Designing Strategy: A Conclusion

by Shannon Appelcline

March 27, 2003 – I've been talking about strategy games — a fairly narrow subset of what we do here at Skotos — for about a bazillion years now. I'm about done. Fortunately, that also coincides with me having discussed all of the topics that I originally wanted to cover (and then one).

However, with 32,000 words on strategy now writ, before I close up this first venture into strategic worlds, I decided that it might be useful to spend a bit of time talking about what I've been talking about. So, here in summary, are my thoughts on strategy, piece by piece. Along the way I'm probably going to highlight a couple of things I missed too.

The Components

Any strategic game needs to start off with components — the physical artifacts that make up the game's material existence. In general, there seem to be three types of components:

1. Environment. The largely unchanging component which all of the others interact with. Also: board or map. The main types of environment are none, scoreboard, abstract, and representative; any of these environments can also be randomized or evolving.
2. Tokens. The components that are moved about the environment. The main types of tokens are: personal, representative, randomized (dice), and arbitrary (cards).
3. Markers. Components used to keep track of something within the game. The main types of tokens are: scoring, trading, power, and modifier.

The main thing I want to add to my original article on components is the idea that different games can have different rules for who controls the components. Most frequently they're personal (e.g., your metal car in Monopoly), but they can also be shared (which is to say controlled by a group of people), or else global (e.g., anyone can move the robber in The Settlers of Catan or the Democratic Monster in Wizwar).

I've also considered that my token/marker distinction might not be the cleanest. Perhaps more intuitive is splitting tokens by whether they interact with the environment or don't. Most markers fall into the latter category, but so do cards, as I realized when writing about Hidden Information.

In any case, any categorization is arbitrary to some extent, as I written elsewhere.

My original article on components was Trials, Triumphs & Trivialities #102.

Hidden Information

Although I covered it last in my series, hidden information is actually related directly to my first topic, components, because it answers the question: what pieces can be see by which players?

I broke this into two fairly broad questions. The first was, what pieces do you hide? Common answers are:

1. Hidden Arbitrary Tokens. The values of cards are often hidden.
2. Other Hidden Tokens. Pieces usually associated with the environment (e.g., your battleships) can be hidden too.
3. Hidden Token Values. More frequently you hide only the value of tokens, not their existence.
4. Hidden Victory Markers. Many games hide some part of your score.
5. Other Hidden Markers. Dollars or other "tick markers" can be hidden.
6. Hidden Environment. In rare cases, a gameboard can be made partially invisible.
7. Other Hidden Information. Finally you can have hidden information core to winning a game, such as in Mastermind Or Clue(do).

The second question was, how is the information revealed.

As noted in the original article, you can, if you wish, totally reveal some part of your hidden information by usage, forced revelation, or at end game. However more interesting is a use of hidden information that hints at values without revealing them. Typical ways to do this include bidding, betting, peer-to-peer trading, and requesting.

Given that I just wrote the article on hidden information, I don't have much more to say on it.

My original article on hidden information was Trials, Triumphs & Trivialities #112.

The Gameplay

Once you have components, you then need to write rules for how all those components interact. This is what gameplay essentially is. It falls into two categories: activity, which defines component interactions during the game; and victory, which defines which components you need to acquire to win.

When I wrote about gameplay I divided it into six general activities and six related victories. They were:

1. Token Interaction Games. Related to Token Acquisition or Destruction Victory.
2. Environment Interaction Games. Related to Environment Acquisition or Destruction Victory.
3. Marker Collection Games. Related to Marker Acquisition or Destruction Victory.
4. Movement Games. Related to Best Racer Victory.
5. Exploration Games. Related to Best Explorer Victory.
6. Building Games. Related to Best Builder.

As noted at the time, most games have multiple activities at different levels: core, important, and minor. Most games only have one victory, and it's usually, but not always, related to the core activity.

I'm still not convinced if these categorizations are the best. They're fairly intuitive and easy to understand, and they do seem to cover most games out there, but I think that relating them more tightly to the components might have helped pull everything together and allowed for a more rigorous examination of what types of gameplay are really available.

My original article on gameplay was Trials, Triumphs & Trivialities #103.

Gameplay is the heart of any strategy game, and thus it's not a surprise that the next two topics related to it directly.

Activity & Timing

In order for your gameplay to work, you need to have rigorous rules about what can be done when: timing. There can be two types of timing: when something happens during the game; and when it happens during your turn.

At the level of game timing you need to be aware of distinctive phases within your overall game. Generally, they are:

1. Setup. The phase when you prepare your components, and the optimum place for weird or complex rules. Types of setup include: minimal, fixed, randomized, chaotic.
2. Activity
   1. Early Game. The first part of the game, a phase that you need to ensure is interesting and also remains important to the rest of the game.
   2. Mid Game. The heart of your game, and probably the bit that you've planned for the most. Here you should also think about how your game can be evolutionary to maintain interest.
   3. End Game. A topic for additional discussion (below).
3. Victory. A topic for additional discussion (below).
4. Cleanup. If you're clever, this phase is no work because your game has sorted itself during play; if not, you should at least make your cleanup simple.

However, game timing is only part of the issue because you also need to think about turn timing, which is how the players take individual actions. Generally there seem to be four types of turn timing that you can choose between:

1. Ordered. Actions take place one at a time.
2. Simultaneous. Actions all occur at the same time.
3. Real-Time. Actions occur as they're input.
4. Slow Real-Time. Actions occur as they're input, but with long delays based upon in-game realities.

Ordered turns are most common in tabletop strategies games, and can be expanded upon quite a bit; for example each player might take a full turn in turn, or alternatively each player might take singular actions in turn. To my previous discussions I'd add one new point: make sure that individual player turns are fairly short, so that everyone remains interested. (This in itself might be a good reason to break by singular actions, rather than overall player turns.)

I recently played Dungeoneer: Tomb of the Lich Lord, and though the game was overall fun, I was struck by the length of each player turn. There were a number of very distinctive actions, including one long phase where you played cards against other players and another long phase where you moved around the dungeon, and thus there could be an extended downtime from one of your turns to another. I thought that a pretty good example of an instance where breaking by actions would have improved the flow on the game (e.g., everyone does their "attack" actions, then everyone does their "move" actions).

My original article on timing was Trials, Triumphs & Trivialities #108.

Victory & Balance

The aforementioned phases of victory and end game deserved an article in and of themselves, which also related to the idea of balance. It's end game where balance really becomes important, because you don't want players to get bored beforehand, because they feel like they can already see the winner.

I had quite a bit of advice on how to balance your game leading to victory, but I'll only bullet out the biggest points here:

1. State Your Imbalance. By putting imbalances out there, players will be able to react to them.
2. Give Your Advantages Disadvantages. Do a simple first-cut balance by mixing virtues & flaws.
3. Offer Multiple Paths to Victory. Which is to say, don't make one (random) route innately worse than another.
4. Explicitly Help Players Who Are Behind. Don't be overpowering, but a boost can help make someone feel they're still in the game.
5. Help Players Stay Involved. By hiding victory information, make it partly transient, and allowing large payoffs for huge risks, you can keep people around.
6. Allow Player Interference. Players will be the best balancers, because they can react to dynamic game conditions; let them.

End game has a number of balance problems all its own; you want to make sure that the final winner is "fair", not someone randomly chosen, and also that the game doesn't drag on forever once you get to the end game.

My original article on balance was Trials, Triumphs & Trivialities #109.

There are two other elements which must be carefully considered when you're thinking about your gameplay: decision sets and random choice. These are directly opposing concepts: the more choice you have, the less randomness there is, and vice versa.

Decision Sets

Decision Sets should generally be constrained to between 5 and 7 options at a time. You can do this, and still have a fairly complex game, by constraining the total palette options in various ways. Possibilities for constraint include:

1. By Game Phase. Take the decisions one at a time according to your rules of timing.
2. By Turn Phase. Make some decisions useful early in the game and others useful late in the game.
3. By Ability. Only allow some players to make some decisions.
4. By Need. Only force some players to make some decisions.
5. By Attractiveness. Make some decisions "better" than others in certain situations.

I also noted that you can constrain by "results", wherein multiple decisions actually lead to the same answer. On the gotcha side, results can also increase the size of a decision set if each decision has several outcomes.

Since I wrote my article on decision sets, I've realized through chatting with a friend that it's OK to have a very complex decision set if it appears fairly random to an inexperienced player. Chess might be a good example here: a good player dramatically ramps up his decision set size by thinking several turns ahead, but a casual players moves a fairly arbitrary piece without a lot of strategy or understanding about the size of the decision set created by "look ahead".

It's perhaps the intelligent but casual player like me who gets into the worst trouble. I'm aware of the size of the decision set, but have no interest in taking the game that seriously — and so go find something simpler to deal with.

My original article on decision sets was Trials, Triumphs & Trivialities #105.

The Random Factor

Which leads us, finally, to decision's flipside: randomness. More generally, games are affected by "unpredictability" which can be classified into four broad types:

1. Randomness. Dice.
2. Arbitrariness. Cards.
3. Chaos. Other players.
4. Uncertainty. Hidden information.

The elements of chance — randomness and arbitrariness — can be further divided into a number of categories:

1. No chance. A lack of randomness or arbitrariness.
2. Unweighted Distribution. All chances are equal.
3. Weighted Distribution. Some chances are more equal than others, as typically laid out by a chart or table.
4. Curved Distribution. Some chances are more equal than others, as laid out my a specific mathematical curve usually generated by tossing a couple of dice.

I think chance is a very good thing in strategy games, but you need to carefully balance it so that it doesn't overpower your game and make it a simple exercise in throwing dice.

And before I close I should probably note that I've never really talked about the "chaos" form of unpredictability that I mentioned here. I think the way that you design inter-player dynamics can dramatically change the shape of your game, and it's probably worth an article or two, but I haven't figured out quite what to say yet.

My original article on chance was Trials, Triumphs & Trivialities #106.

So, how much chance and how much decision is enough for your game? My general advice is that if you want to make a fun, casual game, try and blend both, with the exact proporations related to the casualness of your gaming audience.

Conclusion

And those, in a nutshell, are my thoughts on strategic designs. The other fairly crucial article in this series has been Anatomy of a Game: The Settlers of Catan, which give many of these concepts a real workout.

Overall, I think of this series of articles as sort of a recipe book. I did offer some advice on how to lay out games, but to a large extent I tried to show you all the possibilities, as I think some of the most interesting games are the ones that put together standard strategic ideas in very new ways. So, consider this week's article the ingredient list — not a replacement for the recipes themselves, but a shorthand that'll tell you which recipes might interest you and which might not.

Go read the original articles for the full instructions.

Although I'm done with my overview of strategy games, I'm not quite finished with the topic yet. I still want to: look at how RPGs relate to strategy games; look at how the medium of web games can build upon these ideas; look into how you can expand a successful strategy game; and use this whole series as a jumping-off point for designing a new game for Skotos: Galactic Emperor - Merchant Kings.

But, that's all for the future; in the meantime I see one of my favorite dates of the year fast approaching, and so I need to finish compiling notes on some of this year's funny goings on at Skotos, for inclusion here in my regular April 1st column.

See you in "5".

Recent Discussions on Trials, Triumphs & Trivialities: