Tag Archives: strategy

The Pest-sistence of Memory

Posted on by Jeremy Geist

While it’s entirely possible that what makes a game fun is 100% subjective, I’ve noticed that some things are almost universally regarded as “fun” or “good design”, and some are almost always considered “bad” or “unfun”. Having to do math beyond simple arithmetic is in the latter category, for instance; another is memorizing things, a topic that game critic Erik Twice wrote about previously. However, Erik mostly covered the “what” and not the “why” or “how” of memorization. In this article, I’ll be theorizing on why memorizing stuff generally isn’t fun and offering advice on how to make the high-level play of your game less dependent on memorization.

The Problem With Memorizing Stuff

The concept of memory being an appealing gameplay mechanic is largely a thing of the past. For example, Whist, a predecessor of contract bridge which put emphasis on being able to remember which cards have been played by which players, operated under the idea that these games were a tool for sophisticates to measure their superior intellects against each other. The only modern mainstream board games that make memorization an intentional core mechanic are Memory, which mainly appeals to very young children, and the Simon/Bop It! family of toys, which are 75% fun because of their flashing lights and funny noises.

Modern strategy games only require the player to memorize things as an unintentional part of learning the higher-level strategies. Scrabble is the classic example: A casual player can do fairly well at Scrabble by improvising with their preëxisting vocabulary, but high levels of play demand that the player memorize all of the high-value words in the dictionary so they can maximize their score.

There are a number of reasons that memory is a less fun skill to employ than things like bluffing, prediction, and planning in advance. The first is that memorization is a fairly basic concept that doesn’t challenge the more sophisticated powers of thought that even simple strategy games try to engender in players. You use whichever method you want to commit a card/quantity/etc. to memory; then you see if you did it or not. Compared to things like bluffing, which requires knowledge of your psychology, your opponents’ psychology, the risk and reward of various strategies, and so on, memorization asks you to employ a basic mental concept common to most sophisticated animals. However, despite its simplicity, memorization requires a lot of effort if there’s enough variables to be considered, so it’s taxing without being rewarding.

The second is that getting penalized because you didn’t memorize things correctly feels like an act of cruelty on the part of the game. When all of the information was available at some point, it then feels arbitrary that you aren’t allowed to reference it, especially if this lack of knowledge could lead to substantially different strategy decisions. This isn’t objectively true, as I’ll explain in more detail later in this article, but it certainly doesn’t feel good when the game essentially makes you lose for no reason.

The third is that, paradoxically, losing because you didn’t memorize something always feels like a blunder. You weren’t outsmarted by an opponent; you didn’t make a risky decision that failed to pay off; you didn’t even misunderstand the subtler interactions of the game. If you weren’t able to remember game information in an adequate way, the only explanation is that you, personally, failed, with no excuses or ways to improve beyond hoping you’ll remember better next time. Frustration – the least acceptable emotion in gaming – is born from a combination of rage and helplessness, and loss due to misremembering things produces both in great supply.

The final, and perhaps most important, factor is that discovery and surprise are important to a game being fun, and memorizing the game’s components and strategy does literally the opposite, siphoning a sense of anticipation away from the game even as you become better at it. People will only stay with a game if they think it has something more to offer them if they play it again. (That’s also why I believe that any game with asymmetric powers should have at least one and a half times more powers than the maximum player count.)

Multiple factors affect how important memorization becomes at high levels of play. The three most important are open information that becomes hidden, consistent and complex strategies, and all of a game’s components being guaranteed to appear every game (which is a mouthful, so let’s call it “Completeness”).

Open-Then-Hidden Information

Most games that restrict previously-open information don’t do it to challenge the players, and they don’t do it by mistake, either. They do it because having too much complex information available at once can be paralyzing, especially for new players or moderately experienced players who suffer from analysis paralysis. Many of these games are aimed at casual audiences. That said, even if we’re primarily designing for casual audiences, we can try to do better to make the game fun for people of all skill levels and seriousness.

An effective rule of thumb is that if components can be counted, high-level play will demand that they must be counted. In For $ale, players bid on house cards of differing values and then use those house cards to bid on cards of various values. Each player starts with the same amount of money at the beginning, but are encouraged to keep their remaining cash a secret. Because all of this information is public, you could therefore, with either a keen memory or a pad of paper, know which player has which houses, which houses haven’t been played yet, and how much money everyone has left. This is crucial information to victory, so if one player decides to do this, the others will be obligated to do so to be competitive.

In some cases, there are good reasons to not have all this information available, as having too much at hand can be paralyzing, especially for new players. Returning to our example of For $ale, if every players’ house cards were face-up and visible on the table, players would have so much information to process at once, the game would get bogged down with their thinking.

Consistent Strategy

As your players play your game repeatedly, common tactics and strategies will start to emerge. This is a natural and wonderful part of the game-playing process. However, if strategies within your game are both consistently strong and relatively nuanced and complex, it means that players interested in advancing in skill will have to memorize these strategies. Chess is the perennial example – many people, including myself, have been turned off from chess from the nightmarish vision of having to spend our precious free time memorizing openings.

These strategies become more problematic if they take longer to execute or have little room for deviance. The “Uxmal Gambit” in Tzolk’in: The Mayan Calendar is a gamelong strategy that first involves placing your workers in two specific locations on the board to acquire the maximum amount of workers as quickly as possible, then following a series of tasks that enables you to build many high point value buildings. However, if you don’t remember the strategy offhand, you’ll probably lose after creating many more workers than you can effectively feed. The designers, wise as they are, made the expansion create much more varied random openings that lessened the Uxmal Gambit’s strength in most cases.

Note that a strategy is different from a heuristic – a rule that applies in enough circumstances that you can usually take it for granted. Most heuristics are easily digestible and helpful for players. If your players can establish these heuristics without having to adhere to a lengthy set of guidelines for play, it means your strategy game is likely on the right track.

Completeness

If every component in your game is guaranteed to appear in every game, then it’s good strategy to memorize all of the components so you know which ones haven’t appeared yet.

This becomes more of an issue when there are many different components. The opening hands in 7 Wonders are always comprised of the same cards, depending on the player count. (That is, a four-player game will always have the same set of cards, and a five-player game will have an additional set of cards, etc.) Because you can be sure that every card from the list will appear in the game, good strategy now consists of remembering the card distribution and cost for all three rounds so you can tell when you’re being cut off or not. Furthermore, you have to remember the exact cards that were in each hand as it was being passed to you so you can try to predict what you’ll be handed when it comes back to you.

Fortunately, avoiding the completeness issue is relatively simple – even knocking out one or two components can be useful depending on the complexity of your game. Both Love Letter and New York Slice, which place importance on a certain class of card/pizza slice running out, remove one card and three slices respectively at the start of the game, making exact calculation of all the components you’ve seen so far not so effective that it’s worth the trouble.

Conclusion

It’s hard to design for higher-level play because playing with the same group over and over (in other words, the only way to experience higher-level play) isn’t as important as making sure your game is good across a large variety of players. But in order for your game to be truly good, it has to maintain a sense of fun and discovery, as well as make more thoughtful play more of a joy than a slog. Memorization is an enemy to both of those, and should be reduced as much as possible unless you have an extremely good reason otherwise.

 

And Now…A Post About Starcraft (Sort Of)

Posted on by Jeremy Geist

A while ago, I discovered the blog Illiteracy Has Downsides via a link tweeted by pro Magic player Matt Sperling. Despite being about a type of game  I never play and possess only passing familiarity with (real-time strategy games, primarily Starcraft II), I found a lot of their work to be clearly written and extraordinarily insightful. If you’re a game designer of any type or genre, you’ll find some of IHD’s writing useful.

One article in particular I wanted to focus on is entitled “Why Starcraft II Feels Difficult To Play”. The article primarily discusses the real and perceived skill floor to Starcraft II, but in a tangent, brings up a term called “power mechanics” that shone a light on a key difference between video games and tabletop games: Their learning curves. This article goes into how “power mechanics” smooth out the curve to video games in a way that tabletop games often have difficulty with.

Protoss: Too Smart For Their Own Good

The IHD article defines power mechanics like this:

Let’s compare Zerg and Protoss. Zerg has lots of simple, high priority tasks – injects, spreading creep, moving overlords around, and maintaining constant production. I call the most important of these “power mechanics” – basic tasks that need to be exercised constantly and deliver a measurable, substantial boost to the player each time they’re employed. Injects and spreading creep are examples of what I’d consider “power mechanics”.

Protoss doesn’t have very many power mechanics – things players can do constantly to put themselves in a better position. The game’s design instead calls for the average Protoss player to focus on more complex tasks, such as careful placement of structures, appropriate game sense and scouting of the opponent’s composition, ensuring a unit is on hold position within their simcity, etc.

Note that “mechanics” in real-time strategy games refer to game components that rely on the player’s dexterity and execution.

To broaden this definition, a power mechanic could be any system in a competitive game that relies on mechanical execution, can be improved through practice, but that doesn’t rely on high-level strategic thinking. In fighting games, a power mechanic might be executing a combo; in MOBAs, it might be managing the flow of creeps into your lane.

The main benefit of power mechanics in video games is it gives the player something to improve that produces consistent, visible results. If I practice combos in a fighting game for a week, at the end of that week I will most likely be able to execute that combo better and feel happy about my improvement. Improving my overall strategy and ability to read my opponent is something that is much more difficult and subtler to improve, so a game with no power mechanics is one with a very steep learning curve.

As opposed to video games, where almost all have at least some kind of power mechanic as part of their gameplay, board games essentially have none. This has a lot to do with the real-time nature of most competitive video games; it’s easier to make mechanical execution challenging when you can control the necessary timing of the motion using a computer mediator. Board games are also more accessibility-concerned in that a player not being able to properly manipulate components is seen as unacceptable on the part of the game rather than a sign of the player needing to get good.

While many people, myself included, enjoy not having to repeatedly practice mechanical execution to get better at a game, it does have the downside of making improvement more difficult and less rewarding. This is a bit abstract, so let’s compare a board game and a video game to demonstrate the difference.

Doing Reps in the Lab With Reiner Knizia

The board game is Modern Art, which I chose because it’s relatively freeform, and the actual strategy has to be puzzled out over repeated plays. For those who haven’t played, it’s about buying and selling art whose value changes depending on how many cards from a particular artist have been played in the round.

The video game is Street Fighter V, both because I’ve been watching a lot of Street Fighter tournaments lately and because I used to try to be good at fighting games, so I feel more qualified to discuss the genre than MOBAs or shooter games.

The improvement process in Modern Art primarily consists of learning lessons and applying them to new situations. You’re trying to get as much money as possible and deny your opponents as much money as possible, so many of these lessons have to do with predicting profit margins on different paintings. “Paintings max out at $30 each in the first round, so bidding $31 is guaranteed to lose me money” is a lesson most players learn halfway through their first game; “Try not to end the round” and “there’s 13 paintings on average for each artist in the whole game, so if 10 have shown up in the first three rounds, they can’t come in first in the fourth” might take longer. Regardless, you advance in your level of Modern Art success by learning these big lessons.

Street Fighter V has lessons as well. The broad strategy for most characters is to force your opponent into a corner and hit them with combos until you win. You learn a “neutral” game, a corner game, and ways to escape a bad situation for whichever character you like. However, half of the challenge of Street Fighter is successfully executing these strategies. It’s not just enough to have a theory of how to pressure Urien with Cammy, once you’re put into that situation you have to actually hit buttons with the right timing, complete your combo, and maximize the damage you can get out of it. 

Training your execution is fundamentally different from learning strategy. You can go into training mode and execute your combos over and over again until you start getting them right consistently, then harvest the fruits of your labor as you get these higher-damage combos against other players. While it requires more rote practice than learning strategy, you’re also rewarded quickly and frequently as your combos become more consistently reliable.

To summarize, if you look at a theoretical graph of player improvement for Modern Art, there’s a lot of sharp increases followed by plateaus where your skill remains functionally the same, while Street Fighter‘s graph is a lot smoother of an incline.

Who Cares?

So what can we, as tabletop game designers, learn from looking at power mechanics?

The first lesson is that learning is best digested a little bit at a time. Much of what I’ve written about was achieving mastery, not competence, but even learning the basics of a complex game can be better handled by rewarding the players continually and quickly. For example, Magic Maze is in the unenviable position of having a lot of content and not allowing communication between players, and it handles this by doling out the rules one at a time throughout 15 or so short tutorial games. Most importantly, all of the tutorial games feel fun to play and satisfying to complete, meaning that even as you’re not playing with everything in the game, you’re still having a good time.

The second is to balance rewarding players for discovering new tactical alleys in your game with not punishing people too much for not getting it yet. Of course the person who understands the game better should win, theoretically, but there’s a difference between “I lost but I think I did alright” with “I got absolutely clobbered and I don’t even know what happened.” Variance can help here in a way that’s inappropriate in video games, both to make people at a higher level of strategy have to improvise and to give the less skilled player a chance at victory. Since the learning curve is more punishing for tabletop games in general, make the gameplay less so.

The third is to regularly create satisfying moments in your game that makes your players feel rewarded. Although we can’t replicate the feeling of perfectly executing a power mechanic fully, we can help reach some kind of game journey that doesn’t make the player feel like every decision is fraught with stress. With the exception of extremely tight games like Agricola, it can be helpful to have moments to “breathe” where players collect resources, rally their troops, and so on, so people don’t feel like the game is too oppressive.

Conclusion

The more I critically examine the designs of games that aren’t in a box on my shelf, the more I learn about their differences. Reality TV shows have to make the audience have fun, but not the players. Subjectively judged contests like ice skating need to reward variety and creativity in addition to mechanical execution. And video games can reward players for well-practiced rote behavior in a way that’s impossible in most tabletop games. These contrasts help illustrate the strengths and weaknesses of my chosen medium and allow me to understand how to make the best of both.

Dare to be Stupid

Posted on by Jeremy Geist

I’ve written before about the difference between strategic and experiential games – that is, games that are meant to challenge players intellectually versus games that are trying to generate an emotional experience. While I separated them into a binary for the sake of writing about it, every game has some ratio of strategy to experience to them, and part of game design is deciding where to stake that ratio.

One of the most difficult parts of balancing strategic and experiential value is that sometimes, the best thing for you to do is to make the game less smart – that is, replacing difficult decision-making with something out of the players’ control. This may sound counterintuitive, but sometimes it’s what’s needed to make your game a more complete and enjoyable experience.

Destroying Your Brain Cells

There’s two questions that have to be answered here: How do you make your game dumber, and when do you do it?

“How” can be defined with something approaching empiricism. You simply have to look at the areas of your game that exist on a binary between “experiential” and “strategic”. Some examples are:

  • Hidden information: Having more complete information is more strategic, because you have a better idea of what the game state looks like. Having more hidden information is more experiential because of the exciting “moment of truth”.
  • Randomness: Having no randomness, or “input randomness” where you determine random effects before making choices, are strategically beneficial because the players have more information before making a decision. However, having more “output randomness”, where you don’t know what will happen to some extent when you take an action, means that victory or defeat can hinge on an unexpected outcome, making these moments more impactful than games where things go mostly according to plan.
  • Asymmetry: I’ve mentioned this multiple times at this point, but while asymmetric powers make it more likely that a game will be reduced to who or what you selected before it began, it also creates sides that players can identify with and remember.

“When” largely depends on your design. I share Robert Ebert’s opinion on craft in that a well-made work does the best job of expressing its vision, rather than adhering to a set standard of “good design”. Thus, it’s your judgment call as a designer for when this kind of sacrifice is necessary. What audience is your game aimed toward? What kind of holistic experience do you want to create? Do you have so much strategy it can get paralyzing? These are all useful questions to ask yourself.

The next section is a case study of a game I think took a stuffy genre and made it stupid in the best way.

Camelot: A Silly Place

Tournament at Camelot belongs to the trick-taking genre, which has its roots in classic games like bridge and hearts. The genre has a lot of stiff competition, including the two-player The Fox in the Forest, the co-op The Crew, and Nyet!, a game I’m very fond of and wrote about at length previously. However, Tournament at Camelot has done more to innovate trick-taking than any other game before it, because it’s the world’s first trick-taking game that’s stupid.

Even compared to other strategy games, trick-taking focuses a lot on the ability to analyze what cards have already been played. Classic trick-taking games use a full deck of playing cards divided evenly among four players, and a large component of the strategy involves memorizing what has been played, knowing what remains in your opponents’ hands, and formulating a strategy around this information. The core of trick-taking, essentially, is using perfect information to develop short-term tactics.

Tournament at Camelot throws all this out the window, and arguably for its benefit. You only see about half of the deck, so you have no idea what your opponents could have, and wild suits and “merlin” cards make it impossible to actually decipher your opponents’ hands fully. To add to the chaos, the game belches a half-dozen Godsend cards onto the battlefield after the first rounds, which do things ranging from making a player play with their hand revealed to stealing someone’s character power. It is a horrendous, hilarious mess, less of a calculated battle of wits and more four knights punching each other bare-handed in a pit full of mud. Whereas other trick-taking games require you to be cold and calculating, this one allows you to get in a good laugh about how ridiculous the whole situation has become. It’s a warm, funny, human game.

Not every game should go to the lengths Tournament at Camelot did, but its decision to carve out a niche in its stolid genre by being goofy and unpredictable was an extremely good one. 

Escape From the Castillo

You might think that sacrificing strategy for excitement is limited only to lighter games, or to things like area control games with lots of minis where the experience of getting into a slugfest with your opponents is more important than something tense and balanced. However, things aren’t that clean, and many games we think of as classic Euro or strategy games have embraced stupidity to at least a small extent.

El Grande is one of the best examples. In El Grande, you can choose to place your caballeros (cubes, used to determine area majorities) into the Castillo, an opaque tower that hides the caballeros from view. At the end of every three rounds, all of the cubes in the Castillo are revealed and routed onto the board via simultaneous selection.

The fact that the cubes in the Castillo can’t be seen, despite players having to announce how many they’re putting into it at all times, is a strict negative in terms of strategy. Making players memorize every minor action in order to gain a strategic advantage is unfun and rewards rote thinking over innovative strategy. By removing the Castillo and exiling the caballeros to Strategy Island, where you can clearly see how many caballeros everyone has, it means that the players can make more informed decisions about where to put them during the scoring round.

But removing the Castillo would also destroy one of the most important moments of the game: The theatre of revealing the caballeros inside and determining who has a majority. Aside from the hardcore strategy game enthusiasts, most players of El Grande will only semi-remember additions to the Castillo, so sometimes people will under- or overestimate how many caballeros are actually hidden in there. This means that the scoring rounds, which can otherwise be bookkeeping for an outcome everyone saw coming, gain an emotional element as eight caballeros suddenly storm out of the tower to charge into the Basque Territories you thought you had total control over.

Scoring in most Euros is mostly bookkeeping, but by removing the element of strategy that having visible caballeros would have created, El Grande manages to make it one of the most exciting parts of the game.

Conclusion

Sometimes, the best version of your game is one that defies the easy definitions of a “good game”. Being able to sacrifice what people think are good features of other games in service of your vision is difficult – possibly one of the most difficult things to do in game design – but if done with good judgment and clear vision, can be the difference between an okay but forgettable game and an instant classic that wins people over. When the time comes, will you dare to be stupid?