Friday, March 28, 2008

Old Favorites: Skies of Arcadia Legends

Skies of Arcadia Legends (a Gamecube port of Skies of Arcadia for the Dreamcast) is still to this day one of my favorite RPGs. It was a breath of fresh air as far as RPGs go at the time, due to its uniquely fun characters and setting. However, the aspect of the game that I remember most fondly is the game's battle system. While the system had a few unbalanced elements, it strongly rewarded, and sometimes absolutely required, excellent strategy on the part of the player.

The heart of Skies of Arcadia Legends' battle system was the Spirit gauge: a single meter that fueled the attacks of all four party members. Both magic spells and individual character's super moves use up set amounts of Spirit points. However, each character also automatically restores a certain number of Spirit points to the gauge each turn. Furthermore, a character can use the "Focus" command to double their Spirit contribution for one turn. The result of this system is that every battle in the game comes down to careful management of character actions and the Spirit gauge. So developing sophisticated strategies that involve all four characters is a must.

Of course, this system is at its strongest during difficult boss battles. It usually is not necessary to worry too much about normal enemies, particularly since the system lets the player annihilate most random encounters by using Aika's all-enemy attacks without cost every turn once the party generates enough Spirit. While the game leans towards the easy side, there is a category of enemy that always proves to be a challenge: the optional Bounties. All of the optional Bounty enemies, as well as the Pirate Hunter Piastol, increase in strength to match the party level of the heroes, so it is impossible to beat them easily by taking the time to level up. As a result, the fights against the Bounties and Piastol are some of the most strategically demanding and fun that I have ever seen in an RPG.

The biggest problem of fighting off these enemies is that their offense is often enough to wipe out the entire party in very short order. However, one of the greatest strength of the Skies pf Arcadia battle system is its strong support for defensive strategies. Every character has at least one defensive/support special move that has turn priority, in other words always goes at the beginning of the turn. For example, Vyse has Skull Shield, which shields the entire party from basic physical attacks (and follows up with counter-attacks) for the entire turn. Aika has Delta Shield, which makes the entire party immune to magic for one turn. Fina has numerous powerful healing moves and an all enemy debuff attack called Lunar Winds. If the party is generating enough Spirit, Vyse and Aika can use both of their defense moves to make the entire party immune to regular attacks and magic spells (though enemy special attacks, which most optional bosses have, are unaffected).

So, some fairly complex and versatile strategies are possible for example, a typical strategy might be:
Vyse- using his best attack Special
Aika- using items to heal/buff
Fina- Focusing to build Spirit
Enrique- using Justice Shield to halve all enemy damage

Alternatively, if fighting enemies without Special attacks:
Vyse- Using Skull Shield
Aika- Using Delta Shield
Fina- Alternating between Focus and using items
Drachma- Alternating between Spirit Charge and attack Specials

Simply attacking every turn with every character and healing as necessary won't cut it against many of the super-powerful optional bosses. The game also opens up more strategies by making a full-party super attack if the Spirit meter caps out. So the player can choose between an attack strategy and turtling until the super attack becomes available.

I find this kind of full-party strategic combat to be much more interesting than the somewhat more common "save MP until boss fights, then nuke the enemy with your best spells every round" battle-systems. The only real weakness of the Skies of Arcadia system is that magic is useless in it. Most spells only do the same kind of damage as special attacks, or mimic the effects of items. However, Special attacks can do any kind of elemental damage, and items do not use up Spirit. So healing and buffing spells are completely over-shadowed by items, and attack magic just doesn't have much in its favor.

Still despite that one weakness, Skies of Arcadia has one of the best RPG combat systems I can remember. The only way I can think of improving it would be to couple it with the Lunar series system of enemies broadcasting their attacks, so that the player can devise counter strategies.

SRW:OG2 Battle Mastery

As I mentioned in my last post, the Battle Mastery system of the Super Robot Wars games deserves some recognition, so I will write a bit more about it today.

In the Super Robot Wars Original Generation games, every battle has an optional victory condition called the Battle Mastery condition. This condition has no bearing on the battle itself, but every Battle Mastery you acquire adds to the difficulty of the game; if you get no Battle Masteries, you will be on the Easy difficulty, but if you get many you will be on Hard difficulty, where you will earn less money and experience while fighting stronger foes. Other than the difficulty setting, fulfilling Battle Mastery conditions has a few more effects: some Battle Mastery objectives are tied to unlocking hidden items and units, and, more importantly, you can only play the final mission of the game and fight the true final boss if you have completed a sufficiently large number of Battle Mastery conditions (in SRW:OG2, most sources say that you can only miss three out of 41 and still reach the final mission).

I think the way this system controls difficulty is brilliantly elegant. If a player is inclined to take on optional challenges (in other words, the kind of person who likes to play videogames for the challenge), then the game becomes more difficult to match. On the other hand, a person who just wants to play through the game for the story is not compelled to take on the challenge of the Battle Masteries and is allowed to play through on an easier route. What is more, the difficulty level scales gradually, so to a certain degree the player can adjust the difficult to a comfortable level by simply earning Battle Masteries where he can and ignoring the ones he can't. It is a system which is fair to all kinds of players and controls the difficulty without an awkward difficulty selection choice (well, there is one, but that is only for unlockable extremely hard and extremely easy versions, which also scale). I can't state enough how much I like the concept of this system.

Praise aside, there is one problem I do have with the Battle Mastery system as it is seen in SRW:OG and OG2. The main reward for having a lot of Battle Masteries is being able to see the true final battle, a dramatic and fun finale that is an important conclusion to the plot. If you do not have enough Battle Masteries, there is a sudden and unsatisfying interruption in the flow of events that leads to a hasty ending. What is more, the true final battle of OG1 (which I missed originally) is assumed as the real series of events in the following game, which can result in confusion for people playing the sequel who did not see it (like me). Forcing a player who wants to see the dramatic finale to play through the most difficult route through the game is a flawed concept, in my opinion. It is just a better experience for the player if he can see the whole game as it is meant to be seen, regardless of difficulty.

Despite that one objection, I do like the idea of having rewards for players who are willing to take on a challenge. Certainly, it would be a good idea to link hidden units and items to the Battle Mastery system. At least, I would prefer to do that than use the esoteric conditions so often seen in the SRW series. Another possibility would be to tie Battle Masteries to unlockable game modes, minigames, glossaries (SRW:OG2 could really use a character/mech glossary), or other similar things.

I should also discuss the actual conditions used in the battles... As a whole, almost any condition for earning a Battle Mastery works well, so long as it is more difficult than just completing the normal mission while still remaining possible. I prefer Battle Mastery conditions that can be cleared in the early stage of a mission, rather than ones which enforce a time-limit on a whole battle (mostly because the latter make it hard to build up weaker characters), but that is just a preference. The only Battle Masteries that I think are poorly designed are the ones that must be achieved through relying on luck or the AI moving right rather than player preparation and strategy (such as OG2's Mission 16 or Mission 20 on the Aviano route). The latter, which tend to be things like protecting weak allies or killing enemies timeframe that is so limited that you must rely on the enemies moving in a specific manner, are simply frustrating and are not a good measure of the player's skill level, so they are not fun and not a good way to determine challenge level.

Anyways, I think the Battle Mastery system is a good concept that could easily be implemented in other games. Games have been offering optional challenges linked at least since Goldeneye for the N64, and many other games (like the Devil May Cry series) rate the player based on ability. A system like the one in Super Robot Wars would fit games like those very well.

Wednesday, March 26, 2008

Non-Standard Actions in RPGs

One of the most interesting developments of late 3.5 edition Dungeons and Dragons, and in the upcoming 4th edition D&D is of swift actions (also known as minor actions): an action a character can take in addition to a regular move or attack action. In 3.5 edition, it is often used to perform short-duration buffs and the like. This concept is actually rather new to RPGs, of both the pen and paper and electronic varieties, and I think borrowing this concept can be very useful to video game RPG design.

To illustrate this point, I will use the example of Soul Nomad, a game I have discussed before. In Soul Nomad, every unit leader has a list of special powers called tactics, which use up the unit's attack action when used. These can vary widely in effects, from area of effect magic blasts, to self-healing powers, to buffs. However, the usefulness of these powers also widely varied. While the attack spells are incredibly useful, I found that I simply didn't use many of the other kind of tactic. In particular, I only used the buffs in the period of time when I was still closing in on the enemy, and I never got much use out of the healing tactics. My decisions came from  the opportunity cost of using a tactic.

The problem with spending a turn to buff or heal using tactics was that it often was not a very efficient use of an action compared to an attack. For example, lets assume that a given unit goes into battle 4 times in one mission (not out of the ordinary in some maps in Soul Nomad). If the unit attacks four times, it will do a total of 400% damage, assuming ideal conditions. If the unit instead spends one turn buffing its attack power by 30% (a typical number in Soul Nomad), and then attacks three times, the unit will only do a total of 390% damage. So spending the one turn to buff actually decreases damage dealt if the battle lasts less than five turns (assuming the buff even lasts the whole battle, I was never able to confirm this or not). Spending a turn de-buffing an enemy can be an even worse tactic, if the enemy dies quickly.

While I admit that this is an overly simplified model, it reflects the way a player makes tactical decisions while playing a game like Soul Nomad. It is also worth noting that spending a turn buffing just isn't as fun as watching a character or unit perform a cool-looking attack. Even though there were times I could have gained an advantage in Soul Nomad by spending a turn healing or buffing, I often instead preferred using the turn to launch an attack against an enemy unit. It made me feel like I was actually making progress more.

I think switching some of these tactics over to being a swift action would not have been a bad idea, though it probably would require rebalancing of the game to match. Since the player was limited to using these abilities to only a few times per battle anyways, there would still have been reason to save them for when needed. It also could have expanded the tactical choices of the player, by making the player choose between healing up before attacking and buffing before attacking every turn.

I think the addition of swift/minor actions to video-game RPGs would open up a lot of so-far unexplored design space. I will probably talk more about this topic, and on actions in RPGs in general, more in the future.

SRW:OG2 Victory Conditions and Optional Objectives

I have been replaying Super Robot Wars: Original Generation 2 quite a lot over the last week or so (I have completed 23 out of 43 missions so far), and replaying that game has given me a lot to think about regarding good and bad ways a game can present victory conditions and optional objectives to a player.

Missions in Super Robot Wars: OG2 can be very complex and involved, which I think is a very good thing. For example, one mission is split into two phases: the first requires you to lure a number of enemy units outside of a certain area and then destroy them within a time limit, and in the second phase both you and the enemy are reinforced with additional units and you must reduce the enemy commander's hit points to less than 70% of maximum. Mixed in with these two phases are a number of different story events, hidden requirements for unlocking secret units, many opponents you can fight in order to gain experience and money, and an optional "Battle Mastery" mission objective. This kind of complexity is typical for a mission in that game, and I think that such complexity is great. However, there are a number of things I need to criticize regarding this structure.

First off, I need to mention the way the victory conditions themselves are stated. In the above example, the victory condition in the first phase is stated explicitly, but the victory condition in the second phase is stated as "Defeat All Enemies," even though the only factor that determines actual victory is in fact the hit point total of a single foe. The victory condition "Defeat All Enemies" is in fact extremely unreliable in that game, and is listed in the victory condition menu when the actual victory condition can mean anything from "Defeat 15 enemies, at which point the enemy reinforces and conditions change" to "Either bring the boss's life down to 50% or wait three turns." Also, if there happens to be multiple victory conditions, or a victory condition happens to be optional, it is never specified, and only a single victory condition is listed as if it were the only one. This vagueness can often frustrate me to no end, especially when I misdirect time and energy towards one goal, only to result in me not having enough time and energy to accomplish the real mission objective. I think that games should be honest when they spell out mission objectives for the player, and SRW: OG2 is terribly dishonest in that regard.

The unreliable nature of information regarding victory conditions is especially problematic concerning hidden units and items that need to be unlocked. I really like having a large number of hidden things in a game that reward players who are willing to undergo an optional challenge, and I like that SRW:OG2 has so many (and makes them worth the effort), but a big problem with the game is that there are absolutely no in-game hints to their existence, at all. For example, to unlock one hidden unit, you need to have a certain character (Arado) defeat four out of the five available enemies before completing the first phase of mission 19 (among a few other conditions). There is no indication in the game at all that defeating opponents at that time has any importance, let alone is linked to a hidden unit. Further, if you accomplish that objective, there is no indication that have completed a hidden objective. When I first went through the game, I had no idea about the importance of that particular battle, and I never even had any hints at all about the possibility of acquiring that particular hidden unit. Even now on my second playthrough, when I have successfully accomplished the main requirements for getting that unit, I have had no confirmation yet in the game, and I will not until I actually get the unit a few missions from now. Hidden objectives are great, but there needs to be some way for a player to know about their existence without reading a strategy guide or FAQ.

Finally, I want to talk about the Battle Mastery conditions. I think my brother mentioned these before on this blog, and they deserve their own post, but for now I will just say that they are optional objective in each map that have a few problematic interactions with what I have been mentioning above. The Battle Mastery conditions themselves are explicitly laid out (more so than actual victory conditions, which sometimes helps shed some light on how the battle which actually play out), but some conditions work better than others. Namely, there are some missions in which the Battle Mastery condition says something like "Complete this map in 4 turns", even when the main victory condition is deceptive. Because of this, it can be difficult to determine if you are on target or not for achieving the Battle Mastery, and it is easy to end up missing the Battle Mastery entirely because of that confusion. For example, I once mistook the listed victory condition "Bring Seolla down to 30% of her HP" as the actual victory condition, when the actual victory condition is "Defeat All Enemies" (Seolla simply flees the battle when you bring her down to 30%, changing the listed condition). Since the Battle Mastery was "Complete the map by the end of the fifth player phase", the fact that I held back on attacking Seolla because I wanted to complete a hidden objective before the map finished, and avoided targeting the boss because I thought he was optional, resulted in me failing to complete the Battle Mastery objective (which resulted in me going back to play the mission from the start again because I couldn't afford to miss another Battle Mastery while only halfway through the game).

Victory conditions are the guidelines a player uses to help determine how to manage his own time and own preferred way to play a game. Vague and unclear guidelines simply lead to a less fun experience for the player, and should be avoided.

Monday, March 24, 2008

Metal Gear Solid Story Pacing

While I am a big fan of the Metal Gear Solid series, there is one somewhat consistent problem with the series that bothers me whenever I play many of the games in the series, particularly the original: the plot of the game isn't paced quite right. Namely, the games tend to create a lot of dangling plot threads and unanswered questions, but holds off on resolving these plot details until the very end of the game. This results in the final sequences of Metal Gear Solid games getting significantly slowed down and dragged out, often to their detriment. 

The final sequence of the first Metal Gear Solid, and with it all of the big plot revelations of the game, is initiated when Solid Snake starts using the PAL key to "deactivate" Metal Gear. From this point until the end of the final chase scene, here are the plot revelations that are made:
*Naomi Hunter's entire life history.
*The details of Liquid Snake's Xanatos Gambit.
*Master Miller was Liquid Snake the entire time.
*The existence of a bio-weapon called FoxDie.
*The fact that Snake's job was to just be a carrier for FoxDie the entire mission.
*Grey Fox's motivation for fighting Snake, and that he killed Naomi's parents.
*Most of Liquid Snake's life history, and why he looks like Snake.
*The existence of Les Enfants Terrible.
*Naomi's modification to FoxDie.

These scenes resolve the entire plot of Metal Gear Solid other than the stories of individual bosses and a few side characters. This in of itself would not necessarily be a bad thing, if it wasn't for the fact that many of these plot elements were only seriously brought up in this ending sequence. In particular, FoxDie's existence was only hinted at a few strange deaths earlier in the game. Liquid Snake's relationship with Snake was only developed earlier in the game through his strangely similar appearance to Snake, and through a few cryptic statements he makes during a boss fight. So entire complex plot threads are only seriously introduced and explained over the course of the last two hours of the game.

The result of this structuring of the plot is that between two phases of the final boss battle, the destruction of Metal Gear REX and the final one-on-one duel, the player is treated to a long monologue (maybe soliloquy?) from Liquid Snake where he explains about the entire Les Enfants Terrible project. It takes up a fair amount of time, and as a result weakens the pacing and fluidity of the final battle.

Many of these big revelations didn't even need to be put off until the end of the game. Unlike the information on FoxDie and such, which would have derailed the plot if Snake learned about it earlier in the game, much of the backstory of Liquid Snake could have been told earlier in the game. Knowing that Liquid Snake was Snake's twin brother, or that they were clones of Big Boss, probably wouldn't have seriously changed Solid Snake's actions in the game. Ditto for much of the information about Grey Fox. In fact, revealing more information on Liquid Snake earlier in the game could have made him more compelling of a villain earlier in the game, since he spends most of the story a complete enigma.

Following People Around All Day

I guess as a continuation of this whole "giving identity to lots of characters" vibe that my brother and I have been writing about lately, I wanted to follow up on my last post. After thinking a bit, I realized that there is one more rather good way to make characters interesting, even without forcing the player to pay attention to a character through a difficult character recruitment phase. In fact, this method is pretty good for making both controllable characters and NPCs interesting. This method is seen in at least two games I am aware of: Radiata Stories and The Legend of Zelda: Majora's Mask, two games where every character is on a set schedule, wandering around acting out their daily lives.

In these two games, time passes hour by hour regardless of what the player does. As it does so, all of the characters wander around on their own paths. A single character may leave their house in the morning, go talk to friend, walk around town a bit after that, go to work in the afternoon, work all day, and then return home to sleep at night. At different times, such a character may say different things, offer different services, potentially trigger a story sequence, or simply lead the player to different places. This has the potential to give quite a bit of depth to a character by fleshing out the character's personality and quirks, allowing a character to belong to multiple different groups of people, and making the character seem more life-like and less like a static part of the scenery.

An important effect of giving every character a schedule is that it encourages the player to closely examine the game world and its characters. It becomes fun and interesting to go to different places at different times, hang around at one place all day to see what happens and who comes by, or just follow a person around all day and see what they do. Radiata Stories and Majora's Mask both reward such behavior, which makes the experience even more fun. It is impossible to complete the Kafei quest in Majora's Mask without exploring the game in this manner, and the complexity of that quest makes that particularly sub-plot and its characters memorable even for the Legend of Zelda series. In Radiata Stories, there are some hidden things you can only find like this, such as a character you can only recruit by following the leader of the Warrior Guild through the maze-like sewer as she goes on her daily trip to pay respect at a hidden shrine, or a brief moment in which an arrogant and aloof girl watches over her sick brother late at night.

Another good aspect of such a system is that it allows for the plot of the game to have more meaningful changes to the actions of generic characters and NPCs. In more traditional styles of game, NPCs may change where they are standing or what they are saying, but in this system NPCs may change their entire daily schedule based on a story event, which can highlight the importance of the story upon the setting and make all of the characters more complex. The changes in a region that you can trigger in Majora's Mask by completing side-quests or defeating a boss are a great example of this at work.

There are certainly some problems and limitations with such a system, though. As is shown by the examples I have listed, it works best in a limited, but detailed setting, usually built around a town that works as a hub for all of the player's activity, because it would be absurdly difficult to implement it in a wider setting, and it would be equally more difficult for the player to understand the cycles of a larger setting and number of characters. Also, it can lead to problems where the player as to spend an hour of play time just to complete a minor task he set out to do if he times his actions improperly. That said, I think the potential benefits of such a system far outweigh the potential drawbacks.

Saturday, March 22, 2008

Lots of Characters, continued

Part of the problem with having lots of characters in a RPG is making sure that the player actually has the incentive to use more than a few favorite characters. Sometimes, there are games that have lots of characters and little incentive to use more than a handful. For example, most Fire Emblem games give the player dozens of characters, far more than what the player can field in a single mission. However, there are strong incentives in the gameplay of many Fire Emblem games to focus on only using a handful of favorite characters. Part of this is that many of the later recruitable characters are mechanically near-duplicates of earlier ones, sometimes with inferior stats. There is no reason to build up any characters outside a small group of favorites.

Even though most of its iterations have only a few characters, the Final Fantasy series yields some interesting trends after some analysis. Recent FF games have often given the player a group of six or seven highly customizable characters, yet only let the player use three of them at any given time. Many players have realized that there is no drawback to focusing on building up and using only three of the characters (half the team) and ignoring the rest; in fact, it is an advantageous strategy. In an environment like this, giving the player an optional character late in the game is not a very meaningful reward.

And yet, there is one Final Fantasy game that did reward the player with optional characters: Final Fantasy VI, which had fourteen characters in total, far more than any other Final Fantasy. Arguably, all but 3 of these characters are optional, since the player has to undertake many optional quests to recruit/regain most of them in the second half of the game. However, the game gives the player strong incentives to use all of these characters, and rewards the player for seeking out the most elusive and optional ones.

First off, all of the characters in FFVI are very different mechanically. Not only do the different characters have wildly varying stats and equipment, they all possess powerful, unique special abilities. Some of them, such as Gau, even operate under very different rules than other characters. So, choices about which character to use can go beyond mere personal preference.

Second, FFVI gives the player situations where using different party combinations is advantageous. The Tower of Magus, for example, is a dungeon where only magic can be used. So, a party of nothing but strong magic-users is a good choice to use for clearing it. A party of nothing but physical fighters would work poorly.

Third, there are points in the game where the player can actually use more than a few characters at once, or on alternate paths. In a military battle a quarter of the way through the game, in the Pheonix Cave, and in the final dungeon, the player can switch between up to three teams. Thus, the player can actually make use of having a lot of characters.

Finally, the game forces the player to use certain characters in his party at various points in the game. The player has to send Locke and Celes to the Empire or use Shadow on the floating continent for example. While it takes control out of the player's hands, it both forces the player to keep certain characters in use, and gives the designer a chance to make characters more developed by including them in the story.

Many successful games built around having many characters employ these at least some of these four techniques. Here are some individual examples:

Super Robot Wars Original Generation 2 does a good job of mechanically differentiating its various characters and mechs. Some mechs can dodge almost anything, others can absorb huge amounts of damage, while others can attack numerous foes at once. While mechs tend to fall into general categories, they all have individual quirks and uses.

Chrono Cross is actually very good at encouraging different team choices. Since most dungeons and bosses are dominated by a single element, it is advantageous to always bring at least one character with the opposing element (to attack with) and one character who is of the same element (who resists attacks).

Suikoden III is a great example of a game which gives the player opportunities to use lots of characters. Not only does the mass combat system let the player field up to ten units of five characters each, the multiple perspective system of the game lets the player use various different teams. Not to mention the final dungeon forces the player to create three teams which each challenge their own major boss.

Again, Super Robot Wars Original Generation 2 is good at using the story to make the player use certain characters. In almost every stage, the game forces the player to deploy one or two specific characters who have their plot developed a little more along the course of the battle. So, the player is encouraged to use almost every character.

Friday, March 21, 2008

Difficulty in Recruiting Characters

This topic is pretty closely related to my brother's last post concerning the ways you can differentiate characters in a large cast. There is a major factor regarding that I want to mention: making the act of recruiting a character itself distinct is one of the most important things for making a character memorable. Often, in a game in which you can recruit a large number of optional characters, optional characters will not have any importance to the plot and will have few story elements based on them. In games like that, the only story sequences where a designer has room to make a character memorable is the scene in which you recruit the character.

A good example of this principle at work is seen in the Suikoden series of RPGs. It is a good example for two reasons. The first reason is that, in playing through those games, I have recruited hundreds of characters, but only a fraction of those characters actually stick out in my mind. The second reason is that there is a wide variety of different ways you use to recruit characters in those games, ranging from characters who can be recruited simply by talking to them, to characters who require extremely complex and long sidequests in order to be convinced to join. As a whole, I believe that characters who are harder to recruit are much more memorable and interesting.

A good example of this principle in action is seen in the character of Jeane in the Suikoden series. In Suikoden 1, you can recruit her simply by talking to her and asking her to join. In Suikoden 2, Jeane appears again, is recruited the same way, fulfills the same important role in your team, and I did not even recognize her as a recurring character until I played Suikoden 1 again. However, by Suikoden 5 she is one of the only two characters to have appeared in all five games (even though one is set over a hundred years apart from the rest), and recruiting her involves diving down into an forgotten ancient ruin of a lost civilization in order to "retrieve something she left behind". A character who was easily forgettable in the first few games is transformed into one of the most mysterious and interesting characters of the Suikoden series, simply by moving the unusual aspects of her character into the phase when you trying to recruit her, rather than leaving it wrapped up in hidden scenes and hard-to-acquire information.

Another good example is the set of gamblers who appear in the Suikoden games. Almost every gambler in that series requires that you defeat them in their preferred game of chance before they will join. All of these characters are completely forgettable, and are so uninteresting that I don't even feel like looking up their names. In a variant of this, the character Linfa from Suikoden 5 is a gambler who makes a similar challenge, cheats you out of your money in order to pay off her debts to a suspicious individual, and then challenges you to a game several more times in several more places while showing contradicting goals and motivations, before eventually being forced into joining your group by the suspicious individual. As a result, Linfa is given a chance to be portrayed as a character with her own story, rather than just a random gambler whose only character trait is a love of gambling.

As a whole, I consider Suikoden 5 to have one of the best casts of optional characters in any of the Suikoden games, and that is in large part due to the unusually high difficulty of recruiting characters in that game. There are some potential allies who have very involved and fun sidequests in that game, many of which reveal interesting things about the main plot. In contrast, I can hardly remember any characters at all from Suikoden 4, a game where the majority of characters are recruited simply be talking to them, and of those characters I remember I can't think of many distinct traits.

Anyways, these musings should not be interpreted as a statement that recruitment difficulty and complexity is the only factor in determining how memorable a character will be. Things like concept, character design, novelty, and mechanical usefulness are extremely important, after all. However, a lot of those traits will never be shown off properly unless the player is given the time to actually pay attention and learn about those things, and one of the best opportunities is when that character is being recruited.

Wednesday, March 19, 2008

Differentiating Lots of Characters in RPGs

Every once in a while, a game company will create a game where it is possible to recruit a very large number of characters. The Suikoden series is the best example of this, where each entry in the series lets the player recruit 108 characters (or more). Recruiting this many characters can be a lot of fun, since recruiting a character can be a bigger and more meaningful reward for a sidequest than receiving a treasure. However, introducing lots of recruitable characters involves making dozens or even hundreds of characters useful and interesting, which can be a tall order for a game designer.

One of the way the Suikoden series handles this problem is by giving recruitable characters various different roles and functions. Of the 108+ characters in a Suikoden game, usually only a fraction consist of main combat characters. Other kinds of character can include shopkeepers, gamblers, minigame operators, commanders for the mass combat system, and information brokers. Since no two of these kind of character typically overlap, these characters are always distinct and useful.

However, there still remain several dozen battle characters that the game designers need to make unique. A lot of this is carried out by giving each character a distinctive appearance and unique weapon. Some characters also come equipped with unique runes (the basis for special attacks and magic powers) that only they can use. However, there are usually still a lot of characters with similar stats and uses to other characters. Particularly in later Suikoden games that allow a lot of customization, many characters end up looking a lot like others.

Radiata Stories tries something a little different. There are 177 recruitable characters in Radiata Stories, and all of them are solely useful for battle. However, the game tries to differentiate them and make them useful by completely doing away with character customization outside of the main character. The player can't change the equipment of any recruitable character, and no character other than the hero can learn new attacks or skills through the player's effort. Furthermore, it is difficult to keep more than a few characters up to level with the main hero at a time. As a result, characters can fall behind and become outdated quickly.

Yet, this is actually a decent solution for having lots of characters. By forcing the player to phase out older characters and recruit newer, more powerful character, the player is much more likely to use a larger fraction of the game's cast. Also, fixing the abilities and attacks of side-characters also gives the designers more room to differentiate them than in a system with lots of customization. However, a lot of the characters fall into clear archetypes and often wear similar uniforms to each other, so quite a few characters end up looking like near-clones of other characters.

While it usually doesn't have as many characters as the two games listed above, the Super Robot Wars series can often have dozens of pilots and mechs to choose from. Each kind of mech has a unique set of attack moves and stats, that while they can be upgraded, their basic parameters are fixed. So, every mech has unique attacks with distinctive, over-the-top animations. Furthermore, pilots are differentiated by their stats, which mechs they can pilot, their unique list of spirit commands, and their theme-music. So, even if the characters often have similar stats, they look and feel very different in play.

My final example is Chrono Cross, which had 40 or so characters. While they all had unique looks, personalities, and even accents, many of them were not differentiated enough mechanically. While there were numerous weapons that different characters used, large categories of character equipped the same weapon. And while every character had three unique special moves, almost all of those attacks were either single target attacks or all enemy attacks. So, there isn't much mechanical variation between characters. The only important mechanical choice to make is choosing what element of character you want to bring along, so it is possible to play through the entire game using only five or so characters.

Hmm, I am beginning to ramble and go off of my point. I think I will get into a more focused discussion of some of these elements in my next post.

Super Smash Bros. Brawl: Stage Builder

I think the fact that I am going to write about the Super Smash Bros, Brawl Stage Builder is a sign that I need to find more inspiration for this blog soon. Oh well...

I really like that the game designers put a Stage Builder into SSB Brawl. If implemented well, a system that lets a player create more stages for a game can add a bit to the long-term enjoyability of the game. Also, in my opinion, it is both easier to do and more rewarding than a create-a-character system (which usually ends up either with the player only controlling a character's looks, without being able to change anything that affects actual gameplay, or a limited and difficult to balance system). That said, I am not convinced that the designers of this game actually succeeded in making an interesting Stage Builder.

There are at least two big problems with the SSB Brawl Stage Builder, both of which stem from the limited number of pieces with which you can build stages. The first problem is that there are not a wide variety of different gimmicks and special parts that you can add. I have not unlocked any hidden parts yet (and I have heard that there are not many), but so far the choices are very limited. Other than normal blocks and floor tiles, the main choices are ladders, floating platforms (of one size), falling blocks (of one size), ice blocks (mostly of one size), and spikes (of one size). There are no other gimmicks like flaming traps, retracting spikes, larger moving platforms, platforms that randomly drop those standing on them, cannons to launch characters, etc. Also, it seems impossible to control things like gravity, water level (or presence of water at all), upward or downward flows of air, or any of the other effects that exist in various normal stages and areas of the Subspace Emissary mode.

One particularly harsh limitations in the stage builder that relates to the problem above is that it is impossible to layer any kind of block, or even fit anything but normal blocks closely together. Most objects require a very large amount of space (most of the "Structures" require far too much), and this means that it can be difficult to layer in a complicated design. For example, you need a fairly large amount of space to have a block that will fall out beneath you and drop you onto a spike trap, so much space that it prevents such a gimmick from being used in places where it seems logical.

The other, and perhaps even more problematic, problem with the Stage Builder is the lack of purely aesthetic, or at least mostly aesthetic, components for stages. For each of the three stage types there is only one way for any kind of floor or trap to look. The different stage types have different "Structures", but these blocks are all minor variations of "object standing by itself in your way", and can't be used to replace structural pieces. Also, the differences between the three stage types in the main structural elements are very minor. As a whole, there is almost no way to try to create a stage with a unique or interesting look; it is only possible to control a stage's function. I find this to be unsatisfying.

The Stage Builder is a good idea, but it just seems too simplistic and it has too few options for me to like it.

Monday, March 17, 2008

Super Robot Taisen: Original Generation 2 and Branching Structure

A common element in most installments of the Super Robot Wars series is to have branching plot paths. In Super Robot Taisen: Original Generation 2 for example, the plot of the game branches three times. Each time, the cast of characters divides into two groups to accomplish different missions, and the player is forced to choose which of the two paths he wants to play through. After a few missions, the two paths reunite, and continues again down a single path until the next branching point. Unfortunately, while this structure does have some merit, it wrecks havoc on the the coherency of the game's plot.

The greatest problem is that plot elements which are introduced in a unified story section of the game are often developed or resolved in certain missions on one plot path during a split story section. For example, a major villain named Archibald Grims first challenges the heroes in the first few missions of the game, and recurs constantly through several battles. However, Archibald often only reappears to fight the heroes again on only one of the paths during split path sections. Furthermore, Archibald is finally defeated by the heroes for good during a split path. As a result, a player who takes the opposite paths that Archibald Grims doesn't appear on will be introduced to Grims, then wonder where he went and never get to see the resolution of his story.

This gets even more complicated when plot elements and characters are introduced during one of the branching sections. For example certain villains, such as one of the Inspectors, first appear on the battlefield during some early branching sections, but then reappear during a unified section. If a player picked a different path, then the player is suddenly introduced to a character who the characters in the game have already met. And following the problem listed above, these characters can also have their stories resolved during split story sections.

My brother actually experienced the result of this. During his first play through of the game, the Machinery Children were introduced off-screen, on the plot path my brother passed up. Later, they showed up to fight the main characters during a unified plot section. After that, the Machinery Children were defeated for good on a story branch my brother did not see. As a result, the Machinery Children were introduced off-screen, made a brief on-screen appearance, before they died off-screen.

The story becomes even more convoluted as a result of how the characters are divided up during split path sections. First off, there are two rough plot paths that the player chooses between during each split, based on the teams who go on the different missions: the Super Robot path (which always includes Kyosuke Nanbu, the main character) and the Real Robot path (which always includes the SRX team). However, the characters who make up these teams, and thus the sub-plots the teams deal with, are not consistent.

For example, a major sub-plot of the game is the rivalry between Sanger Sonvolt and Wodan Ymir. Wodan Ymir first appears on the Super robot route during the first split, and the first time you get to fight him (and where Sanger first show up) is again on the Super Robot route during the second split path section of the game. Then Sanger and Wodan both play a prominent role during the third unified section of the game. However, Sanger and Wodan finish their decisive battle during the third split path section on the Real Robot route.

Because of complications like this, the player is 100% guaranteed to run across the weird plot problems I outlined above. The player will never be able to see more than a few plot threads go from introduction, through development, to final resolution in one go through the game. And since the player can alternate between following the two different general routes, it is possible for the player to miss out on parts of every plot thread in the game.

This system has a lot of room for improvement. A good model to follow might be Final Fantasy VI's split scenario system, where the player had to go through all three paths during the one time the game split into multiple paths. A similar example is in Suikoden 3, where the entire first half of the game consists of split paths, where the player has to follow all of the routes to completion in order to reach the second half. If the Super Robot series let the player see both paths, then playing through OG2 would be a lot less confusing.

Super Smash Bros. Brawl: Final Smashes

One of the biggest additions in Super Smash Bros. Brawl is easily the Smash Ball and the character-specific Final Smash attacks that the glowing orb unlocks. Because of the incredible power of Final Smashes, the Smash Ball radically changes the flow of a match each time it appears. The item's movement is quick and erratic, forcing anyone who wants to use a Final Smash to drop what they were doing and chase after it, knocking away rivals while trying to grab Smash Ball themselves. The winner of this miniature competition gets access to a unique move that, if used properly, can radically change the course of a battle. I think the whole thing is a lot of fun.

Here are some things I like about the Final Smash system:

1) The Smash Ball's erratic movements, and the fact that it must be hit several times for someone to acquire a Final Smash, are both really good ideas. The chase these elements create makes it feel like whoever acquires a Smash Ball has earned it, and it makes the whole system is fair to every player. If a Smash Ball was just an item laying on the ground that could easily be picked up, it would not be as exciting, and it might lead to people feeling that luck was more of a factor than skill in the game,

2) I really like that Final Smashes are character specific. In previous iterations of the series, all of the most powerful attacks were items, mostly rare Pokemon, and were not associated with characters. Final Smashes put the emphasis back on the characters, which I think is a good thing.

3) A person can be awarded a Mercy Final Smash if they are badly losing a match. When I first heard about this I didn't like the idea, but after seeing it in play I changed my mind. This only occurs if a player is really far behind in score, usually so far that a single Final Smash will not be enough to turn their luck around. Even so, it gives that player a small hope, which can lead to dramatic comebacks (or even just a good attempt at one) and more interesting matches. Of course, realizing you were just given a Mercy Final smash can be disheartening as well...

4) I think that it is a good thing that you can knock a Smash Ball out of a character and prevent a Final Smash. It means that even the most powerful attacks can be countered, and rewards skill and aggression. However, this also leads to some imbalance, since there are many characters for whom this is a severe weakness, such as characters who need to carefully position before using a final smash like Ike or Meta-Knight, and others who don't need to fear it at all, such as transformers like Bowser or Sonic who can use a Final smash immediately.

Anyways, on to a few character-specific observations and comments.

I like the idea behind Samus's Zero Laser. The concept of a Final Smash that is somewhat more powerful than others (which I think the Zero Laser is), but can't be used often because of the character switch, is a solid one. As a person who heavily used Samus in Melee, I don't like Zero Suit Samus much, so I am not terribly happy with Samus being the one given such a Final Smash, but that is just a personal annoyance more than anything.

I think it was a serious mistake to allow any repetition in Final Smashes. Having three character all use the Landmaster Final Smash just emphasizes the fact that the Star Fox characters are too similar. Ness using PK Starstorm, a move he never learned in Earthbound, feels a bit forced, especially when Ness doesn't even use his own signature PSI ability (PK "Favorite Thing"). The fact that Toon Link uses the same Final Smash as Link feels like a wasted opportunity (maybe the very strong whirling slash from Minish Cap would have been better?).

I really like the cinematic Final Smash that Captain Falcon uses. I would have liked to see a few more that did something like that.

I think that either Marth or Ike should have had a Final Smash that emphasized that they were leaders of armies. King Dedede and Pit both summoned large armies of minions, so I see no reason that it wouldn't work for Marth to do so (I say Marth because his move is a lot less impressive than Ike's). A rain of arrows and spells followed by a Pegasus Knight charge would have been a lot of fun to watch.

I think it would have probably been more interesting if the Pokemon Trainer summoned a special Pokemon rather than use the Triple Finish Final Smash. It would have been a good way to showcase a powerful legendary Pokemon (like Mewtwo) at an impressive level of strength.

Other than that, I guess all I can say is that I generally like the various Final Smashes. They are a lot of fun.

Saturday, March 15, 2008

SSBB: Subspace Emissary, part 2 and Stories in Fighting Games

A story mode can be a very useful thing for a fighting game. Even though fighting games are designed to primarily be a multiplayer versus mode experience, a single player (or two-player in Brawl's case) story mode can help draw a player into the game and further both their enjoyment and dedication to a game.

A really good example of this is Guilty Gear X2's story mode. Certainly, it is a very different experience from Subspace Emissary; GGX2's story mode is simply a progression of one-on-one fights with dialogue taking place between fights to further the plot, as well as a few branching paths in each character's story. However, GGX2's story mode accomplishes something very important: it fleshes out the characters and world way beyond the norm for fighting games. After playing through the story mode, a player develops a better understanding of the characters and becomes more attached to them.

This effect can also be seen in FPS games like Halo. First person shooters are built around multiplayer deathmatches, much like fighting games. However, Halo's story mode is one of its strongest features, even though it involves very different mechanics and structure than the it's multiplayer mode, much like Subspace Emissary does. Without its epic story fleshing out Master Chief's character, the SPARTANs would look like just another kind of space marine. However, that story did create and develop popular characters and built up a strong fan-base for the franchise.

One can argue that this treatment is unnecessary for a game series like Super Smash Bros. which is based on a cross-over of established characters from other series. These characters already have an identity of their own. However, not everyone has played all of the games these characters are from. For example, Lucas and Marth are from games that so far have not been released outside of Japan, and other characters like Pit and Ice Climbers are from very old games. So, it can't be assumed that a player necessarily knows anything about most of the characters. Furthermore, some of these characters (such as the two older ones I mentioned above) were never given much in the way of character depth or personality in their original games. So, there is merit in trying to establish who the various characters in the game are, even if it is just to raise the player's interest in other games.

So, how well does Subspace Emissary do this? Despite having no dialogue whatsoever, I think it does do a pretty good job. The cinematic scenes manage to inject a lot of personality into many of the characters. My favorite example is a scene where Ike, Marth, and Meta Knight observe a monster passing below them as they stand atop a cliff. Ike immediately jumps off the cliff to pursue, leaving Meta Knight and Marth staring at each other in surprise. Meta Knight then turns and follows, while Marth looks rather exasperated for a few moments. This scene establishes that Ike is a gung-ho type of character, while Marth is more of a "let's think things through" type. There are plenty of scenes like this through Subspace Emissary.

However, Subspace Emissary does not do everything perfectly. Most of all, each individual character does not get much screen time. Four characters in particular only appear in brief intro scenes. I would have probably followed in the footsteps of Guilty Gear X2 and given each character their own unique story path. That way, every character gets some time in the sun and a chance to be developed a bit. It would have been a good excuse to throw in some more mode completion character trophies too.

On that note, some of the development of character identity is done by the descriptions of various trophies. However, the discussion of Smash Bros. trophies might be worth a post in of itself.

Friday, March 14, 2008

Super Smash Bros. Brawl Characters

Well, as I continue to play and write about Super Smash Bros. Brawl, one thing that I have to comment on is the list of characters. There are even more characters in this game than in Super Smash Bros. Melee, which is obviously a good thing considering that the series is based around the idea of showcasing the good characters from throughout Nintendo's library of games. However, the details of the character list is something that I am not completely satisfied with. I guess I will break this down by series, just to keep things manageable.

Mario Bros.: I've got no complaints with the full cast of Mario, Luigi, Peach, Bowser, Yoshi, and Wario. Other than minor characters and characters who only appear in a single game, it is a very complete list. Adding any more would probably have been excessive.

Donkey Kong: Adding Diddy Kong to the game was a nice touch, though I still would have preferred if Dixie Kong also was in the game, simply because she was just as important to the great Donkey Kong Country games as Diddy (and even more important than Donkey, oddly enough). No really justifiable complaints here, though. I am actually rather glad they kept K. Rool out of the Smash Bros. series so far, because he is such a blatant and shallow derivation of Bowser that his presence would not add very much.

Metroid: I am glad that they (essentially) just stuck with Samus as a character to represent the Metroid series. It is a series that is entirely held together by only that one character, so I think trying to make any other characters iconic would just hurt the series. Of course, Zero Suit Samus is still Samus, so her addition is not a big problem.

Legend of Zelda: Here is where my real complaining starts. Link is good (though a Wolf form would have been nice, considering the Twilight Princess focus of this game), Zelda is good (though I feel that Sheik, a character who only appeared in one game and is unlikely to return any time soon, is probably unneeded), but Ganondorf is still a clone. They differentiated him from Captain Falcon a little more, but the Brawl version is still so far from any version of Ganondorf or Ganon in the Legend of Zelda series that he just doesn't seem to fit in at all. I think that making him more into a sorcerer like he was in Ocarina of Time, or a Swordsman like in he was in Windwaker, would have worked better. Finally, I like the concept of and addition of Toon Link, but his name is terrible.

Star Fox: Unfortunately, the Star Fox character list is far from perfect. Fox is still good and is an obvious choice, but it was a poor decision to have him be accompanied by both Falco and Wolf. Both of those characters are a bit too similar, since both are aggressive and confident rivals to Fox, and neither of them can easily be made distinct from Fox easily. In this case, I think Falco is probably the one who needs to go, since his moveset is far less original and interesting than Wolf's, and I approve of more villain characters getting a spotlight. It probably would have been better for Krystal to be the third Star Fox character, using the staff from Star Fox Adventures and a different moveset entirely. Certainly, the Smash Bros. series could use more good female characters.

Pokemon: The addition of Pokemon Trainer is such a good idea that I almost wonder why there are Pokemon separate from trainers at all. Pikachu and Jigglypuff could easily be integrated into being part of a trainer's line-up. I think there should be three trainers: the Male Trainer, a Female Trainer, and a Rival Trainer (a set-up which is proven to be central to the franchise in the main Pokemon manga), each with three different Pokemon (maybe each having one of the starters?). Beyond that wish, I have few problems with the current set of Pokemon characters.

Mother/Earthbound: I really think they should have picked either Lucas or Ness, and just stuck with one. Both are interesting, but there are just too similar. If they wanted to have more than one Earthbound character, it probably would have been better to add character who is different from Ness or Lucas, like Poo, or perhaps add a villain character.

Kirby: I think this set is pretty good. In fact, we probably should have gotten either Meta-Knight or King Dedede back in Melee. Those three characters are the three most important and memorable recurring characters in the Kirby franchise, so they really do all deserve to be there and they work well as Smash Bros. characters.

Fire Emblem: This is one group of characters that I am the least satisfied with. Anyone who reads the archives would know that I am a big fan of the Fire Emblem series, so I suppose I might be a bit biased, but just having Ike and Marth seems to be far too little for a series that has such a huge cast of interesting and varied characters. Ike and Marth are differentiated very well in Brawl, much better than Marth and Roy were, but they are both sword-using Lord characters in the Fire Emblem games, and there are many other major types of characters in fire Emblem who deserve a spot in a Smash Bros. game. Even ignoring the many characters who ride horses and wyverns, non-combatant healers, wimpy thieves, and swordsmen who would act like Lord clones, there are still many character archetypes in Fire Emblem that have recurring importance, such as Wind Mages (like Soren, Maric, and the Holsety users), female Light Mages (like Micaiah, Linda, and Julia), villainous Dark Mages (like Garnef, Manfroy, Nergal, and Izuka), Manakete who can take on a draconic form (Chiki and Myrhh), and the newer Laguz from the recent games, as well as various other interesting villains like Ike's nemesis the Black Knight or Marth's final foe (twice), Medius the Dark Dragon. Any one of these would add some variety to the Smash Bros. series. At the very least, there could have been more of these characters as assist trophies or as Subspace Emissary bosses.

Retro Characters: I have no complaints here. ROB, Mr. Game and Watch, the Ice Climbers, and Pit are all good characters. I wouldn't like having too many more retro characters, because I don't think it is a good idea to have a lot of forgotten characters squeeze out characters with long-standing popularity, but having a few around to celebrate Nintendo's long history is fine.

Others: Captian Falcon and Olimar are both good characters, so I have no complaints here, either. Well, I would have liked to see a Custom Robo character, but that is a minor complaint.

Third Party: There just should have been more third party characters in Brawl. Ever since Snake was announced, people have been desperately hoping for more third-party characters to be announced, but they ended up with only two, even fewer than the number Mr. Sakurai stated that he intended to put in the game. Certainly, characters who have been as iconic to the Nintendo console gaming experience as any first-party character, like Megaman or Simon Belmont, deserve to be in the Smash Bros. series. That said, I am just glad that there are any Third Party characters at all.

Overall, I really think the Smash Bros. series could use a bit more focus on character variety. The Subspace Emissary mode highlights the lack of female characters, especially heroic female characters, pretty dramatically (two of the three female characters are perpetual Damsels in Distress), and it suffers from the lack of villains (only Wario, Ganondorf, and Bowser, really, especially since Meta-Knight and King Dedede are not strictly villainous). More variety would make characters more interesting and add flexibility to possible future story modes, and it would also help avoid the perpetual Smash Bros. problem of too many characters having overly similar mechanics and movesets.

Thursday, March 13, 2008

Super Smash Bros. Brawl: Subspace Emissary, part 1

My brother and I completed Subspace Emissary mode in Brawl yesterday, so it is time for my thoughts. I actually really liked the inclusion of a fully-fleshed out story mode in the game. While Brawl is built around its multiplayer versus mode, the adventure mode does add a lot to the game experience. However, I don't think Subspace Emissary is perfect in its execution.

My brother and I played through the entire Subspace Emissary mode using the 2-player co-op option, and my biggest complaint is that the implementation of this was sometimes shaky. In Subspace Emissary, the second player is limited to playing a helper character, while player one is the main character. This is obviously based on a similar mechanic in Kirby Super Star. However, unlike in Kirby Super Star, where the helper controlled by the second player is completely disposable, since it can be easily regenerated by Kirby, the second player character in Subspace Emissary uses the same extra-life count as the first player. So if the helper dies three times, the first player has no extra lives left for himself.

What this means is that, unlike in Kirby Super Star, the two players are a lot more even in expendability. There are four real differences between player one and player two: the screen is always centered on player one, player two can warp to player one's position at any time, only player one can open doors, and the game ends if player one dies and the team has no extra lives, but not if player two dies. Now then, the last one is particularly annoying. There are points in the game, such as the meat-grinder of a final battle, where I might have had a shot at winning the battle if I could have kept fighting after my brother (playing as player one) was eliminated. However, the game arbitrarily prevented that possibility, even though my brother did get to keep fighting a boss after I was eliminated. That experience just feels unfair and pointless.

The camera always being centered on player one was also annoying, particularly when it resulted in my character's death when my brother accidentally ran too fast. While the warp gets around most of these problems, it still prevents player two from fighting enemies or collecting items too far away from player one. Game code that controls the camera zoom and position to accommodate multiple players is used all of the time in versus mode, so making some allowances for the second player to move independently of the first player should have been technically possible.

Anyways, the biggest problem in Subspace Emissary is that it has a lot of very tricky platforming elements involving bottomless pits and deathtraps in a fighting game. While there is a lot of jumping around in battles, many of the characters in the game, particularly the heavy-weights, were simply not built to handle the challenging jumps that sometimes appear in Subspace Emissary. Since there are over 30 characters in Brawl, each of whom has different jumping capabilites, there is no way to balance every trap for every character. As a result, my brother and I lost more lives to bottomless pits than to the battles in the game, which doesn't seem right. Since Super Smash Bros. is primarily a fighting game, it might have been better to make the adventure mode more of an old-school brawler, like Final Fight. The parts of the game that focus on those elements more flow better than the platforming ones.

I have to admit that I like the team focus of the game, where extra lives represented other characters coming in to reinforce the fighting team. It gives the players more chances to play with different members of the game's considerable cast. Breaking the characters into teams for most of the story also worked very well. However, I guess that I will cut this post short, and talk about the story, writting, and directing of Subspace Emissary in my next post.

Wednesday, March 12, 2008

Alternate Costumes

Amazingly enough, this blog post will not be about Super Smash Bros. Brawl (though I suppose it is tangentially related). Instead, want to talk about a more general concept of alternate costumes for characters, particularly in RPGs. I would talk about their role in fighting games and the like, but that is straightforward enough that I don't think I have ever seen it done in an ineffective manner (though the quality of the costumes themselves is always up to debate in some games). However, in RPGs, alternate costumes tend to be poorly implemented.

One major problem facing alternate costumes in an RPG is that choosing a character's costume is not a choice made independent of other concerns. For example, in Ar tonelico, the Reyvateil's alternate costumes are equipment that control stats. In Tales of Symphonia, you need to give up a title that boosts your stat gains at level-up in order to use an alternate costume. In Xenosaga Episodes 2 and 3, alternate costumes take up skill slots (in Episode 2) or an equipment slot (in Episode 3), and while they have stat benefits, they tend to be sub-optimal choices in the long run. In all of these cases, alternate costumes may as well not exist as an option, because players tend to go for options that give the most mechanical benefit, rather than the options with the most pleasing aesthetics. However, other games demonstrate that such a choice is not necessary. For example, Star Ocean: Till the End of Time has a system where you can simply change character costume between several different options independently from any other choice, simply by disconnecting it from any other system. Another interesting option is seen in Wild ARMs 5, where costume is linked to armor, but there are whole sets of armor that give the same alternate costume and similar types of stat benefits. The implementation is not perfect in that game, but the idea of linking different costumes to a sets of armor with different stat benefits is a good one.

A related issue to the above is the direct linking of the appearance of a character's clothing to equipment, seen in many MMORPGs and some other games, like the PS1 RPG Legend of Legaia. I suppose that there is no problem with this in theory, but in execution the effect tends to be flawed. Too often, a player character looks terrible because of mechanical reasons such as optimal sets of gear being created by mixing bits and pieces from different armor sets, coherent sets being difficult to assemble, or coherent sets not existing in the first place. One solution is for the designers to design armor sets so that optimal sets of equipment match together well into a decent-looking character model, and to avoid creating situations where the best option is to mix gear from very different looking sets of equipment. Of course, a good alternative is to use the same system as the Phantasy Star Online games and separate equipment (which constantly changes) and character appearance (which is set when a character is first created), so they are not dependent on each other.

One other thing worth noting about console RPGs regarding alternate costumes is that most of the time they seem to only change for battle scenes, but not for field screens or story scenes, which can be somewhat jarring and annoying. I suppose this is probably a technical limitation, and it might be already be on its way to being a thing of the past if some of what I hear about Lost Odyssey is true, but it is worth mentioning that having that kind of difference between battle scenes and everything else should be avoided if possible.

Anyways, a lot of what wrote about concerning RPGs can be applied to other kinds of games, as well. For example, you can chose alternate costumes in Drakengard, but you have to equip a costume-type Orb as your only accessory to do so, which means you need to give up one of the defensive items or other powerful accessories that are necessary to survival in that game, making an alternate costume a very poor choice (as a result, I never used one, no matter how much I liked the alternate looks themselves). In contrast, the Way of the Samurai games make appearance something you can control whenever you start a new game, and you are free to chose whatever you like without worrying about affecting something important. Also, as I said before, fighting ames tend to get this right almost every time, and set a good example. Overall, I think the freedom to make a character look cool is more important and valuable than the mechanical variety added by making it a choice based on opportunity cost.

Tuesday, March 11, 2008

Unlockable Content

Super Smash Bros. Brawl apparently has a lot of unlockable content. With secret  characters, hidden stages, unlockable music, and special trophies all to collect, I am beginning to think that it will take a very long time to unlock everything in the game. Thinking about it though, why do game designers force players to jump through hoops in order to unlock certain kinds of content? Obviously the designers at Nintendo could have simply made every stage and character available from the get go in Brawl. So why did they choose to make the player have to fulfill all kinds of unusual tasks in order to unlock content for multiplayer matches?

A couple months ago, while writing one of my posts on Mario & Sonic at the Olympic Games, I realized that forcing the player to unlock certain events in that game was a very interesting way of making the player play the game long enough to get used to the controls and learn to enjoy playing the game. As I stated in that post, the game takes a little getting used to, and using the dangling carrot of unlockable content to coerce the player to practice at various events enough to win gold medals was a pretty good method of solving that problem.

In one sense, this use for unlockable content is a sub-set of the more general purpose: rewarding the player for putting a lot of time and effort into the game. A good example of this is Disgaea, which has at least two characters that can be recruited by diving very deep into the game's high-level post-game. While beating a game and watching the credits roll can give a player a lot of satisfaction, most games nowadays come with a lot of optional content, and many different scales of difficulty. Accomplishing a very difficult task and getting no kind of reward for it is a very frustrating experience for a player. Unlockable content generally works very well as a reward like this.

A very good example of a game that uses unlockable content right is Devil May Cry 3. Devil May Cry 3 is a pretty straight-forward action game. Once the player has beaten the game, there is not much more to do in the game except attempt to beat the game again on a higher difficulty level. As a result, the game has unlockable costumes, game modes (new difficulty settings and Heaven or Hell mode), artwork, and even a god-mode (Super Dante). In Devil May Cry 3, these unlockable features serve the main purpose of rewarding players who put in the time and effort mastering the game to the degree required to beat such a hard game on its hardest difficulty settings. They can also serve as bait, to try and get players to attempt higher difficulty settings.

So then, what is the point of making a significant portion of the cast and stages in Brawl unlockable? I think in part it is marketing. There has been a lot of buzz built up over the last year about which characters would be in Brawl, even to the point where people created hoaxes to fill in the information gap. Having secret characters was a good way to help build this speculation and interest in the game. The rest of the reason is reward. Most of the stages and music are handed out as rewards for players who take the time to clear event matches and other optional content in the game.

Monday, March 10, 2008

Fighting Game Controls

The fact that I didn't write anything here yesterday, the day that Super Smash Bros. Brawl was released, is not a coincidence. That game is horribly addicting, just like its predecessor. My brother and I occasionally challenged each other in Super Smash Bros. Melee all the way from its release until a week ago, and we enjoyed every minute of it. It is a game that kept on entertaining even long after we have explored every single-player game mode and fought hundreds and hundreds of matches in its multiplayer mode.

Part of the reason I enjoyed Super Smash Bros. Melee so much over the last six years is because it is a game with incredibly easy controls. Every character has a limited number of normal attacks which are controlled only by combining the A button with directions on the control pad, a list of 4 special moves controlled with the B button, a few throws, dodges, and a single defensive stance that can be used against every kind of attack. It is very easy to learn how to play the game because of the simplicity, and because every character is controlled using the exact same inputs, it is very easy to fully understand multiple characters. No matter how long it has been since I last played the game, I can still remember all of the controls and all of my favorite moves with each of my favorite characters.

As a comparison, I want to mention another fighting game series that I like quite a bit: the Soul Calibur series. I like this series, own both Soul Calibur 2 and 3, and spent quite a bit of time and money playing Soul Calibur in the arcade, but I never played this series anywhere near as much as I played Super Smash Bros. Melee. In part, I think this is because of the difficulty of this series' controls. Every character has a long and elaborate movelist of dozens of completely unique moves, each of which has its own button input (which makes learning new characters the same thing as relearning all of the controls for the game), sometimes involving complex chains of difficult timing and/or tricky sequences like slide input. As such, I can never seem to remember even the basics of a character's moveset for very long. I need to refresh my memory with a half-hour long trip to the training mode practically every time I go to play the single-player mode, I need to occasionally check movelists in the middle of battle, and I can never remember a thing about a character's moveset after putting the game down longer than a few weeks. This makes coming back to the game somewhat intimidating, which has resulted in me not playing Soul Calibur 3 ever since I first stopped shortly after the game was originally released years ago, even though I left it incomplete and like the series quite a bit.

Another game I could compare Melee to is the Tekken series. I rented a Tekken game only once, and pretty much gave up on playing it within a day, after getting stuck, unable to continue, in a few characters' training modes. I think another major fighting game series, the Dead or Alive series, is a lot of fun, but much like with the Soul Calibur games, I can't seem to ever get back to playing Dead or Alive games once I have stopped playing for more than a week.

One series I can keep playing somewhat consistently is the Guilty Gear series, even though its level of general complexity because of various subsystems is fairly high. The game has all kinds of gauges and unusual rules you need to remember (including really unusual things like the Guard Meter, the limitation on combos, and the penalty you receive for not being aggressive enough), but the controls are not terribly complex. Every character has an analogous set of attacks based on the main attacking buttons, so all you need to remember for the various characters are the special moves and their inputs, which mostly are derived from a very limited set of possible inputs and follow certain learnable patterns. While it is not as quick as with Super Smash Bros, I can get back into playing Guilty Gear well within a very short time, so I return to the series fairly often (keeping with it is a different story, since playing it too much gives my thumb blisters).

Overall, I think a good fighting game should emulate the classic series, like the Street Fighter series, and keep things relatively simple to control. There should be as much similarity in control inputs between characters as possible, and these should be kept relatively simple. Requiring many complex button inputs for a character to even be decently viable just results in a game that is only appreciated by a very dedicated fanbase of very limited size, because it excludes too many gamers like me who like the genre, but are not so big of fans that we are willing to dedicate huge amounts of time and effort to play.

Friday, March 7, 2008

Metal Gear Solid Codec calls

In my last post on extraneous back-ground information and encyclopedias, I made an off-hand reference to Metal Gear Solid's Codec system as a means of delivering information to the player. I decided that the Codec system is exceptional enough to warrant its own post. In many ways, it is a very good system for both allowing the player to learn more about the game world and for developing the cast of characters.

In some ways, the Codec is simply a substitute for towns full of random NPCs a player can talk to. You simply choose an NPC to talk to and then listen to the NPC either give you some helpful advice or blab about something pointless, and is used primarily as a means of reminding the player of their current goal in the game. However, the implementation of the Codec system elevates it to a much higher level than just talking with NPCs, like what is seen in most other games.

First off, the most distinguishing aspect of the Codec is that the player can use it to talk to the support cast whenever the player wants, whether exploring a new area, sneaking around, or in the middle of a boss battle. So, the player can interact with the NPCs whenever the player wants to. The player can talk with the supporting cast whenever he feels like it, without having to go to a specific location or anything. Furthermore, this lets the player ask the NPCs for advice even during a boss battle. For example, if a player can't figure out how to defeat a boss like Vulcan Raven in the first Metal Gear Solid, he can always go to Colonel Campbell for a few tips. This makes the Codec characters generally much more useful than typical hint-dropping NPCs, since the player can turn to them for advice when he actually needs it.

And it really does feel like the player can ask the Codec characters questions. For example, if you equip a grenade in Metal Gear Solid 3 and call Sigint up, he will talk to Snake about grenades. Ditto for various other items and equipment in the game. This lets the player intentionally equip various items to provoke various different conversations with Sigint (or whoever else is giving this kind of information in other Metal Gear games), thereby letting the player "ask" him about things.

On top of that, the best part of Metal Gear Solid's Codec system is that the NPCs don't just give information to the player, they actually engage in a real conversation with the main character. Snake actually talks back to the NPCs, and they often go off on wild, yet very funny, tangents. A good example is how asking Para medic in MGS3 about any kind of animal results in Snake asking how good it tastes. Running jokes, conversations about pointless topics, and the like don't further the plot, but they do develop the characters. I have a theory that the best way to make a character likable is to have that character engage in pointless, mundane, silly, fun conversations. It is a great way to get to know characters.

A lot of the successful elements of the Codec system can be found in very different games. For example, I get a similar vibe from conversations in the Phoenix Wright: Ace Attorney series of games. Most of the time in any Phoenix Wright game, there is usually another character following the protagonist around. Whenever you examine something in an area, Phoenix and the other character usually end up talking about it. Furthermore, you really can ask NPCs about things by presenting them with evidence or asking about specific people. These questions also provoke involved, yet entertaining dialogues with the main character, that often involve running jokes and over-the top personalities. It is no wonder that the Phoenix Wright series and the Metal Gear franchise are well known for having memorable and well-developed casts of characters, even though they are very different games.

Thursday, March 6, 2008

Ar Tonelico Part 4

Game Completion: Pretty much full completion of the Misha path through the game.

Well, I must say Ar tonelico took me quite a bit longer than I expected. What I thought was about 50% ended up being more like 30%. In fact, I think this game is in some ways too long of a game, something which I really need to explain more.

Ar tonelico's plot is split into three phases, each about twenty hours long. It actually has a pretty good structure. The first phase introduces the main characters, explores most of the setting, and establishes the central conflict very successfully. In the second phase, the plot splits into two different paths for most of the phase. At the end of the second phase is a "false" ending (which rather kindly unlocks all the post-game bonus features, which is a nice touch). The third phase has a major change to the cast and mostly re-unifies the two paths, leading up into a proper final battle. What is more, in each phase of the game there is a noticeable increase in the difficulty level (though in this game's case, it is a change from incredibly easy, to easy, to having a small chance of providing a challenge). At its core, this is a solid three-act structure that can be very useful. Unfortunately, it doesn't work very well in this game.

The problem with the structure and pacing of this game is that the third part of the game feels forced and unnecessary, and contradicts and weakens the plot of the previous parts. The center of the story of this game is the love triangle between Lyner, Aurica, and Misha, and it works very well up until the end of the second phase. The split path in the second phase is a choice between an Aurica route or a Misha route, and this split works really well. The (in my opinion, annoying) uncertainty of the love triangle fades away in the second phase, replaced by a visible growing relationship between Lyner and the chosen girl, ending with them falling in love and staying together in the false ending. However, the problem is that this good plot development only really ends properly in a false ending, which is obviously only a part of the way through the game and is very unsatisfying. The beginning of the third phase is based on the hero suddenly fixating on a third girl, Shurelia (a minor character up to that point who sacrificed herself at the end of the second phase), which leads to the reintroduction of the love triangle (with Shurelia added), pretty much ignoring the character development of the second phase. The choices you made in the second phase have an effect again at the true ending of the game, but up until that point the plot of the third phase is extremely focused on the third girl, Shurelia, to the exclusion of the more likable Aurica and Misha who have been developed from the start.

I think Ar tonelico would be a better game if it did not transform Shurelia into a major character so late into the game, after the love triangle had already been resolved. She should probably have either remained a minor character (an ideal for Aurica and Misha to pursue, I suppose), or have been a major character and romance interest for Lyner from the start. Also, building the false ending and third phase around Shurelia specifically puts too much emphasis on her and leaves the false ending as too happy of one, so I would probably have the false ending be based on the girl you have chosen up until that point. For example, the point in the second phase where Misha resumes her role as the Star Singer, singing eternally to seal away the villain, is very similar to the situation Shurelia is placed in for the false ending, and is actually much more dramatic (in fact, it is one of the best scenes in the game). The game cheapens that scene on the Misha route by quickly undoing Misha's sacrifice, so I think that putting a scene like that as the false ending would be more appropriate than leaving it in the middle of a phase.

I suppose that is enough about my complaints concerning the plot. On to the things I liked.

First, I must say that, appropriately enough for a game built around something called Song Magic, the music in Ar tonelico is very good. There are a few tracks that are absolutely terrible (the song played whenever a character is being goofy is jarring and painful), but these are rare, and there are many, many good songs. In particular, the plot important "Hymn Crystal" songs are incredibly good and add a lot to the game. I just wish there was a greater link between activating normal Song Magic in battle and actual music.

Second, I still like the Reyvateil characters (even the third, Shurelia) and their Cosmospheres. Shurelia's Cosmosphere doesn't have the surrealistic charm of the others', but it has more real plot and actual battles, which makes it interesting enough. I wish the strong points of Shurelia's long plot and battle system was combined better with the way the other Cosmopsheres reflected personality and subconscious thought.

Third, despite a few places where the plot gets a bit preachy and cliche in the third phase, I like the way the game breaks from the mold of having characters defeat villains with brute force all of the time. The game could have set up Mir as a sympathetic villain a bit better (it is amazing that they were able to do so at all, considering she is so single-minded on total genocide), but the events of the actual final battle, in which the heros try to convince Mir to end her rampage, is a good sequence and matches the basic premise of the game (the hero helping Reyvateils heal their mental scars) very well.

I am still curious about the Aurica route through the game, but I will probably set this game down for a while. I have been holding back a few other really good games far longer than I intended.

Wednesday, March 5, 2008


One feature of various video-games that I always appreciate is a simple glossary that the player can access at any time. Some various games that have had this kind of feature include Xenosaga Episodes 1 and 3, all three Metroid Prime games, Final Fantasy XII, and several others. I have always though that a glossary can be a very useful feature in a videogame with a complex world.

One of the best functions of a glossary is that it serves as a place to put background information in a game without cluttering up the writing on the game's main plot. It can be really easy to go a bit overboard and write a lot of background information for a game, particularly RPGs and other games built around having lots of complex characters. However, if a developer tried to put all of that information into the text of the main plot, the game's narrative would get horribly bogged down by unimportant minutia. However, there are plenty of people who like knowing that kind of information, so it isn't a bad idea to make all of that background information available to the player in some optional manner.

There have been quite a few methods of giving this information to the player over the years. One of the oldest is putting this background information into the dialog of random townspeople in RPGs. Another variation are the optional conversations with support characters in the Metal Gear Solid games. It is possible to acquire an incredible amount of completely useless and trivial, yet entertaining, information about every place, character, and item in those games from these conversations.

However, I tend to prefer having an accessible glossary to be more helpful than those other methods for one primary reason: it can be referenced at any time in the game. In most RPGs and the Metal Gear Solid games, it is possible to miss conversations or pass the point where you can go back and read them again. A glossary lets the player refresh their memory of things at any time. This can be useful for a player who plays a single game over a large span of time, particularly if the glossary includes some kind of ongoing plot summary.

Another useful feature of a glossary that can't be replicated by simple NPC conversations is that it can easily include extras such as artwork and character models. Oftentimes, it can be difficult to pay attention to detailed enemy models during gameplay, particularly since zoomed-in camera angles are rare during any kind of combat. So, it is nice to be able to look at the model of a main character, monster, NPC, or item in a glossary section at one's leisure. I would have liked a feature like this in Front Mission 4, since I rarely get a chance to look at the various mechs in that game up close. A glossary would thus be a useful feature (particularly since Front Mission 3 actually gave a lot of background information and artwork as part of its Network feature, so there would be precedent).

While a glossary is a kind of extra feature, it is the kind of extra feature I always like seeing in videogames.

Drakengard: Caim and the Red Dragon

One thing from the Drakengard games that I have really wanted to discuss is how characters from the first game are used in Drakengard 2, and how the characters are shown to grow and develop. Since Drakengard 2 only follows the first ending from the original game, I will only discuss how the characters are portrayed in that path of first game's plot.

The hero of the first game, Caim, is a man of complexities and contradictions. For one thing, he hates the Empire and dragons because his parents were killed by one of the Empire's black dragons some time previously. Of course, this kind of hatred is an over-used cliche, but unlike in the cliche situation, this hatred is not a minor attribute of the character, it is one of his most defining aspects. Caim's rage is so all-consuming that he is nearly self-destructive in his aggression and is merciless to his foes. His violent nature is so extreme that it terrifies his allies and makes the even the red dragon comment on his cruelty, such as the time he slaughters the Empire's child soldiers. At the same time, Caim is a survivor who does not die without a fight. Rather than die fighting a hopeless battle against the Empire, he makes a pact with one of the dragons he hates so much, bonding his life to hers. The interesting thing is that he is as much driven by his hatred of dragons in that act (in which he at first tries to coerce the dragon into obeying him) as he was defying that hatred. His hatred is terrible, but at the same time it is what drives him on to survive and gives him the strength to defeat his enemy (which is portrayed so monstrously that you never feel unustified in hating it). As such, the game doesn't condemn Caim's hatred, nor does it portray it as a necessarily negative quality.

Adding on to the complications of Caim's character is his genuine protective nature. Caim is a man who looks out for other people and protects them, and this is clearly seen with his desire to protect his sister Furiae. He fights to protect the world, and he seems to have a strong sense of justice. These attributes drive him as much as his hatred does, even though they seem like very contradictory things at first glance (certainly, the way Caim both spares Manah's life and condemns her to life of suffering is a sign of this contradiction, though perhaps more of his just side). Of course, anything more clear than that gets a bit difficult to describe, because after only a short time in the game, Caim becomes completely mute due to his pact with the red dragon, the relationship which easily defines the first game.

The red dragon is certainly one of the most unique and fascinating characters I have seen in videogames. Certainly, one of the most unusual things about her is the fact that she is probably the only female dragon I have ever seen which is portrayed as a powerful creature capable of widescale destruction (female dragons are rare enough). Added onto this is the combination of her pride, honest cynicism and hatred for humans (refusing to even tell humans her name), her lines being all voiced by a very good voice-actress, and her being the most talkative character in the game, resulting in her being a very memorable and entertaining who does not fit into any typical mold. Her talkativeness is in fact one of her most important qualities. Because Caim can't speak, the red dragon pretty much does all of his speaking for him, while also constantly commenting on his actions (usually mocking him for his human flaws and violent cruelty) and expressing her own feelings and opinions. In a sense, she serves as the main narrative voice who describes the action and the characters (though she is certainly not an objective voice, since she expresses somewhat inhuman thought and a lot of disdain for the people around her).

Because the red dragon is so open about her disdain for humankind and pride in her draconic nature, and Caim hates dragon so much, the very act that the two form a pact is an interesting contradiction. In a sense, the fact that they share one life after making a pact is a metaphor for how the two have to give up their own identities as a human who hates dragons and as a dragon who hates humans as a result of their partnership. Caim progresses from a person who pretty much hates everything, especially dragons, to a person who trusts the red dragon, depends on her, and encourages her. The red dragon progresses from an aloof dragon who looks down on humans to someone who respects Caim's strength and is willing to fight against incredible odds on his behalf. The way Caim can encourage the red dragon to fight a legendary dragon by just rubbing her nose with his hand is a sign of that growth, and the red dragon's eventual sacrifice to save the world is the end result of that change.

The bond between Caim and the red dragon is shown to be even greater in the events of Drakengard 2. In this game both the Empire and Caim's sister are gone, leaving him neither his hatred or the person he wanted to protect. In their place are the Knights of the Seal who protect the world and the imprisoned red dragon who suffers as the final living seal. With personal hatred behind him and no more reason to act as the world's guardian, Caim becomes the One-Eyed Man, the enigmatic destroyer determined to save the red dragon, even if it means destroying the world in the process. Nowe, the hero of Drakengard 2, fights Caim to avenge his parents (an interesting repeat of the same motivation) and to save the world, but against Caim's determination to save his pact-partner, even Nowe's noble (and traditional) motivations seem hollow (a fact that Nowe comments upon himself). At the same time, the red dragon's hatred of mankind is revealed to have never been lessened, and has grown even worse (enough for her to try to burn the world), but her bond with Caim has never wavered. She sacrificed herself to be the world's seal to protect Caim, not to protect mankind. Both of the two were willing to give up the world for the other's sake, and while this results in a tragic end, it is still portrayed as a wonderful thing.

Anyways, I only have a few minor observations left. First, the games go with an unconventional theme of condemning naive ideals while making terribly violent actions to selfishly protect loved ones the role model, never condmending characters like Caim and often soflty condemning the ideals of people who hold back in battle where violence is needed and harshly condemning people who choose to sacrifice others rather than protect anything themselves. Also, while I didn't comment on them too much, the character Manah's development in these two games is fairly good itself, and the relationship between Nowe and the blue dragon Legna both serves as both a parallel and an opposite to the relationship between Caim and the red dragon.