Wednesday, September 14, 2005

Game Design Review: Advance Wars - Dual Strike

Advance Wars is the latest iteration of what I like to think of as the casual console turn-based strategy genre. Writing a design review of a highly evolved genre king like Advance Wars: Dual Strike is a distinctly different task than writing my previous review of a relatively new game design like Nintendogs.

Again, this is a game design game design review, not a game review. A game review typically is written for the consumer and is intended to help them decide if they want to purchase the title. A game design review is written for other game developers and is intended to highlight successes and failures of the various layers of the game design. The hope is that we learn from the successful design practices and apply them in an intelligent fashion to our future titles.

In this design review I’ll cover the following:

  • A brief history: What is the historical context of the Advance Wars design?
  • Game anthropology: What need does Advance Wars serve?
  • Layered game design: What is a major design lesson we can take from Advance Wars?
  • What worked and what didn’t: What design decisions worked and which ones failed?
Brief history
Any meaningful review of the game mechanics of a title such as Advance Wars needs to consider its extensive history. My personal memories of the console specific implementation of this class of games go back to Military Madness (1989) on the TurboGrafx-16. A new sub-genre was created by moving some of the core mechanics of the computer strategy game godfather Empire over to a console. The roots of Empire stretch back even further to board game forefathers such as Risk and Axis and Allies. Ultimately, you can trace the core mechanics to a wide number of chess-like games going back thousands of years.

All the games in this particular genre have the same basic mechanics. You move different units around on a board and attempt to destroy all the enemy forces or capture some set of locations. They use an “I go, you go” turn structure similar to classic board games and much of the strategy comes from strategic timing of attacks and positioning of units. Game play tends to be quite quick, with most battles taking an hour or two. Realism and simulation accuracy is sacrificed in order to maintain an enjoyable pacing of the action.

There have been dozens of titles leading up to Advance Wars, each adding small tweaks to the core game play. For the console specific sub-genre, there are at least 15 years of intense genre evolution at work. There have to be at least one or two game design lessons buried there.

Game Anthropology
Every game must serve a market need. Advance Wars: Dual Strike is what I like to call a genre king. It is the premier turn based strategy game on the console. It serves a built-in audience of players who have been addicted in the past to similar games.

The problem that it solves is simple. Very few games in this niche genre are being released any longer, yet the craving still exists. Release a game and make some money. Not a bad marketing strategy.

Turn-based console strategy games are in the niche stage of the genre life-cycle. I put together a list of niche stage game attributes a while ago and it is amusing to see how well they match up with the major selling factors of Advance Wars.

  • Copy the most popular game possible: Advance Wars doesn’t stray very far from the original rules set ages ago by Military Madness.
  • Try to pick up a license for cheap: Advance Wars is a great license that is proven to sell well to the GBA crowd. Using it for Dual Strike is a simple product line extension that no doubt increased the sales considerably.
  • Pay lip service to higher layers game mechanics: Dual Strike pays lip service to things like plot, character development, etc. Why spend money on flashy graphics when your core audience will pay for the experience regardless?
Dual Strike also provides an example of an interesting market launch tactic of being one of the first titles in the genre to be released on the DS. By being first to market on the DS, you decrease the competition and increase your chances of being a genre king. When you are a dealing with consumables like retail games, the release window becomes an important market factor.

First Layer: Core Mechanics
The design of Advance Wars can best be understood in an evolutionary context. You can almost think of it as an archeological dig with each layer of game design building up on what has come before. You start with the core mechanic and then layer on additional meta-mechanics until the final game emerges.

In Advance Wars, the core game mechanics is the direct combat system, which is composed of two risk / reward sequences. The first is the move / attack verb and the second is the end turn verb.

The move / attack verb is quite basic:

  • Action: Move a unit on a square board a set number of spaces. If the unit is next to the other unit, it may attack. Units that attack first tend to destroy the other unit. Units cannot move past other units.
  • Reward: You destroy an enemy unit or gain a strategic position. This opens up new movement possibilities and increases the chance of winning.
  • Risk: You put yourself in a position where the enemy can easily destroy or block your units, gaining an advantage.
The end turn verb is also rather simple

  • Action: At any point in moving their units, a player can declare that their turn is over. At this point, the opposing player can begin moving and attacking with their unit.
  • Reward: By ending your turn when the board is an optimal position, you are ready to withstand the enemy’s attack. Ideally, you’ll thrive during the player’s turn and end up in a position of great opportunity when your next turn arrives.
  • Risk: If you choose a bad board position to end the turn, you can easily be destroyed by your opponent’s reaction.
In this basic mechanic, you have the guts of a nearly infinitely replayable system. The position of the tokens on the board forms a robust player-created environment that constantly provides the player with a new puzzle to solve. The two verbs form a complimentary short term / medium term reward rhythm. Moving a single unit takes a few seconds and results in an interesting reward or punishment. Ending the turn after 4 or 5 minutes presents a bigger risk, but also a greater potential reward.

These games were the first to clue me into the concept of ‘reward momentum.’ The act of consuming rewards gives the player the psychological incentive to keep trying to gain more rewards. Turn-based strategy gamers know this as ‘one more turn’, but the concept is applicable to a wide range of games if you understand it in terms of the rhythm of risk / reward schedules.

Here’s a quick example of the psychology:

  • The candy: At the beginning of each turn, a new series of tiny rewards are immediately present. These low hanging fruit come in the form of obvious moves, an enemy tank you’re your artillery can destroy or a partially captured city.
  • The tease: However, by the time the player is done collecting the smallest rewards, the end of the next major reward is in sight. It seems a shame to end the game when the next big event requires only an incremental effort. So the player moves the last few units and finishes the turn.
  • The cycle repeats: The game immediately presents the player with another set of smaller low hanging fruit. It is nearly impossible for a player to resist repeating the cycle.
Second Layer: Create new tokens by tweaking variables
Now that we have our core mechanics in place, we can extend them by layering on additional meta-mechanics. Advance Wars uses this design technique extensively.

The next layer is created by expanding the token set. Players quickly become bored when all pieces do the same thing. First you identify all the variables that you have to play with. Then you create new tactically interesting tokens by assigning each one a unique batch of variables.

Identifying your variables is often the critical step. Here are some rules of thumb that can be useful:

  • Identify existing variables: Movement is an obvious one. Board size is another. Creating tokens based on variations of movement alone, early game designers came up with Chess, one of the most popular games in the history of the world.
  • Identify binary states and then create variables to represent those states: In Advance Wars, units are not dead or alive like in Chess. Instead they have health points. When a unit is attacked, it loses health.
  • Map out all the other variables in the risk / reward chain: If something decreases in health, then there must be another factor that causes the damage. Ah, another variable to add to your list. Does attack affect all units equally? Add a defense variable to your list.
  • Tie variables into other variables: Now that you have you list of variables, you can look at interesting equations that tie the variables together. In Advance Wars, damage is a factor of health. The lower the health of a unit, the lower the damage.
Now we can create a whole bunch of interesting units. You have fast tanks that move quickly but do very little damage. You have slow tanks that pack a big punch. This is really just the start of this technique, but it is one that we’ll see used over and over again. The creation of new and modified tokens by adjust variables is one of the basic drivers of genre evolution and is a standard tool of game designers during nearly every step of a game’s production.

Third Layer: Classes
Many designers don’t think about it, but a map is really just another series of tokens. You can ask the question, what variables can these tokens affect? Advance Wars uses two obvious modifications

  • If a unit is on a specific terrain token, its damage is modified
  • If a unit is on a specific terrain token, its movement is modified
From here we generate a whole new series of tokens. Roads, plains, water, shore lines, forest, mountains, and pipe tokens all come into play. We also get ships that can only move on water, infantry that can move through all terrain at the same pace, aircraft that can move over any space.

At this point, Advance Wars introduces a class system to keep track of everything. Class systems are useful when the complexity becomes too large for a player to hold everything in their head. My rule of thumb is that whenever you get more than 4 or 5 of variations, group similar ones together in a single class of objects.

In Advance Wars introduces the classes of vehicles, infantry, ships, subs, copters and planes. It also creates a matrix of damage and movement rules around these types. So if you have a tank, it will do more damage against infantry than it will against a Neotank.

These matrices of possible combinations become large and would easily cover a page or two if they were printed out. Having a computer to manage this complexity becomes important since it allows the interface to all this data to be given in heuristics. A tank doesn’t do 5.6 points of damage to infantry. It is ‘very effective against infantry troops’.

Numbers are scary and many players prefer to think in terms of heuristics, or simple rules of thumb. By taking the massive complexity that we’ve introduced with our token explosion and reducing it to easy to understand relationships, we help manage complexity in the mind of the player.

Other Layers
There are numerous other systems that are layered on top of these basic systems, but you should start to see how the layering of game mechanics works to create additional complexity.

Some other notable meta mechanics include:

  • Indirect fire: The ability to attack from a distance, but not during the turn your unit moves.
  • Fuel and ammo consumption: Move to far without refueling and your unit dies. Weak, non-combat supply vehicles bridge the gap between the front and your cities.
  • Unit production and money generation: You can buy units.
  • Fog of War: Units can only see a certain radius.
  • City capturing: Only infantry can capture cities.
  • Unit repairing: Special units and cities can repair damaged units.
  • CO powers: There a meta-powers that alter the balance of units.
  • Tag Powers: You can take two turns at once and use both of the super powers.
  • Weather effects: There are meta-powers in place that affect various variables such as fuel consumption and visibility.
Very few of these are unique to Advance Wars, but they all add additional risk / reward sequences to the game.

The point of layered complexity: Preventing burnout
I’ve mentioned briefly that the point of this complexity is to prevent the player from ‘getting bored.’ That is true. However, it naturally goes a bit deeper.

A game can be described a series of challenges, some small, some large. Some are based on timing, others on spatial comprehension. The player is presented with a set of possible actions and alerted to potential risks and rewards. They are required to predict the results of their actions and choose an optimal path towards reaching the rewards.
All these challenges can be described in terms of risk reward sequences and are composed of verbs, tokens and rules. We’ve covered much of this in other essays.

However, rewards that are too easily achieved lose their impact. Why this occurs is up for debate. I’ve heard several theories:

  • Repetition and subsequent burn out: When a reward is easily achieved, the player consumes too much of the reward too rapidly and tires of it.
  • There exist types of players that seek out challenges and their conquest. If there is no challenge, then there is no point to playing.
  • The brain is tuned to spend effort mastering patterns. Once a problem is mastered, there is no longer any joy to be had in performing the activity.
The layering of complexity increases the number of options for the player to consider. Achieving simple goals becomes a balancing act for the player. Should they move into the forest square? This puts them in range of the enemy next turn, but in the line of fire of the battleship sitting off the coast. Fuel is decreasing, units are repairing, money is flowing in and the enemy is building more ships. Due to the use of highly layered game mechanics, the number of options that the Advanced Wars player has to balance is mind boggling.

Mastering Advance Wars is quite difficult. And that is the point. All the systems exist to create a system that you can always get better at, but you’ll never be able to dismiss as ‘easy’. Burn out is unlikely and even the most skilled player will constantly be discovering new tactics and subtle ways of optimizing their performance. The result is a game that stays highly addictive for longer periods of time.

Benefits of layering as a design strategy
Advance Wars relies on layering for some practical production reasons as well.

  • Reduced risk: When you layer proven game systems on top of one another, it is much easier to build a stable game design. Risk reduction comes from having your core game mechanics as a foundation and anything else you add is bonus. If the sub unit hadn’t worked out, that’s okay. You can lose it without any real risk to the game’s production schedule.
  • Easily achieved complexity: If you need more challenge, you add new elements. You don’t have to worry about pruning old functionality. Just add tokens and you are done. Contrast this to a game like chess or go. Adding new tokens requires a rebalancing of the entire rule set.
  • Obvious learning path: You can gradually introduce the complexity of the title to the player by teaching them one layer at a time. The initial maps of Advance Wars start out by teaching the core mechanics and then gradually introduce new mechanics one at a time. By the end of the first 10 lessons, the player has a huge palette of verbs available.
Problems with layering
Layering is used by almost every leading game design in every established genre I can think of. However, it is not without its own problems.

  • Painful learning curve: By constantly adding new systems you increase the learning curve. This acts as an entry barrier to new users. I’ve shown many people Advance Wars and they see the screen as a mass of hieroglyphics. Only after you gain a substantial appreciation of how all the systems work in concert are you able to appreciate the game for its genius. Contrast this to Nintendogs, which though a simpler game, was immediately appealing to users unfamiliar with the genre. A title like Advance Wars rarely expands the overall game market.
  • Impossible to balance the difficulty: You end up with two very different audiences. The first are the hardcore players who have played previous games in the series and expect more of the same, only harder. The second are the new players who will struggle with what you have already. Already I’ve seen reviews by seasoned gamers complaining that Advance Wars is too easy.

    I’m on the other side of the fence. I’m taking a break from playing because the campaign reached an almost puzzle-like status where one false move dooms my entire mission. To an admittedly incompetent fellow like myself, the game is moving a bit too far away from the casual light hearted combat the originally caught my attention. It's a tricky balance because even though this is a niche title, you want to capture as much of the niche as possible.
  • The risk of accidental simplicity leads to incremental design: No new game system can be too extreme or you begin changing the existing mechanics of the title. A classic example of this are the CO powers. In a game like Cosmic Encounters, these meta game mechanics radically alter the rules of the game. You get crazy mechanics like “attack power and defense power are swapped” that force you to rethink all of your strategies.

    The CO powers in Advance Wars are tame by comparision. As the game design becomes heavily layered, there are simply too many systems to balance. You run the risk of creating ‘accidental simplicity’, a hole in the design where the player can bypass your complex game systems with a simple, highly effective tactic. For example, early RTS games allowed you to win by building only tanks. They had lots of other units, but no one ever used them.

    If you fail to isolate your design mechanics interactions from one another these strange loop holes will emerge and your game loses its challenge. As a result, designers become more tentative and subtle in their modifications. The final game is often a comfortable, mature title that will never be wildly innovative.
What worked
Now that we’ve got the basics out of the way, here are some of my favorite moments

  • Casual strategy: Highly differentiated units, fast paced combat and a refreshing lack of statistics. This is a classic beer and pretzels war game of a sort that I thought stopped being made ages ago. I’m happy I picked up a DS and was able to discover such a delightful example of one of my all time favorite genres.

  • Indirect combat: I loved how indirect combat forced you to make a choice between moving or firing. Indirect combat can win the game for you and I make use of it probably more than I should. This mechanics makes you very aware of the rhythm of turns and how turns affect your tactics.

  • The limited plot: You have a couple of tiny heads and a few lines of text per mission. The plot is the thinnest excuse possible in order to introduce interesting maps, units and COs. Yet it works quite well and was quite likely a relatively small portion of the overall budget. I love it when a game cuts substantial corners and no one really gives damn. You don’t have to turn all dials to eleven to build a successful game.

What didn’t work

  • Critical information revealed halfway through the mission: So far I’ve run into terrain effects that turn on half way through the mission. I’ve also seen massively dangerous CO powers that require building your deployment strategy around them. These are all things that are impossible to respond to until you’ve started the mission once and been smacked in the face. The result is that you need to restart and try again.

    Lesson: Give players critical information up front or at least give them enough info to succeed without restarting. If players are restarting multiple times, you’ve failed to cater to the casual gamer. One of those times they are not going to restart and your game will be shelfware.
  • CO powers and weather: CO powers seems to be the big innovation for this title, yet the impact of these meta mechanics didn’t blow me away. The minor modifications subtly affect your play style, but to my inexperienced eyes, one CO seemed only slightly better than another.

    Lesson: Be wary of overbalancing new mechanics to the point of pointlessness. Sometimes, it pays to be a bit more bold.
  • Linear mission structure: The campaign only allows you to play one mission at a time. If you are stuck on a single mission, there is no chance to improve your skills on an additional mission. You are at a dead end.

    Lesson: Never force the player into a dead end challenge. Always give them alternative paths.
  • Puzzle Maps: Some maps were open ended and allowed for creative solutions. Others required very specific movements in order to succeed. This is a player preference and I’m sure there are people who get insanely excited about puzzle maps. I’m more of an explorer type fellow so any hint of ‘one false move and you’ve lost the mission’ game design is a huge turn off. Luckily almost all the maps open up a bit after the first 4 or 5 turns and you can escape the demented planning that all level designers are guilty of inflicting on innocent players.

    Lesson: Kill all existing level designers and burn any records of their scripting tools before you hire a new batch. Seriously though, the use of heavily scripted scenarios is a design crutch that hints at a shallow player environment. If you are confident in the depth of your game, instead build maps that let players experiment. Always avoid ‘instant death’ scenarios.

I’ve enjoyed my time so far with Advance Wars: Dual Strike. It fits the portable format nicely and is a professional, well balanced example of one of my favorite genres.

There are several design concepts that I’ve discussed in this review:

  • Layered game mechanics: By layering meta-mechanics on top of one another we can extend a proven core game mechanic and reduce long term player burn out.
  • Reward momentum: By stringing rewards of various sizes after one another in a visible, predictable sequence, you can encourage the player to keep playing.
  • Creation of new tokens by tweaking variables: We can quickly increase the complexity of a title by introducing new tokens that are simple variations on implicit design variables.
  • Use of classes to simplify the learning curve: We can manage the inherent complexity of a highly layered design by chunking similar objects into classes and explaining relationships with simple heuristics
I highly recommend looking for other examples of game design layering in other titles. It is a very common technique that should be in every designer’s tool box. However, as we’ve discussed above, it should be used with care.

There are numerous other lessons to be gained from reviewing the design of a game like Advance Wars: Dual Strike and to be honest, I’ve only scratched the surface. With a title with this much history and depth, you could easily write a book that just goes into the implications of each mechanic and the theory behind it. I’ll see if I can schedule another bit of vacation. :-)

Take care

PS: Fixed a couple of typos that folks have caught so far. Apologies for any jet lag induced wackiness in this post.