Theory: Theme As a Mechanism for Discouraging Optimization

It’s generally understood that tournament players of card games will gravitate toward optimal decks and strategies. However, last year a fascinating situation arose in which the players of Legend of the Five Rings (“L5R”) chose not to optimize, and ultimately forced designers to alter the game around that preference. In the process L5R demonstrated that it’s possible to get players not to play the best cards and decks, if a powerful theme creates an adequate incentive to do otherwise.

By way of background, L5R is a card game which goes to great efforts to simulate life in a world inspired by mythic Japan and dynastic China. Battlefields are replete with samurai, while in palaces courtiers jockey for influence. The victory conditions are meant to capture a range of ways in which one might attain respect and power in such a setting: conquering opponents’ lands is one option, but players can also achieve dominance in court or become a religious leader. Players are encouraged to be loyal to one particular “clan,” following it like one might follow a sports team, and to represent it in tournaments. Everything about the game is designed to create a “you are there” feeling, immersing the player in the game world.

In last year’s tournament season one clan was extremely strong, putting up more than its share of victories. That led to a great deal of discussion about where the clan’s strength came from. Some argued that the clan’s cards were too good–a design flaw in the game. Others suggested that the problem lay with the players, who neglected cards that would rein that clan in.

Ultimately the game’s designers gave both sides some credit as they announced errata meant to level the playing field. They conceded that the powerful clan “ha[d] come out of the gates far too strong.” However, they also noted that players were not doing everything they could to maximize their chances of defeating the front-runner. “[P]eople are generally not preparing their decks for fighting [the powerful clan],” they said, citing cards that “are fantastic . . . yet are seeing very little play.”

It’s unusual, in my experience, for game designers to have the problem that tournament players aren’t well prepared. In the age of information, it’s usually the other way around: good strategies propagate quickly, are studied intensively, and counter-strategies then appear promptly. (Alternatively, sometimes it’s determined that no possible counter-strategies exist, and that errata are needed–but the problem in that case still is not insufficient preparation.) How is it that L5R’s tournament players bucked the trend, and were so unready that the designers had to take action?

Certainly, one contributing factor is that relatively less data comes out of L5R tournaments than those of other games. However, players who wanted to know could easily find examples of the strong clan’s best decks. While not the wealth of information that comes out of, for example, Magic: the Gathering events, the data available was enough to point out the utility of the “fantastic” cards that players didn’t use.

Some have also argued that the “fantastic” cards actually weren’t all that good, or that it was too onerous to use them, or at least that it was too onerous to use them in the numbers necessary to hold back the strong clan. L5R’s design team, however, is generally drawn from those skilled at the game. Indeed, its lead designer was once its winningest tournament player. Under those circumstances I’m inclined to hew to their opinion on how good cards are, and how realistic it is to include them in one’s deck.

If the issue wasn’t that players couldn’t find answers for the strong clan, and wasn’t that the answers didn’t exist, what was it? The answer, I think, lies in L5R’s intense focus on theme. Its players are encouraged to choose a clan that appeals to them, to pursue a victory condition that they like, and even to use or not use particular cards as a means of personalizing their experience. Thus, there are devoted players of the Scorpion Clan, players who always try to win by enlightenment and eschew military victories, and players who won’t use cards that are associated with the “Shadowlands” because those cards represent evil forces in the game’s setting. Mark Rosewater would say that it’s a very “Johnny” game. Telling these players that they have to build their deck a certain way in order to compete is often going to be futile. Players who have been hooked by the promise of immersion and even self-expression will not want to break their suspension of disbelief to do something as “game-y” as running athematic meta cards.

To be clear, I’m not criticizing those players. I’ve played L5R on and off for about two decades because I enjoy the personalized experience too. I don’t love using athematic cards in my decks any more than anyone else. My goal here is to understand why people don’t play them, not to rake anyone over the coals for passing them up.

Nor is it my intent to criticize L5R’s designers. It was completely rational to expect that tournament players would optimize in the pursuit of victory. Their decision not to was unpredictable to say the least.

Unpredictable, but interesting. Players’ refusal to change their decks to react to the tournament environment was a problem for L5R, but a superb lesson for game design generally. Even in a tournament setting, L5R showed us, it is possible to get players to forego advantages and play sub-optimal strategies. The competing incentives provided by theme can outweigh the desire to win.

A Question for CCG/LCG Players

Last time I had lots of theoretical questions. This time I have a single, completely practical one.

My understanding has long been that the only acceptable way to shuffle at a CCG tournament is the riffle shuffle. Pile shuffling doesn’t randomize the deck, so it’s useful for checking to make sure a deck has the correct number of cards but is otherwise unhelpful. Side-shuffling (a.k.a. slide-shuffling, mash-shuffling, etc.) allows sleeved cards to stick together and stay together, knifing through the unstuck cards and coming out of the process still grouped. Riffle shuffling randomizes the deck and breaks up clumps, so–so far as I knew–it was the best, and indeed only really acceptable, option.

As a result, I was surprised by this video, which I saw linked on ChannelFireball:

The video demonstrates a side-shuffling technique wherein the cards are held off to one side, so that the player shuffling cannot possibly see the cards.

I understand the value in making sure that people aren’t sneaking peeks while shuffling, but I would be much more concerned about the possibility of an insufficiently randomized deck than I would be about my opponent potentially seeing a card, even an important card. As a result, I feel like I would much rather the opponent be riffle shuffling.

Has the received wisdom on this topic changed?

Theory: Eras of Game Design?

Are principles of good game design timeless?

In the last post we talked about how the Babylon 5 CCG was a lot of fun, even though it did a lot of things differently from how most games work today. It eschewed elegance in favor of a baroque ruleset, and the game’s cards are text-heavy, more so than is usual in current card games. Was the game fun in spite of its diversions from now-accepted design principles, or was it complying with the standards of a different era?

I’ve been turning that issue over in my mind, and I’ve only ended up with more questions:

1. Has technology created a new era? Imbalances in a game are much more likely to be detected, and optimal solutions are arrived at much more quickly, now that players worldwide can pool data and compare notes. Problems with a game’s design that might never have needed to be addressed in the past can show up very quickly in the internet age. We have seen this dynamic at play with Magic: the Gathering, which actually reduced the amount of information coming out of online tournaments because it was becoming too easy to home in on the best decks.

Yet, it’s always been possible to solve games, or at least to find optimal strategies. The Russian Campaign, a classic Avalon Hill wargame, cheerfully provides the optimal opening positions for the Russians in its rulebook. Players of Starfleet Battles were certain that getting the alpha strike was critical until someone showed that precisely managing one’s weapons to maximize damage over several turns led to better results. Did the internet change the situation for designers, or just contribute to one that always existed?

2. Where is the necessary information? Some game design principles might be contingent on historical factors outside the game rules. For example, I think it’s broadly agreed today that cards should have as much information as possible along the sides, so that the information is visible when the cards are held in a fan. That’s only a rule, however, because most people hold cards that way. Were cards held differently in other times and places? For example, cards might be held vertically so that the tops are visible instead of the sides, leading to a different standard for how cards should be laid out. Where would one find that out? How would one even know to look for it?

3. How universally should the rules be stated? Suppose the rule for cards was not “put information on the sides,” but rather, “put information where it will be visible when the cards are held naturally.” At that point the design rule becomes flexible, able to accommodate regional and temporal variations in how people organize cards in their hands. What design principles can be put in such broad terms? Where they can be, should they be? Or are the idiosyncrasies of each time and place part of what denotes eras of game design?

4. Which rules of design can define eras? “The game should involve interesting decisions” has probably been an important standard for many games throughout history. If there was a time when everyone agreed that games should be boring, it seems like that would represent a distinct era, and we might have to evaluate games from it very differently. What, though, about the information-on-the-sides-of-cards rule? Is that important enough that a change in it would represent a distinct period in game design history?

5. What work has already been done in this area? At times I feel the limits on my knowledge of the academic work in game design keenly. This is one of those times; I’m sure I’m not the first person to think about this, and I wonder what conclusions others have reached.

Theory: Drawbacks of Rotating Metagames

A rotating metagame—the situation where every pre-game choice a player might make can be countered by another pre-game choice, so that none of the choices become dominant—is a commonly-cited tool for balancing games with lots of moving parts. However, it is not a panacea. Implementing a rotating metagame creates some new problems, and puts certain pressures on the game’s overall design.

“Rotating metagame” is a term of art, and like all terms of art it might sound opaque. However, it’s simple enough in practice. Think of Magic: the Gathering, with its many cards and decks. Each card has a counter, something that destroys it or negates its effectiveness, and through these individual counters whole decks can be countered. If one card or deck starts to become prominent in the competitive scene, people will play the relevant counters and that card or deck will be taken down a few notches. Then as the counter becomes strong people will start to counter it, and so on and so forth. Through this process tournament play begins to look like a wheel: cards and decks move to the top and then get pushed to the bottom, at which point their counters lose popularity and they start to rise again. Hence, the strategic situation rotates.

A rotating metagame creates game balance in the sense that there is no dominant, unbeatable strategy. However, it is not necessarily desirable for every game, or even for many games. The technique has serious limitations.

First, a rotating metagame offers a good environment, not good individual games. In a rotating metagame one accepts that blowouts can occur when someone is caught on the wrong side of the rotation. When viewed as a whole and over time the tournament scene will look healthy, but the zoomed-in experience of the individual player might be very poor.

This problem is exacerbated by the fact that bad games will be concentrated among new and casual players. Players with little experience or who are less informed about the game are the most likely to be rolled over by rotation’s wheel, because they often will not realize that they need to learn about the current state of the metagame. Deeply committed competitive players, by contrast, will know exactly what they should be playing.

Unfortunately, this distribution puts unfun games in exactly the wrong place. Casual players might quit after an unfun game or two, and new players are especially likely to pass on a game after a single bad experience. These are the groups who should be getting protection from blowouts, but a rotating metagame instead makes them grist for the mill.

A further challenge for rotating metagames is that everything hinges on the ability to rotate. Rotating metagames work when they are in a state of dynamic imbalance; something is always on top, but that something changes frequently. If the changes stop, all that’s left is a game with a dominant strategy.

Ensuring that the wheel of the metagame keeps rotating, then, becomes extremely important—and that imposes several design requirements. First, the resources for rotation must always be available. No strategy can be without its counter, and those counters must appear in a reasonable percentage of players’ toolboxes so that they can do their work.

An interesting example of what happens when this isn’t the case can be found in the NBA. Without getting into the minutiae of basketball’s rules for defensive play, it used to be the case that defenders had to set themselves up against specific offensive players and follow them around the court, rather than staying in “zones” and defending against anyone who came near. As a result, the game began to trend toward a dominant offensive strategy of passing the ball to a single unstoppable player—the gigantic Shaquille O’Neal, the too-quick Allen Iverson, or someone else with an enormous physical advantage—and then having all the other offensive players just get out of the way. Defenders were required to follow the irrelevant players to irrelevant places, and then the hard-to-stop player would beat the lone defender permitted to be anywhere near him and score. Since the resources required to stop the unstoppable player weren’t available—there just aren’t many people Shaquille O’Neal’s size—the game could not rotate away from this lone wolf approach, and basketball strategy began to stagnate. Ultimately the league had to change the rules to allow zone defenses in order to break the deadlock.

Additionally, rotating metagames only work when the barriers to change are relatively low. Shifts in Magic’s metagame are painless because those who are keeping up with tournament play probably have all the cards they need for the change and emotional investment in any given deck is relatively low. By contrast, a game where changing strategies means a massive new investment (e.g., miniatures games where starting a new army involves a great deal of money and time spent at the modeling table) can’t rely on players keeping up with a changing strategic environment. Similarly, games where players have reasons to stick a single deck/army/team through thick and thin will find that the metagame doesn’t rotate. Fans of the Philadelphia Eagles who want to see the Dallas Cowboys defeated can’t transfer their loyalty to another team with a better chance of taking the Cowboys down.

Thus, implementing a rotating metagame means designing and marketing in ways that keep those barriers low. Important counter-cards can’t be at too high a rarity in Magic, even if limited play would benefit thereby, because they have to be broadly available. Team loyalty can’t be too big a part of how the game is sold, lest it stop players from shifting gears when they necessary. Every aspect of the game has to be looked at with an eye toward, not just how it impacts strategy, but whether it could have an ossifying effect on players looking to change strategies.

Finally, rotating metagames are a major design challenge. It isn’t easy to maintain a web of counters and counters-to-counters, all good enough to dethrone a dominant strategy but not so good as to make the countered strategy unplayable. Fine judgment about how strong each strategy is going to be and how effective to make the related counters is vital. Attaining that judgment requires enormous amounts of quality playtest data, which is not always easy to come by.

Given all of these weaknesses, when is a rotating metagame appropriate? The short answer is “when there are too many things happening to balance all of them at once.” Magic has a functionally infinite number of possible decks giving rise to a tremendous wealth of strategies; it’s impossible to arrange for each to have an even game against all of the others, so the rotating metagame serves as a safety valve that gives as many of them as possible a chance to shine. The many benefits Magic gains from its diversity of options outweigh the drawbacks of the rotating metagame they necessitate.

In the end, a rotating metagame is a tool. Like all tools, it places certain demands on its user and can be harmful if employed thoughtlessly or in the wrong situation. Don’t just assume that it’s right for other games because it’s been used successfully in the past; instead, think critically about what it will accomplish, what it will cost, and whether the former is worth the latter.

Something Completely Different: Alternate Mana in Magic

I was going to put up a discussion about how Rock Band succeeds in being fun even when the players are losing, but then I saw the #AlternateMana posts on Twitter and got inspired. Changing the way players get mana–the resource required to play cards–in Magic: the Gathering messes with the fundamental building blocks of the game. Pushing that to an extreme could end one up with a game that still has cards and mana costs and timing rules and all the other elements of Magic, but that’s nevertheless a very different experience.

How about some of these:

Mana is acquired by building a house of cards. The different colors of mana each have a different size and shape of card associated with them, which make some combinations easier and some more difficult (e.g., the red cards and the blue cards are shaped such that they’re stable when used separately, but do a poor job of reinforcing each other). Getting more mana requires building the house higher.

Mana is produced by the overall amount of Magic in the area. The more Magic is being played, the more total mana is available. Some cards’ costs can only be paid at large events; PTQs and GPs aren’t just noteworthy because of the players and the prizes, but because they’re big enough to allow Griselbrand Unleashed to hit the table.

Mana is allocated by a group, which may or may not be made up of people playing in the same game. At the start of each turn, players explain what they want to do and what they need to achieve it. The group then divides the mana up according to whose speech impressed them more. (Imagine how different Commander would be if you had to get people to give you mana by explaining why your deck’s gameplan is fun for the whole table.)

Mana comes from real-world locations. Traveling to a new place and playing Magic there permanently gives the player access to that location’s mana. Get more by further “attuning” to that location: sightsee, become proficient in the local language, etc.

Mana is captured in wargame fashion; it comes from spaces on a board, and players gain mana by taking and holding those spaces.

Mana is a flow, represented by flowing water on the table. Players gain mana by using their cards to divert the flow. (Sleeving cards suddenly becomes very important.)

Mana is acquired through a music equalizer, with sound in different ranges generating different kinds of mana. Players get the mana they need by finding (or playing?) a song that quite literally hits the right notes.

Mana is generated by emotion; to get a certain color of mana, a player must find evidence of a specific emotion in the world via news stories. To get more mana, the player needs to get better at searching up information. Bonus mana comes from finding it in other languages, from different countries, etc. The metagame is influenced, not just by the card pool, but also by the state of the real world.

Now I really want to design games that involve building houses of cards and redirecting water. If only there was a 25th hour in the day . . . .

Theory: Why Players Concede–and Why Their Opponents Hate It

Why people do people concede–or why, on the flip side, they object to others doing it? Answering those questions will provide us with guideposts for addressing concessions as a design problem. If we know what leads players to concede, we can try to avoid those situations; if we know what makes concessions objectionable, we can design the game to make them less so.

The definition from last time will guide the discussion here. We aren’t going to talk about concessions that are explicitly a form of cheating, because those raise different issues and need to be addressed separately. We will, however, call on forms of concession that feel bad or are unsporting. Minimizing bad feeling among players, after all, is an important design objective.

So, what leads people quit?

The rest of the game doesn’t matter: A player decides that the result is inevitable and that there’s no point in continuing to play. I’m pretty sure that this is the most common reason for conceding, and the least likely to be found objectionable (but some people still hate it; see below). It’s seen in games ranging from chess to Magic to Little League baseball games decided under a mercy rule.

It’s important to recognize that there can be reason to continue even if a player is sure to lose. In card games played over multiple hands, for example, it can be useful to stay in just to get as many points (or as much money) out of a losing hand as possible, so as to stay in the overall running. One also sees this frequently in car racing; even if a driver is sure to finish behind the leader, it’s worth finishing the race to accumulate points toward the overall championship. When a player concedes because the game is no longer meaningful, it’s a statement that there’s no substantial incentive even to play the game out.

The game has stopped being fun: Fun is always tricky to quantify, but there’s no denying when it’s not there–and its lack makes players walk away from games. Often a lack of fun is tied to the rest of the game not mattering; if it doesn’t matter what one does then the game’s decisions probably aren’t very interesting anymore. However, a game can stop being fun for other reasons as well. Perhaps the decisions were interesting once, but the game has gone on too long and the player wants to be done. Maybe there’s a situation the rules don’t handle well, and it’s led to an argument that sucked the joy out of the experience. The opponent might simply be a jerk who’s not worth tolerating any more.

Conceding leads to long-term advantage: Some tournaments are designed in such a way that losing a match has the ironic effect of increasing one’s odds of winning the whole event. The 2012 Olympic badminton debacle grew out of this; the gold medal favorites, among others, threw their first-round games so that they would face weaker teams later in the tournament. While those players were caught (it was hardly difficult) and ejected, sometimes players can manipulate the tournament.

A variation on this theme is the concession that protects what someone already has. One sees this in Magic all the time, as players at the bottom edge of the cutoff for prizes agree to draw their match rather than risk a loss that would push them lower in the standings. Television game shows feature this brand of concession as well, with players deciding to stop with what they have rather than risk it all on one more question.

Stated by themselves, all of those sound like fair reasons to walk away from a game. A bit underhanded, in some cases, but logical. Why, then, do people dislike concessions?

Taking away the climactic move: Some players don’t just want to figure out how to win, they want to actually do it. Concessions deny these players the final moment in which they knock over the opponent’s king or otherwise demonstrate their victory. This is especially galling when the victory was very hard-won and involved a brilliant final sequence of plays; cutting such a game short can feel anti-climactic.

Reducing the time spent playing: I see this reason cited frequently by players who don’t get to play their preferred game often, or who have traveled a long way to play in an event. These players want to savor every moment of their games. Conceding necessarily denies them some of those moments. It doesn’t matter that a concession means they win; these players value time spent playing more than the victory.

Others are affected: Conceding can impact others in the tournament, as it did in the Warmachine event that inspired these articles. When Adam concedes to Beth, it can affect Charlie’s strength of schedule (he played Adam earlier, and will place higher the better Adam does), or Dani’s odds of winning (she has a good matchup against Adam, but a poor one against Beth). It can even, in unusual cases, have more direct effects. The Warmachine tournament’s result was controversial in part because the player who conceded was ineligible for the grand prize; had he played his game out and ended up in the finals, his opponent would automatically have gotten the big-ticket stuff instead of, as actually happened, going home with second place.

The list of people who can be impacted extends beyond the players. Consider professional sports: paying fans would be livid if a hockey team decided that the game wasn’t worth bothering with and left the ice at the end of the first period. Among the complaints leveled against the ejected Olympic badminton players was that they had wasted ticket-holders’ money. Shoeless Joe Jackson had to wonder what the kid who asked him to “say it ain’t so, Joe” learned about sportsmanship from his decision to tank World Series games.

It’s unsportsmanlike: For some players, trying one’s hardest is integral to honest gameplay. Choosing not to pursue victory with all one’s strength is just inherently wrong under this view, regardless of why one might do it. By joining a game, they feel, one commits to try to win it until the very end. The circumstances have no bearing on this moral obligation.

(As a side note: listing these, I feel, helps make clear why discussions about concessions so often involve people talking past each other. The reasons to concede are all about the game’s obligation to the player: when the game stops making play worthwhile, they posit, the player does not need to continue. By contrast, the arguments against conceding are about the player’s obligations to others: they want the player to keep going, even given that the activity is voluntary and no longer rewarding, because doing so benefits those others. Since the two sides value entirely different things, it’s hard for them even to engage with the opponent’s arguments.)

Looking these over, I think that they represent a fairly comprehensive statement of why concessions happen, and why some players would prefer that they didn’t. Next time we’ll try to put this knowledge into practice, discussing how games can be designed to minimize the impact of a player conceding.

Theory: Taking Mark Rosewater Out of Context

On Monday Mark Rosewater posted his annual “State of Design” article, in which he reviews Magic: the Gathering’s successes and failures for the past year. It’s an interesting read for any Magic player, but as a designer what I think is most fascinating about it are the lessons that could apply to any game. The article has design rules that are still powerful when divorced from their context.

Take, for example, Mr. Rosewater’s conclusion that “[f]lavor is key.” He explains that Magic used the same mechanic (that is to say, a thing cards did) twice: once with a flavorless label, and the second time with a name that evoked ancient Greek mythology. The mechanic was much better received the second time around, in part because players understood what the mechanic represented in the fiction of the game world and got more excited about playing cards with the mechanic as a result. Accessing “chroma” sounded technical and boring, but showing “devotion” and being rewarded for it was fun–even though in both cases players were doing largely the same thing!

Reading Mr. Rosewater’s comments immediately put me in the mind of Over the Next Dune’s rules for keeping searchers on the map. When explained step-by-step, people often find them rather opaque. Say that searchers “bounce like a screen saver,” however, and everyone understands instantly. Picking the right context helps players understand the rules enormously–so much so that I’ve considered switching to a Tron-esque theme just to be able to make the screen savor metaphor more explicit.

Other lessons presented in the article are similar. His self-critique of Magic’s execution of an “enchantment block” is interesting for anyone considering a game with expansions. The discussion on rescuing a failed idea has something to say about every game where the designer’s options are limited. More generally, the fact that Mr. Rosewater criticizes his own work despite the fact that this year saw “the best-selling Magic set of all time” sets a good example.

Mr. Rosewater is a controversial figure; opinions differ on whether he’s saving Magic or smashing it. Whatever one’s opinion of his “New World Order,” however, there can be no denying that he’s learned game design in an environment where sales numbers provide quantitative feedback, with his job staked on his continued success. Hard-won experience like is is always worth considering, and the lessons he has to teach are general enough at the macro level to make figuring out how to apply them to other games time well-spent.

Theory: What to Do If Your Game’s Not Random

Riffing on the idea that the promise of a rare lucky moment is attractive, I thought it might be interesting to look at the fate of VS System–a game where people didn’t get those lucky moments. I should say up front that I only have this story second-hand, but even if it’s wrong it’s still interesting as a thought experiment. It’s a cautionary tale about a game that ended up being unfriendly to new players, and a chance to learn lessons about what to do if your game doesn’t randomly offer new players help.

VS System was a superhero card game. If you wanted to play something Magic-ish, but instead of knights fighting dragons you wanted Captain America fighting Doctor Doom, this was your game. Its crowning achievement may have been its substantial tournament popularity; lots of people got very good at VS System, played it very seriously, and won quite a bit of money in the process.

Part of what made VS System so popular among tournament players was that, even though it was a card game, luck actually had very little to do with the outcome. Better players almost always beat worse players. For heavily-invested players who had spent a lot of time practicing, that was a valuable feature. Their hard work was consistently rewarded.

New players, however, found VS System quite frustrating. Since everyone was better than them, they lost almost every time. They couldn’t hope for a rare lucky moment; no such moment was coming.

Losing is, of course, part of getting better. I have vivid memories of going to the local arcade in Japan once a week and getting trounced for a full year before I managed to win. However, at that point in my life I was dedicated to improving at fighting games. It’s not unreasonable for people to try a game that they’re only casually interested in, get clobbered, and decide that this particular mountain isn’t one they’re interested in climbing.

That was the decision people often made about VS System. Although a dedicated group of old-timers kept the game going, fewer and fewer new players appeared to reinforce their ranks. Tournament players who moved on were not replaced, and ultimately the game folded for lack of sales.

Compare VS System’s experience to Magic: the Gathering. In Magic, the better player wins . . . most of the time. The fall of the cards can give the worse player a chance, creating the openings they need to win. When it happens it makes the better player crazy–but even Mark Rosewater believes the game is better because those moments can happen, and the sales data support him. Magic lets new players get the lucky moment they need once every so often, and as a result it can appeal to a much larger player base.

So, judicious use of randomness can help new players win, and in the process can contribute to keeping a game popular. What if a game has no randomness? Chess has no dice, no cards, nothing that’s outside the players’ control (and I depart from Mr. Rosewater’s view that opening moves are random; a decision made with incomplete information is not random, it is simply risky). Yet, Chess has lasted for a very long time. Diplomacy has a less impressive pedigree, but it’s also a popular game of long standing with no random elements. How do they keep going?

I think there are two key things that non-random games can do to avoid frustrating new players:

Ranked play: Chess’ ELO system helps players find others at the same skill level. Weaker players can avoid being clobbered by playing people at the same ELO–or can opt into a challenge by finding someone at a higher level. Players control how much frustration they experience.

Make it hard to lose quickly: Even an extremely bad Diplomacy player can’t be eliminated in the first few turns. (The writing, admittedly, might be on the wall by then.) Furthermore, the early turns are often when most of the wheeling and dealing at the heart of Diplomacy occurs, so even the newest player gets to experience what the game has to offer and feel like he or she got to participate meaningfully.

I’m sure there are more approaches to this problem, but I feel that these are particularly effective. If your game doesn’t enjoy the luxury of offering rare lucky moments to give new players hope, doing one or both of these will help ensure that they still have a positive experience. That, in turn, will contribute greatly to the long-term health of your game.

Theory: Balance Matters for All Skill Levels

The common wisdom about game balance–that it only matters for top-level players–is incorrect. It is true that balance is more important to the outcome in matches between highly skilled players than it is when newer players compete. However, balance has a far greater impact on fun for weaker players than it does for stronger ones. Both groups benefit from balance and are hurt by its absence.

“Balance” is a tricky word in game design. It sweeps in a lot of issues and discussions about different types of games that arguably shouldn’t be directly compared. Here, I’m talking about balance in initial choices: between characters in Street Fighter, champions in League of Legends, interstellar empires in Twilight Imperium, and other situations where players select a set of capabilities before the conflict begins.

Players of these games often argue that balance is only important at high levels of play. The argument goes something like this: in games between low- to mid-skill players, the difference in ability between the players decides who wins. Slight advantages in one’s choice of character/champion/empire are swamped by relative skill. It’s only when both players are quite good that those slight advantages matter.

The flaw in that position is that it assumes balance only affects winning. It also plays a role in determining how much fun the players have. For experienced players the role is smaller. With new and less skilled players, however, balance can be the single determining factor in whether or not they enjoy the game.

Top players, in my experience, derive most of their fun from developing mastery. They like exploring the game, understanding it, practicing it, and demonstrating the skill they gain thereby. Whether they do that with this character or that empire doesn’t matter as much as the play and the results.

I can’t think of a better example of this than Mike Flores’ view of Caw-Blade. For those who don’t play Magic: the Gathering, Caw-Blade was early 2011’s dominant tournament deck. “Dominant” can’t be emphasized enough; Caw-Blade won again and again and again, sweeping all competition before it. In its day Caw-Blade was the only reasonable choice for what to play in a tournament.

Mike Flores, a well-known Magic player with a history of tournament success and writer of many influential articles, loved the Caw-Blade environment. He conceded that Caw-Blade was by far the best deck–but, he pointed out, Caw-Blade vs. Caw-Blade games were extremely skill-intensive and rewarded good play. It didn’t matter to him that there was only one valid deck, because that deck enabled players to show their stuff.

Newer and less practiced players, however, often have neither the experience nor the mindset to mitigate balance issues. They don’t know what the good choices are, and if they find out may not feel able to switch to them. As a result, these players can have frustrating experiences when they encounter high-level play.

This dynamic played out very clearly in the old Star Wars miniatures game. If a player did not have a plan for the “Black-and-Blue” strategy, or really wanted to play the Mandalorians even though they were weak, he or she could lose games in rock-paper-scissors fashion even against an opponent of equal skill. High-level players, and those aspiring to that status, took note of the imbalances and moved on; others just got aggravated.

To be fair, the differences in power between SWM pieces was stark. Games with smaller imbalances are less likely to produce these joyless situations. Even small imbalances, however, can build into commanding leads over time–especially in casual games between friends, where everyone involved is a repeat player and streaks are likely to be noticed.

Balance affects which top-level player wins. However, it can also affect which lower-level player has fun. Thus, balance shouldn’t be seen as irrelevant to new players and the lower ranks on the ladder. It’s important to these groups in different ways, but it’s important to all of them.

The Case Study & Theory: Gates

Thinking about how to add on to Over the Next Dune raises the question of whether and how to gate player powers. Of course, that begs the question of what a “gate” is. 😉 To avoid definitional confusion, let’s hammer that out.

A gate is something that controls a player’s access to in-game capabilities. The classic example is mana, as seen in League of Legends or the Final Fantasy games. A player uses up mana each time he or she employs a special ability, and when the mana is gone the player cannot use special abilities until it recharges. Ammunition is also a gate; it limits how much the player can use a certain weapon before having to switch or seek out more ammo.

Gates do not have to be numbers. Many role-playing games, for example, control players’ power via progress through the storyline. As the player explores new areas, meets new people, and learns new things, the player gets new capabilities.

Gates can go one way or bi-directional. One-way gates result in permanent changes. For example, in Burnout Paradise access to new cars is generally gated by completing races. Once you complete the race associated with a car, you have access to that car forever. Mana and ammunition are usually bi-directional gates; you can run out and lose access to a power, but replenishing the resource takes you back through the gate and enables you to use it again.

I believe that that’s a reasonably complete discussion of what gates are. They also have some properties that aren’t definitional but that I feel are worth putting forward:

Gates can be thought of in either direction. This is kind of a weird one, and it’s usually not relevant, but it can be useful. All gates can be described as having something or not having the opposite. For example, in Battletech firing weapons builds up heat. You can think of heat as the gate (too much is bad) or coolness as the gate (not enough is bad). It doesn’t matter, from a theoretical perspective, which approach you take.

Admittedly, this can get kind of silly. You could say that “lack of mana” is the gate, and that a player can use a certain ability because his or her lack of mana has been kept below a certain threshold. It’s a lot easier, though, to say that the player has enough mana.

Basically, this is like flipping an equation to put the variable you’re solving for on the left. It doesn’t really change anything, but if you’re accustomed to a certain presentation it might help you understand what’s going on.

Out-of-game gates are ineffective. Experience has shown that players cannot be limited by resources outside the rules of the game. Money and physical difficulty are two examples of out-of-game gates which have been proven not to work.

Money. If your game is popular, you will have a subset of players who will spend whatever they need to to get a competitive advantage. Magic: the Gathering was originally designed to use card rarity as a gate, on the thinking that players would be limited by their collections. Over time it became clear that tournament players assembled complete collections regardless of the cost. Magic still uses rarity for various design purposes, but not to balance constructed-deck tournament play.

Physical difficulty. It does not matter how difficult a physical task is; if it will help players win, some of them will put in the necessary time to be able to do it reliably. Fighting games often use precise timing as a gate, demanding that players time their moves to 1/60th of a second in order to get the longest combos and the most damage. Many, many players have practiced enough to hit those 1/60th of a second windows routinely.

From here we need to think about whether OtND should use gates at all. I’ll get into that next time.