Atelier Marie: The Alchemist of Salburg

As the origin point for Gust's Atelier series, Atelier Marie (Playstation/Saturn, 1997) is a pleasantly simple expression of the alchemy-based game mechanics that has grown and evolved across its more complicated sequels. This game stands apart from its contemporaries for the way it aims to be small and economical against the JRPG trend towards being ever more epic and extravagant. Nobody is going to save the world in flashy 3D, no ancient prophecies are about to be unearthed via lengthy FMVs; there's just a young woman who runs a small business while trying to graduate from alchemy school.

Really, the plot actually is that compact and ambitiously unambitious. Within the first 45 seconds of starting the game, it's explained that Marie's grades are the worst in the history of the academy. Her only shot at graduating is to spend the next five years in a borrowed a workshop (an atelier, as it were) where she can work to hone her alchemy skills. If she is able to  use alchemy to create an item that will impress her instructor within that time, she will earn her graduation. Shortly after this exposition, control is given to the player to figure out how to go about making an impressive item in that time.

First day on the job.  Uh...now what?

Although the game remains refreshingly hands-off in terms of telling the player what exactly to do next and how, there are enough people to talk to and notes to read to keep things comprehensible from the start. After learning how just a few things work, the player ought to be comfortably caught up Marie's routine, which essentially breaks down into these interconnected tasks: gather materials, make items, and sell items.

To gather materials, Marie has to navigate a few menus to leave the castle town of Salburg and enter one of the various gathering spots nearby. Different places have different items, take different amounts of time to reach, and have different monsters lurking around them. Because Marie is not much of a fighter, she can hire adventurers to travel with her to help fight off these monsters, for a fee, of course.

Once these items have been gathered, Marie can return to her workshop to combine these raw materials into new items. She begins the game already knowing a few recipes, but new ones are found in books that can be purchased or found, or through sudden inspiration (i.e., a character asks for an odd item, and she discovers she knew how to make it all along). Most of the items she makes will be sold to people who have specifically requested them, either by asking in person at Atelier Marie, or via the owner of the tavern. The rest can be used in one way or another, healing or attacking in battle, as materials for other items, or even sold at the academy shop from time to time.

Unlike in later games, there are no differences in item quality or characteristics that can change from either the raw materials or the items they produce. The only complication here is that Marie can get weary from working too hard, which makes it increasingly likely that she will botch an item. To reduce her fatigue, it's possible to rest, leave the workshop, or -- bizarrely -- drink a bunch of wine.

Mixing some items to make some new items! Alchemy is actually pretty straightforward

At first, Marie has to do all of the gathering and synthesizing herself, which each activity bringing her deadline a little bit closer (and contributing what what we're left to assume is a growing alcohol problem). The smallest unit of in-game time is one day, one or more of which are consumed while making an item or going out to gather materials. Months are all thirty days each, with twelve months making a year. So, there are exactly 360 days in a year, giving Marie 1800 days to reach her goal.

Not long into the game, Marie has the option of hiring elves to do the work of gathering and synthesizing for her. In exchange, she pays each one a monthly salary, agreed at the time of hiring, and dependent on how fast that particular elf can work. Having these additional workers can greatly increase the efficiency of Marie's workshop, and is where the game truly starts to become an interesting simulation of what it's like to run a business.

Doing a deep dive on work flow, making sure we're running key deliverables efficiently

Each elf can be given one task, which they will repeat endlessly, or until they run out of materials in the case that they are synthesizing items. A smart manager will set a team of elves to work on complimentary tasks, having some out gathering the materials that others will use to make certain items that are then needed to make even more complicated items. This frees up Marie to be doing whatever it is she needs to do, while elves crank out explosive devices to make adventuring easier, expensive items to keep money coming in, and so on.

As for what else there is to do, there actually isn't that much outside of gathering/synthesizing/selling. There are events on certain days of the year where Marie can accompany a friend to a bazaar, or one of the festivals held throughout the year. These take up a little bit of time, and can sometimes open up new items or recipes, or even lead to other new events. I hesitate to call them "story events", because I consider Marie's work to be the actual story of the game, with these events acting as peripheral character moments that keep the player from feeling completely isolated in a world mostly devoid of life.

Turning in some requested items for big silver.

In that way, I appreciated these diversions as momentary breaks from work. The current Atelier series (the Dusk trilogy, consisting of Ateliers Ayesha, Escha & Logy, and Shallie) sometimes feels like the side features are threatening to take over entirely. In many ways, the Atelier series has crept back into JRPG territory, with an increased emphasis on conversations with party members, bigger and more complex battle systems, redirecting the fate of the world, etc. At the same time these sequels have made item synthesis more complex and interesting, they have also made it a less significant part of the game in order to make room for everything else.

I think that's a large part of why I enjoyed Atelier Marie so much: it's just a nice, uncomplicated place to go back to; where a person can put in an honest day's work without much interruption.

Where is it Now? : Valkyria Chronicles

There are a number of RPG series that have gone missing over the years, the victims of studio closures, disappointing sales, changing tastes, and so on. Every now and again I'll shine a light on some of these dead or dormant series by way of summary of the reasons why the games were ever interesting in the first place, followed by as much detail as I can manage regarding where the series left off and what has become its legacy.

My first pick is Valkyria Chronicles, a relatively recent example of a series being built and abandoned in the span of only a few years.  It debuted in all territories during 2008, to be last seen in North America during late summer 2010, and late 2011 in Japan.

Why it's Interesting

Even before knowing anything of substance about Valkyria Chronicles, it's easy to be drawn in simply by how good it looks. This is the result of the CANVAS Engine, which creates an unusually attractive water-colory-cel-shaded, paper-textured, edges-sketched-in style. It did an excellent job of imitating the style of 2D animation with 3D models, especially for the time when it was released for the PS3. The anime-styled JRPG as a genre was off to a slow start during that generation, with many of the higher profile titles going instead to handheld systems, or holding out on the previous hardware as long as possible. To see a game in this style looking as good as Valkyria Chronicles on a then-modern system definitely grabbed attention.

Pleasantly, the actual game is as good as its visuals. It is essentially an SRPG in that the player and AI opponent take turns moving units around a battlefield, engaging with enemy units and leveling up while doing so. What makes it unique to the genre is in the way it integrates the idea of tactical planning with third-person action. This is expressed by alternating views of the battlefield from the detached perspective of a tactical map, and the immediacy of directly controlling individual ground troops at their level.

The tactical map view

The 3rd person action view

As a result of this system, the player is both general and infantry, with access to different sorts of incomplete information while inhabiting each role.  The tactical map shows where all known enemy and friendly units are positioned, along with the location of buildings, cover, and other structures, but doesn't always communicate topography clearly or reveal exactly where unseen enemies are lurking. To get this detail, it's necessary to drop down to the viewpoint of an allied unit. From there, the immediate surroundings are perfectly clear, though the detailed tactical map is replaced with a mini-map that is so small and simplified that it really only functions as a reminder of what the big map looked like.

Limiting access to information in this way introduces potentially frustrating gaps in intelligence leading to foolish decisions made by the player who can only see from one perspective a time. Though this may seem like poor design, I believe it was a conscious choice, made with the intention of simulating in some way the differences in viewpoint between the people giving orders and the people carrying them out. If the systems seems limited and incomplete, I would argue that it's because it is based on a real system that can be limited and incomplete, given that the theme of this game is -- in it's most understated form -- "war is problematic". Unfortunately, I can't find a place where narrative addresses the specific point regarding perspective, so I have to admit that I'm merely theorizing that this is what they were going for.

Getting back to more concrete information, Valkyria Chronicles further distinguishes itself from its peers is with the implementation of crossfire. Whenever a unit is in motion, it can be fired upon by certain opposing unit types who are within range.  This makes it possible to set up traps and choke-points to  halt enemy progress, and make it much easier to get enemy kills even when it is the AI's turn. It's also one of the only ways that the 3rd-person control scheme could ever work without looking silly. In a game like Fire Emblem, where all movement is performed by moving units around a map, the fact that a swordsman can walk right past two other melee fighters to take down an archer is fine because it's never seen from those units' perspective. I can only imagine that a similar scene would look bizarrely and recklessly un-warlike if played out in Valkyria Chronicles' third-person view.

Where is it now?

There have been two sequels, both for the PSP -- a move that has meant dropping the CANVAS engine and other aspects of the presentation and scale of the game.  Even so, the core game has remained roughly the same as far as alternating between map and movement, with the addition of new character classes and other refinements and complications to the relatively simple mechanics of the original.

Thematically, however, Valkyria Chronicles II takes a hard turn from the original. Set several years after the events of the first game, it focuses on a group of military academy students who are unexpectedly called on to fight in a civil war. As might be expected given the setting, the characters are pulled from the shallow pool of high school anime tropes, tingeing the plot with a lighter, sillier tone.

The series returns to form with Valkyria Chronicles 3, this time following a secret unit of nameless soldiers who take on the most dangerous missions, in a timeline that runs parallel to that of the first game. Unfortunately, this one never made it to America, though I understand there is a fan translation patch for those who have access to the Japanese game.

Aside from these direct sequels, Sega has also used a Valkyria-like battle system for the otherwise not-particularly-remarkable RPGs, Shining Blade, and Shining Ark, also for the PSP. Incidentally, there's no doubt a "Where is it Now?"-style article to be written about the Shining series, which continues to exist by hopping genre every so often, leaving fans to wonder if they'll ever get a sequel to a Shining game they might enjoy.

In this same vein, Valkyria Chronicles does sort of live on in the form of an online customizable card game called Valkyria Chronicles Duel. I haven't played it at all, but nevertheless condemn it in my ignorance.

機動装甲ダイオン aka Kidou Soukou Daion aka Dion aka Imperium

Mobile Armor Dion -- most emphatically NOT a Gundam -- launches into battle!

Released at the end of 1992 for the SFC and SNES, Mobile Armor Dion (機動装甲ダイオン, or simply Imperium in North America) is notable for its wide variety of weapons and power-ups, the clever way it makes itself both accessible and impenetrable at the same time, and most interesting to me, the way it incorporates a couple of RPG elements into its already unique vertical-scrolling shoot 'em up (shmup) formula  And of course, I can't ignore the excellent use of pseudo-animation in the game's opening sequence, as seen in the GIF at the top of this article.

Beginning with the weapon load-out, Dion launches with a basic machine gun that is quickly expanded to include an energy blast that tracks enemies, a powerful laser, and a beam weapon that fires in the opposite direction to where Dion is moving (and can be locked in place by holding down the R button). Each has its own strengths and weaknesses that can exploit the strengths and weaknesses of the various enemy types that appear in each level. Learning which weapon to use when is a key part of making progress through the game, though there are usually several ways to approach any given situation.

Enemy-seeking energy blasts end multitudes of fast, weak enemies before they can even get a shot off

Part of the reason for this stems from the way Dion gains -- and loses -- weapon power ups. Rather than powering up by collecting items dropped by enemies, Dion uses what would normally be a score count to function very much like EXP in an RPG. The upper left corner of the screen shows two numbers: EXP and N.EXP, which together indicate how much EXP has been gained, and how much is needed to reach the next "level". Note that these aren't necessarily levels in the typical RPG sense, rather, each weapon has its own level, and they take turns leveling up whenever N.EXP reaches zero. In doing so, a machine gun that originally fired two shots at a time will begin firing four in a wider pattern, the single laser becomes a twin laser, and so on, greatly increasing in power as more and more robot enemies fall to them.

Keeping this power in check is the penalty incurred whenever Dion is hit by enemy fire. Any time this happens, Dion loses one HP (displayed as the five red bars on the top right side of the screen), and the weapon currently equipped goes down one level (to a minimum of Level 1). As a result, there will be times when the best weapon for a given situation is underpowered relative to the others, adding some complication to the moment-to-moment decision of which equipment to use. Even worse, leveled down weapons slow progress towards getting the strongest weapons. It's possible to level up the weapons again, but it becomes much harder to do so while relying weaker weapons to defeat the stronger enemies that appear as the game progresses.

The Level 3 laser edges closer to the border of ostentatious

The dim ray of mercy this game shines on the player comes in the form of the previously mentioned HP gauge. Dion can take five hits in a row before it's game over; a welcome improvement over the much more typical one-shot-and-you're-dead mechanic common to shmups. Making this approach even friendlier is the fact that each time a weapon levels up, one HP is recovered. Clearly, it's best to not take any hits, but I appreciate the way this game gives the player a way to recover, so long as they've got the skills necessary to do so.

For me, this is especially useful in the early stages of the game, where typically frustrating errors are rendered virtually consequence-free. In most shmups, being hit during the first stage probably means a quick restart, given that it it will be hard to make up for whatever power ups were lost at a point where death should be easily avoided. In Dion, getting hit early in the game won't level down any weapons, since they are already at level 1 and can't downgrade further. The only real impact is that one HP is lost, though it can quickly be recovered.  Make no mistake, Dion is still a difficult game that rewards excellent reflexes and pattern memorization; it just doesn't penalize the player for mistakes as severely as others in the genre.

The beautiful city backdrop seen here in the JP version is replaced in the NA release by a digitized trash heap

As of this writing, both the NA and JP versions of the game can be had for a reasonable price from your online seller of choice. I'd consider the JP version to be the definitive one, as some background effects were stripped down and/or outright changed in the NA release to its remarkable detriment. The enemey-seeking weapon was also replaced with wave beam that, while neat-looking, strikes me as a fix to a problem that didn't exist. Aside from the instruction manual, no Japanese language proficiency is needed, so it's a pretty safe and highly recommended import.

Alcahest

Alcahest is a top-down action game developed by HAL Laboratory (Kirby, Mother) and published by Squaresoft (Final Fantasy, Seiken Densetsu) in December 1993. It puts the player in the role of Aren (アレン), a warrior tasked with gathering the four elemental Guardians whose power will allow him to defeat the evil god, Alcahest.

Receiving a sweet blade and shield from the Guardian of Fire

Each of these Guardians is trapped in a different part of the world, locked away behind a boss battle across several discrete stages. Once freed, each Guardian strengthens Aren's sword, both in terms of the damage it deals, and through the addition of a new sub-attack.  By holding down the attack button, Aren charges a sub-attack that, when released, unleashes a flurry of stabs, spins him around in a slicing whirlwind, et cetera.

In addition to these sword skills, Guardians can also be summoned at any time to fire a few energy blasts or otherwise perform a unique special attack. These are typically sufficient to fell several regular enemies or to do significant damage to a boss, so long as Aren has enough MP to use them. Although his MP pool is limited, items that replenish MP are found relatively often throughout each level, so it's not necessary to be overly stingy in using it. It's completely possible to call on the Guardians several times while completing a stage and still have enough left to work on a boss.

Also aiding in Aren's quest are a group of what are essentially temporary party members called Partners. They follow directly behind Aren to attack whenever he does, though typically with a longer reach than Aren's sword. Some Partners fire projectiles that track enemies or bounce off walls, while others use a long-reaching mace or equivalent. In general, they make a positive impact on how quickly and safely Aren is able to dispatch enemies.

Similar to the Guardians, Partners also have their own unique skills, though theirs consume SP rather than MP. Most of these skills are, again, meant to wipe out a number of regular enemies quickly, or to seriously damage a boss. I'm particularly fond of the Princess Elixir, who will automatically heal Aren whenever his HP gets too low, rendering them both practically invincible as long as they're together and have at least one SP remaining.

Princess Elixir offers to help you annihilate the rest of Level 2

Unlike the Guardians, who join Aren permanently and can be selected at any time for summoning, Partners come and go depending on how each level is scripted, and only one will join at a time. This is justified in the narrative as the Partners having their own goals that they pursue in their own ways, which I found to be helpful in fleshing out the world enough to show that Aren isn't the only adventurer in it, nor is he automatically the most important. At the same time, because Partners do actually help Aren directly when their paths align, they avoid the typical conceit in which an NPC nominally vows to help the player, then runs off to do something off-screen that is unlikely to have any practical effect on the gameplay.

While on the subject of gameplay, I think it's worth mentioning that although the game interface clearly shows EXP, VI (vitality), SP, and MP, I would hesitate to call this an action-RPG.  The EXP count really functions more like the score in a shmup in that reaching a certain number of points will grant the player an additional continue, rather than increasing any particular stat. That said, stats do increase, though in pre-determined ways such as finding better equipment or items in specific treasure chests, or through the boost to Aren's sword that comes with each new Guardian. Because of this, and in spite of its strong RPG pedigree, Alcahest feels more like A Link to the Past than a true action-RPG.

Further encouraging this comparison, most stages have a number of hidden items that must be collected to continue on. However, unlike the versatile tools found in Zelda games, Alcahest's items are all fundamentally keys, even if in the form of a gas mask or special boots. Each one has a very specific use that otherwise has no real effect on gameplay, except as a means to open up the next part of each level. Regardless, the real appeal of Alcahest for me is the combat, and not necessarily in the less-than-Zelda-esque adventuring.

This golem's no liar, though his pants are surely on fire

As the game progresses, Aren gains an impressive variety of ways to deal damage.  In the space of a few seconds, he can swing his sword, cuing his current Partner to perform an additional attack.  He can then summon a Guardian to drop in and start doing damage on its own, while beginning to charge a sub-attack.  Then he can cause the Partner to use their special skill while at the same time letting rip a powerful sub-attack. It takes a little while to gather everything needed to do all of these things, but once the game gets going, the skills at the player's disposal continue to grow and change on a regular basis.

If you're interested in checking this game out, there's a Longplay video up on YouTube at the moment, based on what looks like a perfectly serviceable fan translation. There is also a walkthrough on GameFAQs, in case you'd prefer to play the game yourself, but could use some help with the pervasive and often important Japanese text. In either case, I do recommend spending a little time on this relatively unknown example of what HAL Labs and Square can do.

Fairy Fencer F -or- Fasty Faster F(ast)

Galapagos RPG's first game, Fairy Fencer F, will be released in North America next month, making this the perfec--

*skip*

The basic flow of the Fairy Fencer F consists of using a menu to choose people to speak with in your party, then using a similar menu to talk to people in town -- especially the girl who points out new locations -- before going out to the world map to select and play that new level. The levels themselves are a series of indoor or outdoor pathways, leading through green fields, ice caves, warehouses, lava caves, and the like. Enemies are visible while walking through the levels, and battles are initiated whenever the player makes contact with them. At the end of most levels there is a boss that must be defeated before going back to the menu that lets you talk to your party and so on.

I'm close to the end of my fourth(!) playthrough of the Japanese version of Fairy Fencer F, a fact I am confident would not be true if it weren't possible to skip almost everything mentioned in the previous paragraph.

Loosely translated : "Shut up and everything will be fine"

Dialogue with the party? Initiate it and skip it. Talking to people in town? Get them started and then skip it. The enemies in each level? Run past them (i.e., skip it). Boss fights? Still necessary, but it's possible to skip all the attack animations and even the victory screen at the end. By skipping as much as possible, a group of tasks that could take 20-30 minutes to complete will end up taking more like two. By extension, a complete playthrough that might take 40 hours will take more like four.

This will no doubt sound familiar to anyone who has zipped through the Hyperdimension Neptunia games. Both series are fundamentally identical, right down to the button layout and the suspicion that the game exists primarily as a hangout spot for anime tropes. That said, FFF does have a much more traditional JRPG setting and plot compared to the games industry world of Neptunia, so there are at least narrative differences between them.

And yet despite my advocacy for ignoring the story entirely, I didn't skip anything the first time through, and found some parts to be fun or interesting. The main character is so lazy, for example, that he resists being rescued from prison early in the game, preferring to stay in a cell just laying around waiting for his three meals a day. I'll admit to having a similarly short-sighted fantasy during the laziness of my youth, so I found him to be endearing, even if somewhat tiresome.

Just a sampling of the people I will never speak to again

I also enjoyed the idea of collecting Fairies and Furies, basically a system of matching up an accessory to a set of modifiers that apply to an individual character when equipped, or to a level when stuck in the ground. A character might get, say, a bonus to fire damage when equipping one of these, or the entire party would get 30% more experience while doing 10% less damage if applied to a level instead. For me, the fun was spending the time to balance the buffs I wanted for any particular character against the gains/losses I wanted to get out of each stage.

After the first time through, however, I've been skipping through the game as fast as possible, grinding levels, collecting Fairies, and all around turning small numbers into bigger numbers. I may complain sometimes about the way games like this pander to otaku sensibilities when it comes to characters and characterization, but I can't escape the allure of the simple, often mindless otaku task of leveling up over and over again.

This is the reason I keep coming back to FFF: the game has been designed to accommodate excessive leveling without requiring an excessive amount of time to do so. Galapagos RPG knows that attack animations just get in the way of dealing damage as fast as possible. They know there's no need to fight low level enemies, especially on subsequent playthroughs when the party starts at the level they ended the game with. They know that the story stops mattering when shifting into the postgame grind. Monster variety and nifty animations and dialogue all have their place, but so does stripping a game down to a few profitable runs that can be completed as fast as possible.

In that respect, at least, Fairy Fencer F is no doubt one the greatest JRPGs of all time. Certainly worth a look at some point after it comes out in North America on September 16.

Final Fantasy XII - Gambit Love

I hate AI-controlled characters.  Never have I thought, "thanks, AI-controlled character, you really saved us back there" in any genre.  Instead, my thoughts usually have more to do questioning their use of a healing item, choice of target in combat, or worse: the suspicion that the AI doesn't know that "friendly fire" means we'll all die if they stand in the back and spray bullets everywhere.

While some of this nonsense can surely be traced back to poor programming, I suspect that even if the AI were making excellent decisions every single time, it would still come out looking foolish if a) they aren't the same decisions the player would have made or b) the player can't understand the reasoning behind a seemly odd decision.

Final Fantasy XII solves these issues quite handily with the introduction of the Gambit system: essentially a series of IF-THEN statements the player uses to instruct the party AI how to respond to various combat situations.  Though the syntax is a bit different than in these examples, the simpler sorts of decisions Gambits can be used to make are, "if a character is poisoned, cure their poison", or "if there is an enemy in range, attack it", but it goes even deeper than that.

The Gambit menu, mon ami.

Some of the more interesting Gambits are based on the percentages of remaining health for an ally or an enemy.  They make it possible for a character to steal from an enemy only at the beginning of the battle (enemy health = 100%), or to gang up on an enemy that's almost dead (enemy health < 10%, or enemy with lowest health) to get it out of the way rather than let it continue to attack.  On the ally side, these Gambits can be used to heal characters whose HP are below a certain percentage, and can even select which healing spell or item to use based on how low their HP gets.  They can even automate buffs and debuffs -- which I'll admit, I rarely used in JRPGs before FFXII -- and take advantage of enemy weaknesses (if enemy is weak to fire, cast Firaga).

In short, Gambits automate a significant portion of FFXII's combat by selection actions the player's behalf.  In fact, they can be active for every member in the party at the same time, meaning that it's possible for the player to put down the controller and let a battle play out completely on its own.  This leads to the common criticism of Gambits: that they allow the game to play itself.

The problem, as I understand it, is that part of the appeal of JRPGs is in selecting actions for each party member.  In particular, a good boss battles is a chance for the player to work out a strategy for keeping everyone alive while continuing to damage the enemy.  At their best, these battles often require careful consideration of what each character should do next as they slowly approach victory.  When Gambits are enabled (note that they can be disabled), every battle takes off without needing any input from the player.  There's no need for strategy, no weighing the odds of how the next turn will help or hurt, just three AI characters running around beating up on some monster automatically.

I think it's a fairly reasonable argument, and I hope I've represented it accurately, but I nevertheless think it fails to consider the way that the Gambits are put into place and the limitations they have on them.

Other than the first ones meant to introduce the player to the concept, the game doesn't automatically set any Gambits.  It is up to the player to determine what actions they want each player to take in response to various triggers.  I know, as a player of JRPGs, that if an enemy is weak to a particular element, I would do well to attack using that element.  I also know that using the proper element will most likely mean entering a menu for magic, and then scrolling to the appropriate spell each time I want to use it.  What I appreciate about Gambits is that they allow me to tell the game that I would like to use that appropriate spell whenever appropriate, and then it gets used whenever appropriate.  Likewise, if I don't care about using magic, I can just as easily tell the game that I want to just attack, no matter the weakness, and off it goes.

The point is that it is still up to the player to decide how the battles will play out, but it will not necessarily be up to the player to navigate the menus to make it happen each time.  For me, the ideal party would be one that has such carefully tuned Gambits that they never need any input from me once a battle starts.  It would be the perfectly balanced party, prepared for any situation. I can get close to that, but there are still times when player input is useful, due to the limited number of Gambits each character can have, and a limited selection of IF-THEN statements to choose from, both of which do increase as the game progresses.

What this means is that there will often be exceptions to what the Gambits can handle, so the player will need to step in and issue commands.  Doing so pauses the action, providing time to actually go through menus and select the action each character should take next.  This can be repeated as often as necessary, bringing the game to something of a turn-based system, even if only temporarily.  And once the battle ends, the player can then refine the Gambits to effectively improve the AI for the next encounter.

Unfortunately, there are times when there won't be enough room to add another Gambit, or the IF-THEN statement that would fix everything isn't available yet, because the store that sells it is still out of reach.  Having to buy gambits in shops is the part of this whole system that irks me.  I suppose it could be overwhelming to have every Gambit available from the beginning, and there's the danger that a new player might get a list of the best Gambits to use online and never bother learning to use it themselves.  While I agree that experimenting and perfecting a character's Gambits is a significant part of the fun, I do wish they would trust us enough to let us sort out for ourselves how we want to engage with this in all aspects.

That one gripe aside, I do love the Gambit system, and I hope it gets used again in some later Final Fatnasy, or in whatever JRPG happens to copy it.  The pseudo-evolution of this system that shows up in Final Fantasy XIII is still good at streamlining battles, though I miss the more granular tweaking that can be done with Gambits. For anyone who passed on FFXII because of Gambits, -- as I almost did -- I'd encourage you to give it a shot, if for no other reason than this is one of the few games that, if you don't like the AI, you can change it to better suit you.