some of you may have read some of the “alternate history” articles, stories, and books. I have a “serious” book called What If sitting on my shelf, and some authors such as Harry Turtledove (who has a Ph.D. in Byzantine history) have made a career out of alternate history stories. Sometimes the alternative is more believable than what really happened, possibly because it really was more likely.
At some point this becomes a philosophical or even “religious” question. If you believe that “everything happens for a reason”, that “the best team always wins”, that everything in the world is deterministically inevitable, then you will completely disagree with my assertion that history involves a lot of chance, and we have no common grounds for agreement. But let me make one try to convince you otherwise.
Randomness in Life (and History)
I've said that history is heavily affected by random occurrences. For those who don't believe that, I'm going to make one try at an analogy to see if I can change a few minds.
Does the best team win in team sports? Some people actually believe so, yet a look at recent (2008) results would indicate very much otherwise. The New England Patriots had won 18 American football games in a row, yet lost to the New York Giants in the 2008 SuperBowl. If the teams played ten times, New England would certainly have won a majority (as they won against NY earlier in the year), but they lost that one time. So which was the best team?
One game doesn’t determine who is the best team, that’s why many playoffs in American professional sports are best of seven games (baseball, basketball, ice hockey). It would be best of seven in football, if this were practical.
Let’s look at the results of recent single-elimination (or near it) tournaments. In the 2008 NCAA college basketball tournament all four top seeds as chosen by the selection committee made it to the “Final Four” for the first time ever. Frequently, the winner is not one of the top seeds. I happen to be a Duke fan, and I think there have been years when Duke was the best team in the nation but didn’t win the tournament, and at least one year when Duke clearly wasn’t the best team but won the national championship.
I like to point out to my game design students that even if a team has a 90% chance of beating its opponent--an unlikely high percentage, given the country's best teams are in the tournament--that team only has a 53% chance of winning six in a row, as required to win the tournament.
As another example, in the English FA Cup (soccer) competition of 2008, only one of the four semifinal teams was a member of the top English league division (divisions are arranged in order of strength, so the top division contains all the top teams, the bottom division (of four) the weakest teams). In the same year, the four teams at the top of the league standings, Manchester United, Chelsea, Arsenal, and Liverpool, who dominate the league year after year, all made it to the final eight of the Champions League, and two played the championship game, in a competition for all of the best teams in Europe as a whole. Yet none of them were in the FA Cup semifinals.
Does the best team always win? No.
Here is another example of the randomness of life. In 1973 I had a high enough number in the draft lottery–lowest number drafted first--that I was pretty sure I would not be drafted. A friend of mine (who I didn’t know at the time) had a low number, so he joined the army. He is now a retired officer. How different would our lives have been with different birthdays (which is what the lottery was based on)? How many people who might have made fine generals never entered the army, because they had high numbers? How many who might have been great musicians were drafted into the army, and could not continue to develop musically? And so forth...
To continue the theme of the randomness that affects history, consider that after the fact we tend to impute more planning, more awareness of long-term objectives, than actually existed. When the Visigoths moved out of the Balkans early in the fifth century, did they have an overall plan, did they intend to end up in Iberia before the end of the century? No, at first they just wanted to pressure the Romans so that their leader Alaric could gain a major leadership role in the Roman army. When that didn't work, their major goal was finding food. Alaric likely wanted to go to Carthage because it was a major source of food surplus, second only to Egypt. When his ships sank and he died in 410, the Visigoths wandered ultimately into Gaul, and settled for a long time in Acquitaine when the Romans offered them the revenues (if not quartering) of the area. They went to Iberia the first time as part of a deal with the Romans to harry the Vandals, the second time because the Vandals and Alans had abandoned the place for Africa at the request of a Roman official in Africa.
There is no practical way, in a boardgame of large scale, to design this "for Cause". We can't reflect the randomness of food sources and rumors of food sources, of sunken fleets and Roman court intrigues. In a game it's possible to give the Visigoth player points for ending up in Iberia, the Vandal player points for ending up in Carthage. Even with point incentives, very strange history can result from a "sweep" game like Britannia. Anyone who thinks such games are "too scripted" hasn't played or watched a few dozen games.
If we go back far enough we have the additional limitation that we don’t really know what happened. Can an ancients battle game show how the Romans fought? The Romans who wrote military treatises assumed everyone reading them would be familiar with battle either directly or through others easily referenced, so they are unclear about essential points, such as how the three Roman infantry lines dealt with the gaps between units. Nobody knows. Some historians like to pretend otherwise, but to a remarkable degree, we really don’t know what happened in history, let alone why it happened. Everything we know about Alexander the Great comes from tertiary sources such as Plutarch, writing 350 years after Alexander’s death. How do we create games in those circumstances that might “design for cause”?
"War is the province of uncertainty: three-fourths of those things upon which action in war must be calculated, are hidden more or less in the clouds of great uncertainty."
Karl von Clausewitz
Next we can consider the effects of foresight/hindsight. Perhaps the greatest cause of everything in history is uncertainty. Yet much of that uncertainty does not exist in the typical situation modeled by a wargame.
We know the Germans made the right wing strong in WW I, we know they counterattacked at the Bulge, we know the Americans surprised the Japanese carriers at Midway, and that knowledge skews a game in many cases. The best "simulations" don't try to represent a particular actual event, but possible events, such as scenarios in squad-level games.
It’s nearly impossible to design for Cause when modeling a single battle, because one of the biggest causes of what happened in an historical battle is the uncertainty about what was going to happen. We, after the battle, cannot experience or duplicate this uncertainty yet still model that particular battle.
We can come closest with tactical games that enable lots of “what if” scenarios, but even in such cases, we may know much more about the effectiveness of weapons and troops than the historical participants knew.
It seems to me you really have only two choices: accept that hindsight will change the player's behavior, or put the player in a straitjacket (a la many old SPI boardgames) and give him no *practical* choice other than what happened in history.
We can still have good games about specific battles, but not generally ones that are designed for Cause, and not ones that model the problems and feelings of the actual participants--because the players know much too much.
Can you design a game about future conflicts "for Cause"? It's much more practical, because you've removed much of the foresight/hindsight. Some people point to simulations used by the military as examples of designing "for cause". Perhaps:
1. Military simulations are not historically-based, that is, they're not modeling a specific event, so there's much less hindsight/foresight.
2. Military simulations are often computer-based. You can do much more "for Cause" with computers than not, because the computer gets you closer to the complexity of reality--provided you have enough data. It is less impractical to design for cause in a computer game. Players don’t know what the causes are inside the program. Much more detail is possible because the computer tracks and calculates it, not the players.
3. Many people convince themselves they're designing for Cause when, in fact, they are not. Moreover, there are many instances in history of "war games" fooling military leaders (who often let themselves be fooled).
4. Much of the time, military simulations are designed to confirm whatever the current conventional wisdom may be. That is, they aren't designed for Cause, they're designed to evoke a particular effect.
My experience designing battle games. I have designed only one game about an historical battle, the Battle of Hastings. This is a simple game using cards as unit markers and a single die. In early playtests it fairly well resembled what little we know of the actual battle--and it was dull and much subject to chance. That's because leaders of ancient and medieval battles rarely had much control over what happened, once hostilities commenced. They could lead the troops they accompanied and rarely could affect what happened elsewhere--if they lived. In the game, the player could (unlike the real world) move all the units as he wished, but there was little room to maneuver--it was a slog. Yet sheer chance, in the death of a leader or even the rumor of the death of a leader, could have a great effect, and I included that possibility in the game.
In the interests of gameplay I changed the rules to allow more maneuver, and more resilience amongst the troops, to try to make an interesting (though simple) game of it. The result is indeed more interesting and less chancy, but it's not a game that truly represents the actual battle, because that would be too dull and too fraught with chance to be a good game.
Games are designed to be played. What do players really want?
A problem for designers of historical games is that players often expect the game to resemble whatever history actually occurred, rather than what was most likely--least unlikely?--to occur.
The result? A writer on Boardgamegeek, Seth Owen, put it this way: "improbable events that actually occurred are normed in many designs, while sometimes ordinary events become outliers." To put it another way, you cannot achieve something that resembles history when designing for Cause, if (as often happens) chance skewed the actual results, if we got something toward an end of the bell curve rather than right in the middle.
In the "old days" of wargaming, SPI games tended to force particular results in order to “recreate history”. Yet insofar as history has a strong random element, and the “actual history” is often an unlikely history, this effort distorts everything.
Learning history from games
What about the people who use games to learn history? There are two objectives, to learn "what happened", and to learn "why it happened". You can think of these as Cause (why) and Effect (what), if you like.
I confess, if I want to learn history, I get much more from a book than from a game (at the "long" end of historical scale I especially love historical atlases written by Haywood and by McEvedy). But I recognize that I am in the minority of "learners from books", whereas most people are much more "learners from doing", which is where games become important.
In particular, I have never played games to learn “why it happened”: games (at least, the non-electronic variety) are insufficiently complex to show why it happened even if we didn’t have the problems of randomness and of foresight/hindsight. I can learn much more about why from reading a book than from playing a game. But we do have many people who enjoy playing games much more than they enjoy reading books, or they may not have access to appropriate books. They play games to learn something.
If the game is anything like "true to history", designed for Cause, then owing to complexity, the randomness of life, the affects of individual actions, and hindsight/foresight, the result of most plays will vary a great deal from history. If players are to learn "what happened" from games, the designer must design for Effect. If he can do that, and still make a game that is entertaining to play, he'll have done a fine job. But he could just as well try this and end up with something that is neither particularly good history, nor a good game.
I do not play miniatures wargames (too tactical and too imprecise, among other reasons). It appears to me, though, that the historical value of miniatures or miniatures-like boardgames such as Advanced Squad Leader--assuming you agree that there can be any historical value--is in "what if" scenarios, that is, ones that do not represent an actual battle but represent possible situations in history. THEN one player can truly not know about the existence of certain enemy units or unit capabilities, if a commander at that time wouldn't have known. And then, if it appears that the result of the game makes sense from what we know of history, the game rules *may be* "Good History" (again, if we can talk about such a thing...).
As a teacher, I firmly believe that designing a game to represent what happened, designing for Effect, is more instructive historically than designing for Cause, with the possible exception of very tactical games.
You can learn what happened from games, but you're much less likely to learn in detail why it happened. Yet you can design for Effect, and still highlight Cause. For example, using Britannia again, the Roman professional units are much better fighters than their opponents, and the Roman organization as reflected by the Roads makes an enormous difference in mobility. The main Britannia systems are designed to reflect migrations of nations, so the Romans have exceptional rules.
Did I determine that the Romans were X% more effective fighters and incorporate this into the rules? No, no such data exists (and if it did I would not have resorted to it). As I've said, we aren't even sure how Roman tactics worked. But we know they were a lot better than the "barbarians" at the time of the Roman conquest, and that can be reflected in the rules. In a more detailed version of Brit I have reduced most of the Romans to normal armies, late in their stay in Britain, to reflect the deterioration of Roman forces after the crises of the third century, but this level of detail wasn't appropriate in the original game.
For me, the trick is to design for Effect so that the game can resemble history, yet design a good game, a friendly competition. SPI games were often criticized as poor games. This is not so surprising, when SPI expected half the players to be playing solitaire, where competition does not matter.
The reason sweep of history games such as Britannia are singled out for criticism is because the “for Effect” is so obvious in how the victory points are allocated, and because it's particularly impractical to design "for Cause" on a scale of generations. Yet “for Effect” is just as obvious (to me) in any game that models a single battle, we just accept it there because there’s certainly no other practical way to do it.
The bottom line here is that history is in considerable part a very complex accident, and if that is so, a game that truly reflects history will rarely follow a course that appears to be much like "the real history". We need to recognize that historical games are models, constructs, that cannot come close to accurately modeling history, and even if they did, the result wouldn't LOOK like history. If we accept that, we'll avoid many pitfalls, we'll be able to make better games, AND we'll be able to make more games that educate, than we do now.