Another discussion of randomness

A recent episode of Three Moves Ahead took on my latest favorite topic, the role of randomness in games. Three Moves Ahead occasionally delves into boardgames, though its real bailiwick is computer strategy games. However, if you start talking about randomness, it's hard not to focus on boardgames, since they expose the random elements in ways that computer games don't. Or, as the panel in this podcast explore, maybe randomness plays a more important role in a game like Successors or Command & Colours: Ancients than it does in Starcraft? A great discussion, definitely worth a listen.

Randomness in boardgames: simulation

[For other posts in this series, click here, here, and here.]

When I started this series of blog posts, I deliberately avoided talking about the most obvious reason for including random elements in a game. For a game to be a credible simulation of a situation, even if completely fictional (for example, elves and orcs playing jai alai), it needs randomness. For reasons that I don't fully fathom, this argument for randomness sometimes leads into very heated discussions, so I thought I'd first tackle the less controversial but no less important reasons for randomness in boardgames. But there's no avoiding the necessity of randomness in anything that resembles reality.

No human activity is immune from having Dame Fortune appear unexpectedly to ruin your plans. Some situations may introduce fewer random elements than others: for example, you'll face fewer problems crossing the street than traveling to Vladivostok.You might stumble crossing the street, which is unfortunate. Traveling to Vladivostok, you might lose your passport, bad weather might cancel your flight, a pickpocket might steal your wallet, and no end of other mishaps might occur, orders of magnitude more than might happen crossing the street.

EXPECT PEOPLE NOT TO LISTEN TO YOU
Nowhere can you see the importance of randomness in simulations than in wargames. There is a lot of randomness in real-world combat, so it's hard to suspend disbelief when playing a completely deterministic wargame. Designers may choose to simulate different random elements, even when they are depicting the same historical event. If you play a Gettysburg wargame that users a chit pull system (for example, Across Five Aprils), you're not certain when a particular general will get his troops moving. Another Gettysburg game, such as Gettysburg: Badges of Courage, might give the surrogate Meade or Lee more control over when corps commanders order their troops to move or fight.

Every wargame designer imposes some upper limit on the amount of randomness in a game, and where to include it. Many designers are hesitant to include too many random elements in the command and control aspects of the game, even though the unpredictability of C3I is a major issue in real-world conflicts. After all, one of the most famous anecdotes about the battle of Gettysburg is Ewell's failure to take Cemetery Ridge, despite Lee's direct (but vague) order to do so. Whatever side you take on Ewell's hotly-debated decision, the issue boils down to one of C3I.

The importance of randomness in simulations goes beyond the narrative of a particular battle. (What if Ewell had advanced on Cemetery Hill? What if McClellan hadn't received a copy of Lee's Special Order 191, wrapped around a bundle of cigars?) Uncertainty shapes how military organizations structure themselves, and how they operate. If randomness is important for the narrative, it's also important for how you depict the major characters and their actions.

BUT PREPARE FOR THE UNEXPECTED
If you're not a military history buff, you might not understand why historians stress divisions over other levels of organization in accounts of WWII battles. In a book or documentary, you'll more likely hear about the Screaming Eagles of the 101st Airborne division, not the 2nd Brigade of the 101st. The uncertainties of command and control are a major reason for why divisions were the atomic unit of battle for the US army in World War II. Every army is a bureaucracy, with the same limitations on the amount of flexibility you can allow into the organization without introducing too much chaos.

The Army decided to make the division the smallest unit "capable of independent operations." Or, to put it another way, since the Army wanted to expect commanders at some level of the hierarchy to take the initiative, it decided to grant that latitude to the divisional commander. When assembling for battle, the Army could then attach or detach lower-level units (artillery, armor, etc.) to the division to modify its combat power. This model also made the lines of accountability clear more clear for success and failure on the battlefield, despite the tendency of junior and senior commanders to point fingers at each other.

Uncertainty shapes the way in which commanders make their plans. The more complex the plan, the more points at which it can break down. A chief defect of Yamamoto's plan for the invasion of Midway, for example, is his decision to break the fleet into four separate task forces. The probability that all four task forces arriving on station, as planned, was extremely low, leading many later commentators (including some Japanese admirals) to highlight  Japanese overconfidence (or "victory disease"). Had Yamamoto not scattered his ships and planes in this fashion, the Japanese fleet would have been better deployed to provide mutual protection while simultaneously executing strikes against the US fleet and Midway island.

PICK YOUR COMFORT ZONE FOR RANDOMNESS
How does a wargame simulate the risks that complex plans introduce? Frankly, most of them don't. You rarely see the importance of planning in actual combat, since units always take orders, they never get lost, and they don't encounter other mishaps when moving from point A to B. Wargames that do include the planning element, such as Fields Of Fire and the Tactical Combat Series, stand out starkly from the hundreds of other wargames that omit this feature.

Designers have some justification for not introducing too much randomness into the most basic tasks in a wargame, such as moving units around the map. Many of their customers complain loudly when a game doesn't allow this kind of unrealistic control over the battlefield, so designers are just responding to their market. To keep the experience of playing a wargame credible, you need to keep some element of randomness somewhere -- if not in the C3I aspects of the game, then in the combat results mechanics, or the timing of unit activation, or something else.

Randomness in boardgames: Forgiveness

[The latest in a series about randomness in games. Here are the links for the first and second parts.]

I have a reputation in our little gaming group for hating train games. I only partially deserve it, because the reputation started with things I actually said. I really dislike Ticket To Ride, for reasons that aren't relevant for this blog post. And I really hated my first session of Age of Steam, for exactly the reasons I'm writing this post. The fact that I dislike two train games doesn't mean that I'll never play a train game, but don't bother asking me to play either of those.

Age of Steam irritated me because, as a new player, it wasn't fun. Our game consisted of two noobs and two pros, and whaddaya know, the pros mopped the floor with the noobs. Turn one, I felt as though I was making a sub-optimal choice, but I had no idea how bad it really was until later. Every turn gave me greater awareness of exactly how dark the shadow of doom o'erhanging me really was. Decisions made early in the game came back to haunt me, so that by mid-game, it was clear that I was blocked from doing anything interesting.

I might have learned something from that experience that I could have applied to the next game, but frankly, Age of Steam just didn't grab me enough to make me want to play that game on that theme for another two or three hours. Maybe I'm wrong in making this conclusion, but Age of Steam did not feel like a game that would give me much room for trying out new strategies. The idea of playing a couple of turns, finding myself blocked again (it just seems like that sort of game), and then sitting around waiting for the end just didn't grab me.

To be honest, I'm not that interested in railroads, so the theme didn't give me impetus to play again, either. However, it's also the case that, unlike other Martin Wallace games I've played and loved, Age of Steam did not give much room for improvisation in the middle of the game. Nor did it feel, unlike Princes of the Renaissance, that the final outcome was really up for grabs until the end of the game.

In Princes, smart plays do increase significantly the odds of victory. But it's also entirely possible for new players to get enough of a clue to give more experienced players a real challenge. Princes provides just enough randomness where it matters, such as battle die rolls, to make it possible for players to catch up through smarter bids or more productive wars.

One of the most common complaints about randomness in boardgames is that it undermines the decisions you make. From one perspective, that's certainly true. In a game with no randomness, such as chess, you're completely in control of your fortunes. No flip of a card or roll of a die is going to screw up your strategy.

However, the lack of randomness makes learning within a game of chess possible, but not immediately applicable. If you're so new to the game that you don't understand how important the opening is, you'll lose to a more experienced player. Assuming you don't run out of patience with chess altogether before the game ends, you'll apply the lessons learned to the next session.

Chess isn't for everyone, however. It's not a casual game for friendly get-togethers in the same vein as Carcassone. You have to want to be a chess player to buy into the series of brutal lessons you'll endure as you gradually improve your understanding of the game. If you just want a quick diversion that poses an interesting challenge, and which rewards you with a score at the end, play something else.

If you want your game to appeal to more than just the most dedicated players, this is an important principle. Sure, bad luck can mess with you, but randomness also opens opportunities you might not otherwise have. Rather than watching the cascade effect of early mistakes unfold turn by turn until the inevitable conclusion, you can capitalize on a few lucky die rolls or card draws. The leader can be somewhat confident of a victory, but there's always the risk of an unexpected stumble. (Which makes the game more interesting for the leader, too.)

Randomness in boardgames: determining or managing?

[In the first post, I talked about how randomness can define the feel of a game. Now, I'm going to talk a bit about the type of player skills that randomness puts to the test.]

A recent episode of the podcast Flash of Steel talked about build orders in real-time strategy games, including Starcraft 2. The new Starcraft was, on release, a finely-tuned game. Since then, Blizzard just keeps tuning it. If someone figures out a build order that gives a significant advantage, in the same fashion as the infamous Zerg rush of the original Starcraft, lots of players adopt it. If it's imbalancing, Blizzard patches the game to re-balance it. And then online play continues until someone finds a new imbalancing strategy.

Does that sound like fun to you? Obviously, for millions of Starcraft online players, it must be. However, it's also a big turn-off for more casual players. Starcraft has turned into a very chess-like game, in which turn orders dictate the relative capabilities of the players. Turn orders (build a barracks, build four marines, etc.) are very deterministic, and they're arguably more important than battle tacti

cs. The marginal advantage you get from knowing where to position your siege tanks is much smaller than being able to manufacture siege tanks at a faster rate.

There's randomness in Starcraft 2, but it's not in the places where it can provide some breathing space for a new player. (Except, of course, if he's playing against an equally new player, which is one reason why Blizzard put a great deal of effort into matchmaking features.) Randomness is almost invisible in the game, so players don't give it much thought.

Boardgames sometimes follow this design approach. While I have little experience playing Age of Steam, I know that a more experienced player will make much better decisions governing what to do each turn, and where to do it. I'll be at a severe disadvantage against very experienced players, who eliminate bad choices with the same efficiency as good chess players.

If you were to make these games more random in the places where they're deterministic, a veteran player will still have a significant advantage over an inexperienced one. However, the game experience will be much different. Rather than feeling as though you're cracking

the puzzle of What's the optimal move this turn? you feel as though you're making choices among competing strategies. In the first case, you're trying to calculate what will work. In the second case, you're playing with options for what might work.

During one Advanced Squad Leader session, my friend and regular opponent Paul was having a very bad game. It looked as though, by the middle of the game, I was cruising to an easy victory. I urged him to hang in there, since (a) Paul is an excellent player, and (b) his luck could easily turn.

Paul won the game, after making some excellent tactical decisions. Luck mattered, both early in the game when it slapped him around, and later when Dame Fortune seemed a bit more even-tempered. However, Paul's skill as an ASL player really determined his victory. To win at ASL, you need to plan a couple of moves ahead, while still adapting to unexpected twists of fate. That is almost the definition of "the management of violence," the essence of military strategy in the 3-D, non-cardboard world.

I prefer games that challenge me in that fashion over the more deterministic games like Starcraft 2. Not everyone shares that preference. Managing risk and opportunity defines my kind of game, from ASL to Cosmic Encounter. Determining risk and opportunity in a game like Tammany Hall (a very deterministic design) just isn't as fun for me. I'm also not a big puzzle solver, other than crossword puzzles. The two preferences are definitely connected.

Randomness in boardgames: dramatic effect

[First in a series of posts about random elements in boardgames.]

One of the virtues of Boardgame Geek as a site stuffed with "user-generated content" is the insight it provides into those users. From reading that BGG content, it's clear that boardgamers have widely diverging attitudes about randomness in games. To some extent, these attitudes are a question of taste. But beyond the cursory de gustibus disclaimer ("This game is just not for me..."), people hold strong opinions on this subject.

For some boardgamers, it's an anathema, so that the enjoyment of a game is in inverse proportion to the influence of random elements. For others, randomness adds to the gaming experience.Unlike the other faction of gamers, this cohort doesn't have quite as simple a formula for calculating the effect of randomness on games. Enjoyment may not be directly proportional to randomness, but below a particular threshold of randomness, the game loses its attraction.

Take, for example, the fans of the games Automobile and Dungeonquest. Automobile has no random elements, other than the decisions other players make. While other players can and do mess with you, the game rewards some level of planning, particular within a single turn. If I build distributors, you may interpret that move as a threat, and respond accordingly. Taking these measures and countermeasures into account, I plan my turn accordingly, timing actions (building distributors, opening factories, building cars, etc.) based on both what I want to do, and how you're likely to interfere with me.

Dungeonquest provides the opposite gaming experience. There's little interaction among players, other than an implicit race for the exits when it looks possible to score points before the game ends. Randomness is not only greater than in Automobile, but it's almost the defining characteristic of the game. Random tile and card draws provide tension (Am I pushing it to try to steal one more treasure from the dragon?) and black humor (Not another piiiiiiiiiit!!!), the major reasons why people play the game. The game might not have as many interesting choices as, say, Paths of Glory or Twilight Imperium, but there is a strategy that rewards a player's ability to balance potential reward against definite risk.

Greg Costikyan's excellent and frequently-cited article about randomness in games does a great job of explaining why, in games like Axis and Allies that are chock full of random el

ements, strategy is just as important as in games with no randomness, such as chess or checkers. Some elements of those strategies, such as anticipating counter-measures, are the same. Others, such as pushing your luck, don't exist unless luck is a factor.

Costikyan uses the word aesthetics to describe the experience of playing the game. In the PC and console gaming worlds, first person shooters like Call of Duty have a different aesthetic than role-playing games like Dragon Age. Different aesthetics appeal to different kinds of gamers, or the same gamers when their momentary tastes steer them towards a particular kind of gaming experience.

Let's take that word aesthetics a bit farther than Costikyan did to understand the role of randomness in creating that aesthetic. Randomness is not only a tool for enhancing the generic qualities of a particular class of game, making it more forgiving to beginners, increasing the replay value, and better simulating the randomness in real world situations like warfare. Randomness is also essential for creating dramatic elements that are otherwise impossible to achieve.

The recently-published card game Space Hulk: Death Angel is a prime example. The original Space Hulk boardgame attracted legions of fans, in large part because of its aesthetic. Much like its popular culture progenitor, the movie Aliens, Space Hulk depicts a nail-biting conflict between well-armed space marines and a seemingly endless swarm of deadly aliens. The most dedicated fans have dissected the three editions of the game, weighing the qualities of Space Hulk as both a game and a dramatic narrative. On t

he first dimension of comparison, Space Hulk as a game, many players disliked the second edition for dropping the human player's option of saving command points to use during the opponent's turn. On the second dimension, Space Hulk as an experience, fans applauded the high production values of the third edition, such as the map tiles, to the extent that they enhanced the Space Hulk aesthetic of extreme peril in confined spaces.

Space Hulk succeeds because of its ability to balance skill, the product of your actions, and suspense, which depends on elements you cannot control. Creating a formation of terminator marines, carefully positioned to cover every possible axis of genestealer attacks, depends on the player's skill. Creating moments of suspense depends on guns jamming, blips turning out to be dummies, critical fusillades of small arms fire missing their targets, and other moments when your best-laid plans gang aft agley.

Suspense can't exist without the unexpected and the unplanned. When you're watching North By Northwest for the first time, your pleasure depends to a large extent on not knowing what's up with that airplane dusting dead crops. The ending of Casablanca is diminished if you know what Louis is going to say. Similarly, the great moments of Space Hulk depend on not knowing when events will go south, or what will happen when they do.

The card game version of Space Hulk, Death Angel, re-opened many of these discussions of what makes Space Hulk a great game. (At least, for some.) This review on BGG is typical of the person for whom Space Hulk, in any form, is the wrong game. Anyone who says, "The game has a mechanic where you can spend tokens to reroll attacks and defends, but it is not something to be relied upon," is largely missing the point of Space Hulk.

Imagine yourself back when you saw Aliens for the first time. Now, picture someone sitting next to you, loudly proclaiming how disappointing the movie was. "You call that small unit tactics?" he yells at the screen. "Where are your reserves? Why don't you recon a position before sending in the entire squad? And what idiot would leave the heavy weapons team that exposed?"

Of course, this person is missing the point of Aliens. Further realism in small units tactics wouldn't make Aliens a better suspense movie. In fact, too much attention to small units tactics would distract from the suspense. Knowing the outcome ruins suspense. Keeping the outcome a mystery, despite the protagonist's best preparations, is the essence of suspense, in movies or in boardgames.