The Stack

I’ve survived another trip to dungeon level 4. It was a close thing, though: fully half my party was dead by the time I made it out. (Fortunately, they all survived resurrection.) This time, though, it wasn’t due to any individual encounter. It’s because I spent too long wandering the dungeon and getting into fights. Not deliberately, either: I hit a teleporter I didn’t know about, and had to find a way back to the exit through uncharted ground. The upside is that it’s not uncharted any more.

Since the primary distinguishing feature of my last session was more fights than I wanted, let’s talk a bit about what fights are like. The Wizardry combat system is the basis of turn-based combat in so many RPGs, notably including Final Fantasy. It’s all turn-based, or, more precisely, it’s what’s been called a “phased” system: you give every character their instructions for the round — whether to attack or cast a spell or whatever — and then there’s a certain amount of randomness in the ordering of the results, weighted by experience level and Agility. This includes the actions of the monsters.

Monsters come in homogeneous stacks, with up to four stacks per encounter, but act individually. The stacks (or “groups”, as the docs call them) are significant to the mechanics: when you specify a target for a spell or an attack, you specify the group, not the individual. Many of the combat spells affect an entire group at a time, which makes a sufficiently-advanced mage vastly more powerful than an equally-advanced fighter, who can kill at most one creature per round. Past a certain point, the fighter has only two purposes in the party: meat-shield, and conserving the mage’s spells by killing the less-threatening monsters the slow way.

Not that this makes them less crucial as parts of the party! Mages really, really need their meat-shields, because it’s the only kind of shield they can use. Although the interface doesn’t indicate this clearly, the party is divided into two rows. Only the first three slots are in melee range of the monsters. I don’t mean that there’s a reduced chance of hitting the back row, I mean the back row can’t be targeted by physical attacks at all. This is why samurai and lords are so valuable: they’re spellcasters that you can put in the front row without getting them killed. They are, in effect, their own meat-shields. Priests are almost as good in this role, in that they can wear armor that’s almost as good as a fighter’s, and in addition can carry enough healing power to compensate for the difference. But I’m still somehow not comfortable putting more than one priest in front.

Mind you, even in the back row, you’re vulnerable to spells. This is something I remember coming as a bit of a shock when I first encountered spellcasting monsters on dungeon level 2 of Wizardry I. I had gotten used to wiping out the monsters with group damage spells, and suddenly they were easily wiping me out using exactly the same techniques. It seemed somehow unfair, despite being symmetrical. Or, well, not entirely symmetrical: spellcasting monsters tend to come in groups, which increases their chance of getting in the first cast, which can determine the outcome of the entire battle. It’s a good thing that your initiative goes up with your experience level.

Frequently, combat begins with a surprise round, in which only one side gets to attack. The interesting thing about this is that you can’t cast spells in a surprise round, which completely changes the dynamic. The priestly ability to dispel undead isn’t considered a spell, so that’s pretty much your only option for disposing of monsters en masse. Until, that is, you realize that this is what scrolls are for. I had written them off as worthless at first, just an expensive way of getting the equivalent of an extra spell slot, but a Katino (sleep) scroll used in a surprise round against against a group of spellcasters has on more than one occasion prevented them from getting off a single spell.

Well, it happened, just like I said it would: I got cocky. My rapid mastery of dungeon levels 2 and 3 led to a total party kill on level 4. The really galling thing here is that my first foray to that depth was relatively placid: I explored a small area, enough to find a quick one-way route back to town, which seems pretty important, given how long it takes to get to and from that level by the front entrance (and how many encounters you’d have along the way, and what shape your party will be in on the way out). On the second trip, I tried to just do a sweep to that exit, but got slaughtered en route. I’ll want to pick up their corpses at some point, but for the moment, they’re out of reach. The one time I attempted that level with another party, I was beaten back; by the time I completed the return journey the long way, only one member of the party remained alive. Most of the rest resurrected successfully, but it’s clear that I’m not going to be able to make more progress until I’ve improved my defenses to the maximum possible extent.

And that’s been a bit of a problem. The dead party had better armor than you can buy in the store, and better also than what I’ve been able to find since. Even worse, my new priest has been strangely unable to master Maporfic. This is a priest spell — all of the spell in the game have nonsense names — a spell that improves the entire party’s armor class, and lasts for the entire expedition, until you go back to town. (Or until you hit an anti-magic field, I suppose, but I haven’t run into any of those yet.) Since it offers potentially unlimited protection for a single casting cost, Maporfic quickly becomes a part of your regular routine on entering the dungeon. I certainly don’t dare consider braving level 4 again without it. But at this point, it’s starting to look like the quickest way to get it will be to create a new priest.

To fully understand my problem, you have to understand how leveling works, and just how random it is. And since leveling up is the focus of my attention at this point, it seems a good time to describe it in detail. As usual, killing monsters (or even just passively standing and watching the rest of the party kill monsters) yields experience points, which yield character levels at exponentially-increasing intervals. Unusually for a CRPG, you don’t gain experience levels in the dungeon; you have to explicitly stay at the Adventurer’s Inn back in town. Sending each party member to the inn becomes part of the post-adventure routine, just like selling your loot. If you failed to level, you get a report of how many more experience points you need; if you succeed, you get a report of exactly how it changed you. You always gain at least one hit point on leveling, but frequently only one — although sometimes you’ll wind up gaining twenty. (As a point of comparison, all level-1 characters have 8 hit points.) Some of your stats will increase by one, but others will decrease by one, with no discernable pattern, except that higher experience levels seem to make increases more likely. At low levels, net stat loss on leveling is not unusual. (I have one fighter in my party whose IQ stat is down to 0; I think it’s possible for it to even go negative.) I know I keep saying this, but: No one would design a game like this today. The whole idea of rewarding experience with random decreases in power seems downright perverse by modern standards. But really, it isn’t as bad as it seems, because the increase in overall power from just being a higher experience level more than compensates for such losses.

Leveling is also how you learn spells. It’s the only way to learn spells, and it’s just as random. Spells come in seven levels, and there are limits on how soon you can learn them (with pure mages and priests getting access to higher spell levels at lower experience levels than bishops, samurai, or lords), but once you have access to a spell level, you’ll just pick up the spells in it at random as you gain experience levels. But it’s quite possible to start picking up level 5 spells before you’ve got all the level 4’s. Quite possible, and quite frustrating.

The one upside of spending three or four experience levels trying to get a particular spell is that, when you’re done, you have a character who’s three or four experience levels higher than before. Even if I decide to give up and power-level up a new priest, having a high-level healer around will expedite it. The amount of time you can spend in the more experience-rich parts of the dungeon is limited mainly by the amount of healing magic you have available.

The bulk of dungeon level 3 is taken up with a big mass of one-way walls. These are in some ways equivalent to one-way doors, in that they allow you to pass through in only one direction, except that from the passable direction they’re completely invisible. The effect is fearsome. Before this point, you could rely on having an escape route behind you most of the time, and even the introduction of one-way doors meant that getting cut off was the result of a conscious decision to go through an unexplored door. But now, any step can cut you off. Even worse, though, the geography simply doesn’t make intuitive sense any more.

The underlying model must involve separate records for each map tile for what lies in each of the four directions, because there’s clearly nothing enforcing consistency between adjacent tiles. Presumably the renderer has some way to determine which tiles to consult about what to render for each particular wall slot — for example, tiles to the left of the player’s view determine the visibility of their own left walls but not their right walls. This is all more complicated than the contemporary 3D dungeons in the Ultima series, which were simply grids of blocks, each of which could be either solid or empty. In fact, it’s kind of like a primitive version of portal rendering, with each map tile treated as its own sector. (The effect of looking into a teleporter square is particularly suggestive of this: the renderer displays the sectors adjacent to the teleporter’s destination. Which can be really confusing if you don’t know there’s a teleporter there.)

Note that when I say “primitive”, I mean primitive. We’re talking low-res black-and-white line drawings here. Any significant dungeon features other than walls and doors — whether it’s a staircase, a signpost, an altar, or a mysterious cloaked figure beckoning to you — is rendered as a smudge on the floor. Even worse, the player character apparently has tunnel vision: the view is only three tiles wide. I suppose it was optimized for corridors, where all you need to see is the walls to your left and right, and the immediate entrance to any side corridor. But there are a lot of wide-open spaces in this game — or, in the case of dungeon level 3, spaces that look wide-open from one side. You can be facing a distant wall (where “distant” means four tiles away, the longest distance you can see) and see only three little wall segments in the middle of the screen. The kicker is that Wizardry III is a step up from the original Wizardry engine. The original versions of I and II put their line drawings in one small corner of the screen in order to make room for the party stats and other information. Wizardry III (and versions of I and II ported to the Wizardry III engine) renders full-screen line drawings, and overlays information windows on top of it as and when needed.

Still, I can’t help but feel like the graphics here have stood up to time better than the graphics in more advanced games like The Bard’s Tale. As with the pixel art beloved of indie game developers, it’s primitive enough to have a minimalist aesthetic. There is no unnecessary detail, just enough to convincingly put you in a barren and claustrophobic network of corridors. Which means it’s a little embarrassing when the in-game text tells you that you stand before a mighty castle or you’re on the shore of a lake or something. Just let it be what it is, guys.

Dungeon level 2 is thoroughly explored now, or at least the parts that are reachable initially. This level introduces one-way doors, which can really mess up your plans to retreat to the exit at the first sign of trouble. If I’m not mistaken, this is terra incognita for me, an area that I didn’t figure out how to get to the first time I attempted this game. I had skipped to level 3, which you can reach directly from level 1, but the passage to level 2 eluded me, and I erroneously thought it must have something to do with the few bits in level 1 that aren’t directly reachable without a teleport spell (which is one of the last spells you get). There’s some guidance in the game, but the most direct statement of what I was doing wrong and how to fix it flashed by too fast to read, and possibly even to fast to notice. This is one of the few cases where the game makes the mistake of assuming things about the speed of your PC. DOSBox to the rescue!

The key is that the level is alignment-locked. You aren’t allowed in if there’s anyone evil in your party. For all I know, you might not even be allowed in if there’s anyone neutral in your party; I’ve been shying away from neutrality as limiting my characters’ options for advancement. See, there are alignment restrictions on class. Samurai can’t be evil, Thieves can’t be good, Lords can only be good, and Ninjas can only be evil. Priests and Bishops can be good or evil, but not neutral. There are no neutral-only classes.

The chief way that alignment affects gameplay is that characters aren’t allowed to join a party containing anyone of the opposite alignment. (Neutral characters are always welcome.) For the most part, then, “good” and “evil” are just arbitrary designators for two teams, like “red” and “blue”. I can’t speak about the habits of other people, but when I personally played Wizardry I, I tended to maintain two separate party rosters, one for each team, with some neutral crossover characters, who consequently tended to level a lot faster. After all, I wanted to try every class.

There is one sort of moral choice in the game, however: every once in a while, randomly-encountered monsters are “friendly” — they don’t attack you, and you get to choose whether or not to pick a fight with them anyway. (This seems to only happen with monsters that are significantly weaker than you, which makes the designation “friendly” seem like a polite euphemism.) If you fight them, there’s a chance that some good characters in your party will turn evil; if you don’t, there’s a chance that some evil characters will turn good. In this way, it’s possible to wind up with a mixed good/evil party, but only for the duration of the current session: every time you start up the game, you have to form a party from scratch, and you won’t be allowed to choose the same characters as before if some of them are on opposite teams now. Interestingly, neutral characters never change alignment this way. I recall reading somewhere that this mechanic was only introduced in Wizardry III, although I can’t personally confirm this: I can’t run my original Wizardry I disks on my current system (which lacks a 5-1/4 inch floppy drive), and the version included in the Ultimate Wizardry Archives is actually a port of Wizardry I to the Wizardry III engine, with the same alignment mechanics as the latter. If true, it strikes me as a pretty major change to the game mechanics for an otherwise-faithful port. I mean, without the ability to change alignment, you’d never have Ninjas and Lords adventuring together.

Systems of moral choice in games haven’t really come very far since those days. In a lot of cases, the difference between good and evil is simply a matter of which menu-based dialog items you choose. Wizardry III at least grounds its moral system in game mechanics. But the difference between good and evil is still mainly a tactical one. There will come times when friendly monsters get in your way while you’re making a break for the exit, depleted of spells. You really want to just let them go, but you’re trying to train up a Ninja and don’t want to spoil things. What do you do?

By now, I’m no longer just patrolling the corridor immediately around the stairs out. I’ve explored dungeon level 1 rather thoroughly, making a map as I go. Mapmaking is an essential part of Wizardry. And I don’t just mean that in the sense that you need to make a map or you’ll get lost and not be able to find the exit and run out of healing spells and die in the maze. All that is true, of course, but what I really mean is that mapmaking is a vital part of the feel of the game.

For one thing, adding an automapping system would take away a large part of the game’s challenge. The designers used all sorts of tricks to make the dungeons confusing: map tiles that spin you to face a different direction, teleporters that send you to identical-looking areas without you noticing, one-sided walls, and so forth. Yes, even the architecture hates you in this game. There’s a mage spell that tells you your absolute position in the dungeon, but your spellcasting capacity is limited, and the particularly devious areas have anti-magic fields. The first dungeon level of Wizardry III is relatively gentle: all it has is some disorienting corridor layouts. But I recall from my previous explorations, years ago, that there’s one level that’s basically a big grid with a spinner square at every intersection. ¹[Addendum 14 January] My mistake: it turns out that this was in Wizardry IV. Care and meticulousness is required.

But also, there’s simply a kind of joy, largely lost in modern titles, of creating something tangible as a result of playing a game. Mapmaking is lke a little arts-and-crafts project, with an end product that you can tack to the wall, and look at, and remember the effort that went into it. (It’s even better when the game gives you an excuse to use colored pencils, but that would seem a little strange for Wizardry‘s monochrome wireframe dungeons.) While it’s true that text adventures also support mapping by hand, that feels very different to me. My adventure maps, when I bother making them, tend to be rough scribbles, not the neat and attractive grid of a Wizardry. Partly this is because adventure game maps are essentially the same sort of thing as taking notes: they provide situationally useful information, and that’s it. Wizardry maps provide situationally useful information and reveal large-scale regularity, patterns that enhance the experience even when recognizing them isn’t useful to beating the game. I don’t want to overstate the artistry involved here, mind you. One of the levels in Wizardry 1 was based around a set of corridors in the shape of the author’s initials; that’s the kind of thing I’m talking about here. But it is what it is.

After many trips into the dungeon, and many deaths and TPKs, I finally managed to get one of my Bishops up to level 2. This is a big achievement! Having just one level-2 character provides enough leverage to get others over that hurdle, and before long I had a party of about level 4 on average. I’m still not taking great chances, though. It’s all too easy to get cocky. I haven’t even really begun exploring the dungeon yet; at low levels, you really want to bolt for the exit after every fight, to rest and replenish your spells. (You can’t rest in the dungeon itself, although, since you can’t cast spells in town, the dungeon is the best place to heal.)

Adventuring parties consist of up to six characters. You can send fewer if you like — if you intend to drag characters from previous expeditions back to town for resurrection, for example, each requires an empty slot in your party — but keeping a full roster of combatants is the best way to win fights. After I became more powerful, though, I started putting a thief in the party. Level-1 thieves don’t really count as combatants, because they’re so rubbish at combat. Their armor restrictions make them so vulnerable, you pretty much have to keep them in the back row with the mages. But at least the mages are useful back there. The thief can’t do anything from the back row but wait for everyone else to finish the fight. His sole role, at this stage of the game, is to help you get cash faster by removing traps from the chests that monsters seem to carry around a lot. But that’s a pretty important role, because it’s virtually impossible to get enough cash to resurrect anybody without those chests.

In fact, before I brought in the thief, most of my money came from new characters. Every character you import is completely stripped of experience and equipment, but gets to bring a small amount of cash to get started — just enough to buy some basic armor and a weapon. But you don’t really need to buy armor and weapons for every single soul who whirls through the game and into the graveyard. As long as someone in the party makes it back to town alive, you can strip the dead guys of their belongings and hand them over to the new guys. The more characters go through that revolving door, the more unused cash you can siphon off of them. Taking advantage of this mechanic is pretty much necessary. A patient player could even keep churning until they have the best purchaseable equipment in the game without entering the dungeon at all. But I am not that patient, and besides, I kind of look askance at such abuses. But not too far askance, because I think it likely that the designers had this gimmick in mind.

So, let’s talk about character creation a little, because that’s something I’ve been doing a lot of. This is something that doesn’t change at all in the first three Wizardry games, because all character creation for the first three games is done in Wizardry 1, even if you’re creating a character to use solely in one of the sequels.

Like most early CRPGs, Wizardry draws heavily from D&D. You’ve got the same four basic character classes (fighter, mage, priest, thief), and a set of six stats that are basically the same as D&D‘s but with different names — the only real difference is that Charisma has been dropped in favor of Luck. In early editions of D&D, stats ranged from 3 to 18 because they were generated by rolling three six-sided dice. Wizardry uses a point-buy system instead, but it still makes 18 the maximum out of sheer conceptual inertia.

The creation process goes like this: First you choose a name, then you choose a race, then you assign points to stats, then you choose a class, then you decide whether to keep the character or throw it away and start over. Aside from the name, which could easily have been made the last step, the mechanics pretty much dictate that it has to go in this order. You need to choose the class after finalizing the stats because the stats determine what classes are available, and you need to assign the stats after choosing the race because the race determines the base stats. Dwarves, for example, always start with Strength 10, IQ 7, Piety 10, Vitality 10, Agility 6, and Luck 6. Hobbits ¹Yes, it’s “hobbit” here, not “halfling”. This game was made before the Tolkien estate’s lawyers became the all-seeing eye of flame that they are today. have the highest total initial stats, which you might think would make them an attractive choice, but in fact it’s mostly in Luck, the least-attractive stat.

Every character class has a requirement of at least 11 in some stat: fighters need 11 Strength, priests need 11 Piety, etc. In addition to the base classes, there’s an odd assortment of specialized classes: Bishop, Samurai, Lord, and Ninja. (Yes, this means you can have dwarven and gnomish samurai, which is an entertaining thing to contemplate.) Samurai are fighter/mages, Lords are fighter/priests (in other words, paladins), and Ninjas are unarmed combat specialists who randomly do instant kills. (I have a habit of thinking of Ninjas as fighter/thieves, which would make a nice symmetry, but it isn’t really accurate.) Bishops are combination priest/mages, with the additional ability to identify items. They’re generally considered weaksauce, because they don’t gain higher spell levels as fast as the pure priests and mages, but I’m finding them tremendously useful at level 1, where they can cast four spells before running out of slots, compared to the normal caster’s two. Presumably because of this early usefulness, Bishops are the one special class you can reliably make as a new level 1 character, with stat requirements of just 11 IQ and 11 Piety. Lords and Ninjas are basically impossible to generate at level 1, and can only be produced by changing a character’s class after a number of stat increases from gaining experience levels — in particular, Ninjas require a 17 in every stat.

Samurai, now. Samurai can be produced at initial creation, but only occasionally. This is because the number of points you get to assign to stats is randomized. From the base stats for dwarves, elves, and gnomes, it takes 18 points to meet the Samurai requirements; getting this many points to spend is rare, but it does happen every once in a while. Furthermore, the randomization is oddly irregular; it usually stays between 5 and 11, but occasionally leaps to 18 or 19, and I’ve even seen a 26 come up. This is a design decision that it would be strange to see today. Random factors are all very well for transient events like combat, but for a single dice-roll to affect a character’s options from the very beginning like this is to invite the player to reroll over and over until they get what they want. This is the problem with random generation that point-buy is usually supposed to solve. So it’s very strange to see a point-buy system that goes to extra effort to bring the problem back. Maybe the authors felt that rerolling characters was an important part of the RPG experience.

Wizardry is one of the foundational CRPGs. There was a time when CRPGs were commonly described as either Ultima-style or Wizardry-style, the former referring to ones with a tile-based movement on a large map (such as Final Fantasy and The Magic Candle), the latter to ones with first-person navigation through grid-based dungeons (such as The Bard’s Tale and Dungeon Master). I myself played the original Wizardry as a child. I remember it took a very long time to complete, and seemed a monumental achievement. (I still have my official certificate of completion tucked away somewhere. I can’t imagine sending off for something like that today.) Of course, back then, the very basics of the genre were yet unfamiliar. Common practices like creating a balanced party and putting the mages in the back row had to be discovered by trial and error.

I didn’t play the immediate sequels when they were new, however. I wound up skipping ahead to Wizardry IV: The Return of Werdna, which is more of a direct sequel to Wizardry I than Wizardry II or III is, and didn’t get to trying the others until they were anthologized in 1998. I made an attempt to play through the entire series in order, but got stuck in the middle of Wizardry III when I seemed to run out of dungeon to explore. (I later learned what my problem was; I’ll go into detail later.)

Wizardry II, it turned out, is more like an expansion pack to I than like what we’d recognize as a sequel today — a symptom of the infant games industry still struggling to figure things out. III altered the mechanics and UI quite a bit, but still can’t be run as a completely standalone game, as it provides no way to create characters. They have to be created in Wizardry I and imported, a process that apparently involved swapping multiple floppy disks in and out. Even just playing Wizardry I involved a fair amount of swapping on a single-floppy system, what with its separate diskettes for character and scenario data. The anthology edition was thoughtfully altered to use disk images on the hard drive instead, and to carry out the swapping automatically.

Now, I still have some of the characters that I used to play Wizardry III ten years ago. But I don’t intend to use them, except as emergency support. To get the full experience, I’m creating a brand new party, with brand new characters. And I’m getting them killed a lot. This is part of the experience. Level 1 characters stand little chance of surviving their first encounter, and it takes at least three or so encounters before they get to level 2. And getting killed doesn’t mean resetting to a save point or anything nicely forgiving like that. If your entire party gets killed, you roll up a new party. You do have the ability to recover bodies from where they fell — this is what I mean by “emergency support” — and you can can take them back to town to resurrect them for an exorbitant fee, but even the resurrection spell has a significant chance of failing and rendering them lost forever. (And you don’t get a refund when this happens.) Understand that characters are reduced to level 1 when imported; losing a certain number of your characters is simply part of the plan. So, in contrast to most RPGs today, it’s best not to get too attached to them. I gave up long ago on the idea of giving everyone a distinctive and memorable name; I name my guys in batches like “Fitt”, “Mitt”, “Pitt”, etc. (The initial letter indicates the character class, for greater ease in building a party out of the survivors.)

I really haven’t played much yet, though. Most of today’s game time was spent making preparations: digging out the graph paper, printing up a crib sheet with all the spell descriptions on it, deciding where and how to play it — the game uses a CGA graphics mode that my usual gaming rig doesn’t even support, which forced me to use DOSBox, which made me realize that I might as well be playing it on my Macbook. And of course there was the time spent rolling up all the characters, which can take a while if you’re fussy. So there’s a lot of anticipation going into this. It’s a grand thing, and also a memory of a simpler time, with simpler computer systems.