Sunday, August 31, 2008

Megaman Starforce: Geo and Omega-Xis

My brother and I have been putting a lot of time into Megaman Starforce 2 (both versions) for the Nintendo DS over the last week. I have been a big fan of the Starforce series and the original Megaman Battle Network series for years now. However the two main characters of Megaman Starforce, Geo Stellar and Omega-Xis, have become my favorite characters out of the entire combined series. As a pair of characters, they regularly produce fun scenes and hilarious dialogue. Without their interactions, the two Megaman Starforce games would not be anywhere near as fun.

In the original Megaman Battle Network series, the equivalent characters to Geo and Omega-Xis were Lan Hikari and Megaman.exe (otherwise known as Hub, Lan's twin brother). Like most Net-Navis in the Battle Network series, Megaman.exe was in many ways an exaggerated or slightly distorted mirror-image of his operator. In general, Lan was the slightly more aggressive and confident of the two characters, while Megaman.exe tried to act as Lan's conscience. However, the two were both pretty straight-forward heroic characters, and thus were generally in sync with each other. Unfortunately, this meant that the character interactions between the two were often somewhat bland. While there were some truly touching scenes built around the fact that they were two brothers (found mainly during endgame sequences), many of their regular interactions often boiled down to Megaman.exe nagging Lan to do his homework.

In comparison, the conversations Geo Stellar and Omega-Xis have even during routine gameplay range from mildly entertaining to downright hilarious, thanks to the strongly contrasting personalities of the two characters. On one hand, Geo Stellar fits perfectly into the reluctant hero archetype: he is a rather timid boy who lacks drive and confidence who is dragged into dangerous situations against his will. However, Geo possesses strong morals and dreams about becoming someone like his father. In comparison, the alien warrior Omega-Xis is an incredibly aggressive character. Omega-Xis thrives on combat and excitement, is always eager for a fight, and has no qualms about doing things like poking around people's secret data files. It is difficult to imagine two characters more different than these two.

These contrasting personalities fuel a huge range of funny and entertaining interactions and conversations. One of the most distinct is the hilarious scenes where Omega-Xis (who dwells in the computer Geo wears on his left arm) literally dragging Geo around by the arm because Omega-Xis got bored and wanted to go somewhere more exciting. We see Geo protesting when Omega-Xis wants to invade someone's computer to take a peek at people's diaries, before Omega-Xis ignores him and does so anyways. And we see Omega-Xis constantly teasing Geo about the various girls who have crushes on him. Because both characters have such contradicting perspectives on everything, it is easy to write a wide variety of interesting material for them.

Perhaps the most useful outcome of Geo and Omega-Xis's personalities is that it gives a very good justification for why a reluctant hero like Geo would actually become a game's main character. In Megaman Starforce 1, Omega-Xis makes it clear to Geo from the get-go that he knows something about what happened to Geo's father, who mysteriously disappeared some time prior. So, Geo has a very good reason to put up with Omega-Xis. From that point, Omega-Xis simply drags Geo from one dangerous situation into another until Geo has slowly become accustomed to being a hero. So there was no point in the game where I asked myself "So why is this guy the main hero?".

In the end, I think the direction the developers took with the personalities of Geo and Omega-Xis was the right one. It has produced really entertaining results so far. While the risk that the writing for the game will get stuck in a rut still exists, I hope that these two characters will be able to avoid that fate.

Megaman Starforce 2: Battle Cards

I have probably mentioned this before, but both my brother and I are big fans of Capcom's Megaman series, and that includes the Battle Network and Starforce sub-series. As such, we have been playing Megaman Starforce 2 quite a bit over the last days, and I have really been enjoying it so far. I already have quite a bit to say about the game, however, since Battle Cards are so important to the game, they will be the first aspect of Megaman Starforce 2 I will discuss.

I really think that the Battle Card system of the Megaman Starforce games is a significant improvement over its predecessor, the Battle Chip system from the Megaman Battle Network games, mostly because it abandons the alphabetical code system. In the original Battle Chip system, every individual Battle Chip has a letter code in addition to the chip type, so you could have things like an "Cannon B" chip that is similar to, but slightly different from a "Cannon C" chip. All "Cannon" chips are used in the same way and have the same strength, but chips with different codes can be combined with different other chips. Because of that system, you can use a "Cannon B" chip the same turn as a "Protoman" chip (Protoman is always B), or you can use a "Cannon B" and a "Cannon C" at the same time, but you can't use a "Cannon C" chip and a "Protoman" chip at the same time. This means that there is an incredibly strong incentive to build Chip Folders using as few codes as possible, in order to maximize the number of Battle Chips you can use in one turn. This incentive is so strong that it practically overrides any other consideration, and often makes Folder creation an overly complicated endeavor. Also, because chips of particular codes may be more rare than others, it occasionally made gathering a good set of Battle Chips a lengthy process. This system worked well enough in the early Battle Network games, when building the Chip Folder was the player's only concern, but what advantages it had began to break down in later Battle Network games.

The big problem for the Battle Chip system came with the introduction of Style forms, Double Soul forms, and other such systems in later Battle Network games, as well as with the continued improvements to the Battle Chip list that came with each game. More and more, considerations like needing a large number of powerful Fire Chips to fuel the abilities of the FireSoul form or needing to include a mix of good Normal Chips and Electric Chips for ElecCross form began to create an incentive that contradicted the code-based incentive. More importantly, the code system made it difficult or impossible to freely use multiple forms with a single Folder. As a whole, the entire system ultimately promoted a limited set of "good" Folders based entirely on what was pre-determined by the game designers, rather than what the player desired.

Megaman Starforce 1 changed all of that when it completely abandoned the idea of letter codes and replaced it with the column rule. Now, in the Starforce series, Battle Cards are laid out as a grid of three columns of two cards rather than merely a list of available options, and cards that have been randomly assigned to be in the same column can be used together in the exact same manner as cards of the same code could be used together in the old system. As such, the player no longer has the option of gaining an advantage through careful Folder construction, which actually means the player is free to chose whether to use a Battle Card or not on the Card's merits alone. Unusual Card combinations are possible in a way they never were before, and Cards no longer have artificial synergy just because they share a letter code. What is more, these restrictions are eased while still preserving the original reason the code system was introduced: adding some strategic complexity by preventing the player from being able to use every card given to him in battle at once. In fact, it pulls it off more successfully because there is no way to work around that restriction in the Starforce system. It also preserves the "wild card" value of the old asterisk-code with the "White Card" designation. A lot of negative side-effects of the system were removed and almost all of the advantages were preserved.

One new feature of the Battle Card system that I find to be particularly good is the change made to the "favorite chip/card" system. Previously, the Favorite Chip was a Battle Chip that was guaranteed to be available at the start of battle, and it was often restricted by a numerical "memory" value that made it hard to make proper use of. In the Starforce games, this is replaced by the Favorite Card system in which you choose a set of four or six cards that become "White Cards", letting them be used more freely, with no over-complicated memory restriction. This change simply lets you use cards you like more easily, and it is a lot easier to manage than the old system, so I prefer it.

There is a lot more I can write about, such as the new Star Card system, but that will have to wait for another day.

Tuesday, August 26, 2008

Final Fantasy IV: Counterattacks

One of the most significant gameplay differences between the original versions of Final Fantasy IV and the new DS remake is the radical transformation in how the game's major bosses fight. Scarmiglione, Cagnazzo, Barbaricia, Golbez, and many less significant bosses have so far all been given brand new counterattacks in order to increase the game's difficulty. Many regular enemies have similarly been given new or enhanced counterattacks, to a degree that I have never seen in another RPG. However, the results of this experimentation are actually really good; the addition of counterattacks as a major part of most battles has radically increased the amount of strategy that I have needed in order to win compared to earlier versions of the game, and has made many fights much more dynamic and interesting than they originally were.

Counterattacks are a pretty simple way of making up the gap in the number of actions a boss has compared to a large party of characters. If a boss is limited to making only one action in the time frame where each party member is capable of taking one action, then it is at a significant disadvantage and will be quickly overwhelmed by even simple strategies. This is why many RPGs, such as Final Fantasy 3 or Dragon Quest 8, have bosses capable of making 2 or 3 actions in a single combat round. However, the Active Time Battle System makes it harder to judge how many actions either side is going to make in a certain time-frame, and making multiple actions in a row would look inconsistent under the ATB system. Many later Final Fantasy games got around this by giving bosses multiple parts, each of which could attack independently. However, it would have been too inconsistent with the original game to completely rebuild boss designs to accommodate this method. So, giving bosses lots of counterattacks was a pretty elegant way to overcome the player's number of actions and make fights harder.

The addition of counterattacks has pushed FF IV's dominant strategies in a completely different direction than the original game and even most Final Fantasy games. In most entries in the series, the dominant strategy is a scorched earth approach, where the player continuously pounds on a boss with physical attacks and the party's best magic attacks. The counterattacks in FF IV DS force the player to adopt a much more conservative and defensive strategy instead. For example, Barbaricia counterattacks with an all-character lightning spell whenever she is hit with an attack. If a player was to recklessly attack her with all of his fighters, he would find himself listening to the Game Over music in short order. Instead, it is necessary to focus on defense and healing, so that the party can safely absorb the counterattacks without dying. Other bosses who counterattack with negative status inducing spells similarly force the player to focus his energy on restoring those debilitating status conditions. The fact that FF IV DS has this different dominant strategy is actually really refreshing. It makes the game feel new and exciting instead of the typical Final Fantasy experience.

A particular benefit of having bosses (and regular enemies) use lots of counterattacks is that it makes power storing abilities useful for the first time. In FF IV, Yang has an ability called Focus where he sacrifices a turn to store up his energy so that his next physical attack will do double damage. While this ability has been in not only many Final Fantasy games but other RPGs as well, it has never been useful. Spending an extra turn to do double damage only does as much total damage as making two regular attacks in a row, and performing two regular attacks always has the added advantage of at least doing some damage on the first turn, in case the enemy only has a few hitpoints left. However, the addition of reliable counterattacks makes a single big attack advantageous over several weaker attacks; fewer attacks means fewer opportunities for counterattack. So much to my surprise, Focus went from being a move I never used to an important element of my strategy.

I also need to say that Barrier Shift, a staple of the Final Fantasy series, works much better as a counterattack move than as an actively used power. When Barrier Shift is actively used, it results in two problems: the boss wastes too much time Barrier Shifting to actually fight, and it is so unpredictable that it is likely going to frustrate the player into just ignoring it and focusing on non-elemental attacks. Making Barrier shift a counterattack solves both of those problems. Since the boss doesn't have to spend a turn Barrier Shifting, it will be focused on doing damage throughout the fight. Furthermore, the fact that the boss will only Barrier Shift as a counterattack is predictable enough for a player to actually go to the effort of scanning the boss's weaknesses and exploiting them.

As a final note, one of the best features of the counterattacks seen in Final Fantasy IV DS is that they are always explicitly marked. Every time an enemy uses a power as a counterattack, it is marked Counter: Attack, Counter: Lightning, and so forth. This way, the player knows for certain that an attack is done in reaction to one of his actions, and can plan accordingly. Without this clarity, it is possible that a player could get the mistaken impression that the boss was just doing all of its actions at random, and end up getting frustrated from constantly falling into the enemy's trap. A boss who counters every physical attack is hard enough on its own without forcing the player to figure it's pattern out blindly.

All told, the boss fights I have seen so far in Final Fantasy IV have been really fresh and exciting thanks to the newly added counterattacks. My only complaint is that it might be a little too consistent, since it is fun to beat a boss down with lots of basic attacks now and then after all. I am really curious what the next several bosses in the game will fight like.

Final Fantasy III DS: The Job System

After a few failed attempts at defeating the final boss, I finally managed to complete Final Fantasy III yesterday. Other than a frustrating lack of save points and a few truly annoying enemies (things as powerful as the Red Dragons should not just sneak up on you at random), the final part of that game was quite a bit of fun and I enjoyed the game as a whole. Now that I am done with the game, I want to write a bit about the game's defining element: the Job System.

The Job System is one of the greatest systems to ever emerge from the Final Fantasy series. It allows characters to change their abilities around in a highly customizable fashion, so even a small, unchanging central cast of characters can be as versatile as an army of characters, which helps greatly in enabling variety in the gameplay while keeping the main cast steady (which can really help the plot). And, unlike many of Square's later game systems that also aim for that same goal (like the FF7 Materia system or FF8 Junction system), it presents character options in manageable chunks, so the player doesn't have to get bogged down with micromanaging a large number of small, independent elements. To put it plainly, it is much easier and more meaningful to make the choice between making a character a Knight or a Black Mage than it is to choose between assigning Firaga spells to a character's Strength or assigning them to his Hit Points. When a Job System is implemented well, it can provide a very good gameplay experience.

In the particular case of Final Fantasy III, though, I am not sure if the Job System really is implemented well or not. Certainly, later games such as Final Fantasy V and Final Fantasy Tactics made radical improvements to the framework established in Final Fantasy III, and as a result the earlier game suffers from the comparison. At the very least, there were many times throughout my playthrough of Final Fantasy III that I wished I could change the equipped abilities of a Job like I could in Final Fantasy V (it would certainly have made some endgame choices a bit easier). At the same time, Final Fantasy III does have a number of unique elements that could not be replicated in later versions of the Job System that work well in this game, such as the distinct tradeoff between lasting power and flexibility seen in the Magus/Devout vs. Sage comparison. Since I can't really say whether the system as a whole works well or not, I guess all I can do is ramble on about some of the smaller details...

1) I don't like the transition period characters undergo after changing Jobs, in which they have reduced stats for a set number of battles while they "learn" the new Job. Because it doesn't really accomplish anything that isn't already controlled by other factors (switching in and out of mage Jobs is limited by the way it affects current MP, etc), and it can be completely bypassed through careful planning and high Job Levels, all it does is get in the player's way and impose some arbitrary connections between certain Jobs.

2) While I prefer some of the ways this is implemented in later games, I do like the simple fact that you get better with a Job the more you use it (the higher your Job Level rises). It creates a reason for a character to stick to a Job for a long period of time, so Jobs become long-term investments rather than a temporary state designed to defeat a single boss with a particular strategy. Without this, there would be no reason for a player to always turn to the same character whenever he needs a White Mage. Of course, it would have been better if the mechanics of raising Job Level and benefits for doing so were laid out more explicitly.

3) I don't like the fact that spending a large amount of time raising Job Levels in one Job becomes meaningless once that character switches to a different class, particularly in the case of "upgrade" Job pairs like the White Mage/Devout pair or Monk/Black Belt pair. When the Black Belt Job become available, the Monk Job becomes obsolete, so any levels dedicated to that class (and thus all of the time the player spent raising those levels) completely go to waste. There should be some sort of global benefit linked to raising a Job's Level high, even if the Job itself doesn't have any more use.

4) I like the way that you never really need any one Job for any given situation, so you always have choices. Every Job performs various roles in a unique way, so there are an incredible variety of viable teams. Certainly, some Jobs are just better than others (I can't imagine a use for the Scholar, and the Red Mage might outclass every other starting Job), so there are balance issues, but this is a game where I was able to effectively use both a Ninja/Summoner/Sage/Black Belt team and a Knight/Devout/Bard/Black Belt team in various attempts at beating the final boss (the former was close but the latter succeeded), so the game clearly allows for a variety of styles and strategies, an essential quality for a fun game.

5) I think you just get Jobs too late into the game. You are stuck with a very limited set of Jobs for the first fourth of the game, and by the time you get the final set of Jobs you are already at the entrance to the final dungeon. I think it would have just been more fun to be able to play around with some of these classes for longer periods of time. After all, I did like the fact that, in Final Fantasy V, you get all but one Job by the end of the first third of the game. Even better is the Final Fantasy Tactics approach, where the player can essentially determine for himself how quickly he unlocks most of the jobs.

6) I like the amount of detail the designers put into making each character look unique for each Job. This works better with some Jobs than others (earlier Jobs like the White Mage tend to have greater variation for each character than later Jobs like the Dark Knight), but it is a nice touch.

I really don't think I have much left to say. Overall, once you begin to understand the system the game holds up surprisingly well for an NES-era RPG, and I am glad I played through the game.

Thursday, August 21, 2008

Items you don't know what to do with

When I was first going through the first big dungeon of Final Fantasy IV DS, the Damcyan waterway, an enemy happened to drop a rare item called a Rainbow Pudding, which I realized I could sell for a ridiculous amount of money. However, I held off on selling it at first because I thought it might be needed for a side-quest later in the game. Hours later into the game, I finally ended up selling it to finance refitting my party members with fancy new Mithral weapons and armor. In a cruel twist of fate, two dungeons later, I find an NPC asking for Rainbow Pudding, at which point I nearly threw my DS out the window. I only became more frustrated when I learned that the item in question has only a 0.4% drop rate. So I now find myself condemned to spend hours farming an old enemy to get my hands on a rare item that I already had.

An incident like this should not happen. All it does is frustrate the heck out of players for making the "wrong" choice when they do not have enough information to base their decisions on. Unfortunately, it is not that uncommon for this uncertainty and confusion to occur. Recently in FF III, my brother found a stash of eleven Golden Swords, which were useless as equipment, but could be sold for semi-decent money, but he held off on selling them for a while because he had no idea if they were needed for their own quest or not. I experienced this same phenomena many times in FF XII as well, where it is unclear what I was expected to do with certain quest rewards, particularly since selling certain items in that game can unlock expanded store selection.

The problem comes from a conflict between two common game devices in RPGs. The first part of the problem is that many RPGs reward players for finding and holding onto rare items. A great example is Xenogears, where the player can redeem the Mermaid Tear, a weird item found in the first hour of gameplay, for a reward at the very end of the game. A similar phenomenon can be seen in Dragon Quest VIII, where all kinds of esoteric items are needed for item synthesis. This kind of thing is so common, that I now almost habitually hold onto items and equipment that I can't buy in shops, even if I don't have a need for it anymore, just in case.

The other factor at work is that most RPGs expect the player to sell stuff to make money. Final Fantasy XII is a major example, since the only way to make money in that game is to sell the stuff that enemies drop. Many RPGs also give the player items whose sole reason for existing is to be sold for cash (these are usually gold nuggets or jewels, but not always).

So, when a game gives a player an item that has a monetary value, is not sold in stores, and cannot be used, both of these forces come into effect. On one hand, experience tells the player that items like this are often needed in quests, but on the other hand, items like this also often exist solely to be sold. Without any other information to go on, a player has no choice but to guess whether holding on to the item or selling it is the more advantageous course of action. And with 50/50 odds, many players are going to end up making the wrong choice and later become frustrated when they discover the truth. The only solution is to give the player more information from the get-go on what the player is expected to do with an item. Making items that are necessary for side-quests unsellable completely eliminates this confusion, or adding phrases such as "quest item" or "can be sold for money" into the item description and staying consistent with their use.

Final Fantasy III DS: Equipment Woes

Because my brother has been playing Final Fantasy IV DS, I have been inspired to go back and finally play through Final Fantasy III DS, which I have owned for quite a while but never played too much. It has been a very interesting experience. Even with the new 3D graphics and modified main cast of characters, this game still has all of the elements of an old-fashioned RPG made for the NES. Some of these are a nice change of pace, such as the loose directions that tend to encourage exploration, but there are just as many elements that remind me of exactly how far videogames have progressed since those days.

At one point around halfway through the game, the plot dumped me in a new place with access only to a single town right after I acquired a new set of Jobs to play around with. The jobs I just acquired were the Evoker, the Dark Knight, the Dragoon, the Bard, and the Viking. The town, though, only had armor and weapons for Thieves, Rangers, Monks, and Vikings, as well as a spear for the Dragoon and a harp for the Bard. As such, the town provided enough equipment for three of these new jobs to function (because of equipment available earlier, everything but the Evoker and the Bard works, but I will get to those two later), but there was a problem: the enemies around that area and in the next few dungeons were much deadlier than previous enemies, so much so that anyone without improved armor would quickly get torn apart. As such, only Thieves, Rangers, Monks, and Vikings were really viable until I could reach some new towns. I could have bought a spear or two and had a Dragoon who could fight, but he would die so quickly to enemy attacks that it would be pointless.

The problems for Bards and Evokers were much more severe. The Evoker class needed an entirely new set of spells in order to be able to do anything, but these spells can only be acquired a few dungeons after you acquire the Evoker class. Similarly, the Bard requires harps in order to function, since its "Sing" ability depends on the equipped harp, but the only harp available in that one town merely provides a protective effect, which isn't something that can carry a character through several dungeons. In addition, neither job can benefit from the armor available in that town, and are stuck with a set of armor that was outdated long before the sudden increase in enemy power. As such, it would be almost suicidal to try to bring either of these classes into a dungeon when they first become available.

This lack of viable equipment actually becomes a recurring problem for several classes. Warriors, Red Mages, White Mages, Black Mages, Knights, and Dark Knights all have to go quite a long time from the point enemies get a sudden large increase in power to the point that they get a full set of armor capable of protecting against attacks. One glaring example is the Dark Knight job, which, from the point I acquired it (9 or 10 hours into the game), had to make do with armor I acquired a long time prior (6 hours into the game) all the way until I reached a new town recently (17 hours into the game and several dungeons and bosses later), and even then that armor upgrade was incomplete. In the meantime, I acquired several full armor upgrades for the Dragoon, something like five new kinds of lance, five or so Viking weapons, two full upgrades for the pathetic Bell and Tome weapons, and three whole levels of new magic spells.

The real problem with this lack of equipment is that it severely delays the player's ability to have fun with new Jobs. The various Jobs are the most exciting new toys to be had in the game, but because of poor equipment and spell availability the player isn't really free to have fun with them. It is something much akin to giving the player a new kind of gun in a First-Person Shooting game and not giving them any ammo for it until five more hours into the game. It just destroys the fun of experimenting with new things and enjoying hard-earned rewards.

Overall, the more widely-usable equipment should have been available across the entire post-Water Crystal section of Final Fantasy 3. Instead of providing a lot of class-specific equipment that favored eccentric and unusual classes, the game designers should have included a lot more armor and equipment that could be used by a large number of classes, as well as made the equipment necessary for certain classes (Evoker spells, Bard harps, and Dark Knight dark blades) available earlier.

One thing that I will say the game does well, though, is that it tends to use equipment availability as a way of hinting at optimal class choices very well. The game gives you Scholar gear around the time you need a Scholar. It gives a sudden increase in Thief gear right before the point where having a Thief is handy. It makes good Dragoon gear available right before a period in which Dragoons are extremely powerful. Dark Knight gear finally appears right in the midst of the time where it is most needed. This kind of hinting is quite effective, though I wish they didn't sacrifice overall flexibility and class viability for the sake of it.

Tuesday, August 19, 2008

Final Fantasy IV DS: Party Based on Plot

Final Fantasy IV used a very unique approach when it came to creating a party of characters for the player to control: party members moved in and out of the party based solely on the demands of the story, rather than through direct player control. While this approach takes some control over the game away from the player, I think it can allow for a game with a much more personal story than is seen in many RPGs.

There are two approaches to RPG party construction that tend to be much more common than FF IV's approach. In the first, as seen in games like Dragon Quest VIII or Wild ARMS 3, the party consists of a handful of characters who join early and are never swapped out. In the second major method, as seen in every Final Fantasy from VI on, the player has a large pool of characters, of which only a fraction can be in the party at a time, but can be freely swapped in and out. The second approach in particular is almost a standard feature of RPGs. However, these approaches force the game developers to write the plot of a game in a particular way, limiting the kind of plots that are possible.

The second approach's main problem can be very visible: the game developers can rarely customize the game's story and events for specific party compositions. No matter which characters are involved in an event, the event will have to resolve the same way. Let's look at Chrono Trigger as an example. Chrono Trigger gave each character in the team custom dialogue for every scene. However, the characters in the game were only playing out generic roles in each scene. For example, after the Ocean Palace, one of the characters in the party takes over a leadership position and asks the local elder some questions. No matter which party member is placed in that role, the same general information is communicated to the player, particularly since the NPC dialogue is inflexible. So even in Chrono Trigger, one of the best games at handling a swappable party in history, can at best offer only minor variations in dialogue and minor additions to scenes based on party composition. In games with larger casts, such as Final Fantasy VI, developers often have to resort to using generic dialogue. In many cases, developers do not even give dialogue to characters who can be swapped out. The result of this overall approach is that it mandates that the story has to be more about the team as a whole than the individual party members.

The set-up where there is a single unchanging party can avoid the problem of generic dialogue and uncustomized scenes, but does require certain plot considerations. Most notably, the entire party typically has to be introduced within the first several areas of the game. Furthermore, once the party is together, it can no longer be separated for any long period of time. A good example is Wild ARMs 4: the first three party members all join in the opening segment of the game, while the fourth joins soon afterwards. Once the four characters are together, the party is only broken up for a couple short scenarios before it reunifies. This actually produces similar results to the previous approach: the stories of the individual members who make up the group are subsumed into the story of the team as a whole, which can potentially limit the growth of individual characters. The game need to put the party together could also result in rather forced introductions in games (in other words: "Why are all of the characters from the same place when the world is so huge?").

The strength of the approach where party members join and leave freely based on plot is that it makes very few demands on writers. The only demand that it does make is that party members join and leave the party under plausible conditions. However, the act of a party member joining or leaving can create a lot of drama in a game. Tellah's confrontation with Edward in Damcyan castle, Leviathan's attack, Palom and Porom's sacrifice, Rydia's dramatic return in the underworld, and so on are all plot points created to justify the appearance or disappearance of a character, but they include most of Final Fantasy IV's most memorable and dramatic scenes. This also puts the focus of the game's plot on the individual stories of the characters. Tellah is a good example: he joins Cecil at first when their paths go the same way, but leaves after his daughter's death to pursue revenge on his own. When Tellah returns to the party, he joins only to get a chance at taking his revenge, which he ends up sacrificing his life for. The entire time Tellah is a part of the story, his tale of loss and revenge is an ever-present undercurrent in the game that is never completely subsumed by Cecil's journey of redemption. The plot-based approach opens up the opportunity for specific characters to introduce their own sub-plots, develop them, and then bow out of the main story when their own plots are resolved, without ever having to pace out their stories based on the flow of the game as a whole.

The biggest weakness of Final Fantasy IV's story-based approach to party membership is that it is not well suited to game mechanics that involve long-term character customization. It would be really disappointing for a player if he sinks hours into carefully building up a character only to watch that character leave after the next plot event. However, the story approach does work well with game systems like FF VII's Materia system or FF VIII's Junction system, which allow the player to customize characters by equipping them with abilities kept in a party-wide pool. I am actually rather surprised that I have never seen a game that combined such a system with plot-based party membership.

I would like to see more games that used Final Fantasy IV's way of doing things. It is a system that has never seen as much use over the years as it deserves.