Feeds:
Posts
Comments

Posts Tagged ‘autopilot’

Balance Part 1: Tao of Picasso

Balance Part 2: Asymmetry and Art

…and now for something a little more concrete.  I’m taking a look at the backbone of character progression in a game I’m designing and digging a little into why I’m making my choices and how I’m incorporating “balance” within the game, specifically how it relates to pacing and balancing player abilities against the game’s design.  (I’ll handle other balancing aspects of the systems in another article.)  Let’s call it a practical theoretical exercise, since I’m not sure this game will ever actually be created much less published, but I’m designing it as if it were something I’m happy enough to put out there as a finished game.  One that I’d even buy and play, at that.  (So if you’re sitting on money with nothing good to do with it, by all means, build something good with this system, or make it better.  All I ask is for steampunk in there somewhere.)

At its heart, my game is a hearty blend of Final Fantasy Tactics and Battletech mechanically (tactics on a hex grid), with a slight zombie flavor and a wacky premise with room for such weirdness as this hippopossum I sketched (yay for sketching in ballpoint pen!).

There’s even a bit of Kingdom Hearts 358/2 Days in the mix, so that’s where we’ll start.  In that game, character progression is based on placing panels on a grid (that expands as the character levels up).  These panels come in a variety of sizes and shapes (some like Tetris pieces, for example) as well as the occasional mutually exclusive choices.  It’s the character’s gear, skills, magic loadout and level, as well as FFX’s Sphere Grid or FFXII’s License Grid all rolled into one, with some quirks.  (OK, that’s a bit of an insufficient explanation, but it’ll do for now.)

My game uses something I’m calling the DNA Codex, a riff on those systems with some flavor for spice.  It combines my older idea of autopilot character progression and high flexibility that I’m so fond of, up to and including the potential to “break” the game.

DNA Codex, empty, level 20

In a nutshell, it’s a customizable grid (3 rows of 5 hexes at level cap, 2 rows of 3 at level 1) that allows a bit of “skills-based” wiggle room within a sort of “class-based” shorthand.  I really want to reward exploration and experimentation within the game’s systems, without penalizing those players who just want to get on with things.  Players who really dig into the system should be rewarded with varying levels of brokenness or crazy fun, while those who don’t want to bother with the “skills” can simply play and be assured of adequate tools to meet the challenges.

It could almost be suggested that there’s a balance between broken near-abuse of the systems, which requires one sort of skill (“systems analysis” if we’re being generous, “munchkinizing” if we’re being snarky), and playing with the default tools, which requires a different sort of skill (better strategy and tactics, generally).  There’s actually nothing that says those are truly mutually exclusive as far as players are concerned, as a player can have both skill sets, the balance comes in when considering pacing and player vs. computer play.  Conquering the game’s challenges requires mastering one or the other or some balance of the two, allowing tactical geniuses to cruise through on their strengths, and munchkins to do their thing.  Total mastery of both tactics and systems would be effectively overpowering the game.  At least… on Normal mode.

It might be noted also that there’s a third rail here that gets used in some games, but that I’m undecided on here.  That’s the leveling fudge factor.  Players in many games that incorporate levels as a character progression tool (and I am, though only 20 levels) can often forge their own edge by grinding for a while, making their characters stronger than they otherwise would be at a given point.

On the other hand, some games totally ignore this, say, something like a Fire Emblem.  Those games typically demand players learn the game and play near-perfectly.  There are harsh punishments for failure and no opportunity to grind up experience and gear by killing inconsequential foes for a while to overpower the game’s challenges with brute force.

Deep down, I really like the leveling fudge factor, as it is a way of letting players self-regulate the challenge of a game.  On the other hand, some masochists love games that demand a lot of the player.  I think in the end, I’ll include a leveling fudge factor, but turn it off for higher difficulty levels.  It seems a fair compromise.

Speaking of difficulty levels, though, that’s another factor to consider.  Who is the average user?  What is their anticipated skill level?  Can the game be tuned well enough to be able to challenge a spectrum of players, maybe even reactively to player choice “under the hood” instead of something as heavy handed as difficulty settings?  Are the Easy, Normal and Hard modes appropriately spaced on the bell curve of player ability?  (Say, almost anyone can get through on Easy, Normal asks a bit more, and Hard demands a high aptitude… and Hardcore will punch you in the head and take your lunch money, then mock you for not asking to be hit again.)

This isn’t so much balancing game design elements against each other to make the game internally consistent and interesting, it’s balancing the game against the potential players.  Perhaps it could be called a sort of “metabalance”, but considering that games tend to need players at some point, especially if there’s even a vague hope of commercial viability, it’s still something useful to consider.

There needs to be a default setting, where the bulk of your target audience can handle the game and derive fun from the experience.  In this case, as it’s a tactical game, I believe that I’m generally looking at a smaller audience, but one more acclimated to a harsher climate.  Kinder, gentler difficulty settings might open the game to more players, but then, how many casual players pick up tactical games in the first place?  I’m not convinced it’s a large number of people, which is sad, but a consideration for the market.

Within the game itself, this “default” is the automated character progression I’ve written of.  Players will automatically have their units on autopilot when it comes to developing the DNA Codex (though they will be notified the game is doing this, and they can override it at any time).  A strength-based melee unit will automatically slot melee skills focused on strength into their Codex.  An agility-based ranged unit will pick appropriate skills, all without the player ever needing to dig into the Codex.  Those skills will provide the player with adequate tools for the challenges at hand at relevant levels.

On the other hand, someone with different tastes might dig into the Codex and start tuning their units to their particular playstyle and tactics.  That’s where the RNA Codes come in.  Here are a few for a strength-based melee unit: two special attack Codes, a five-cell Joust and a three-cell Grand Swing, one two-cell defensive Code, Counter, one three-cell healing code, Regenerate, and two single-cell utility Codes, Shove and Flare.

RNA Codes, strength based melee, level 20

These are the components that a character is built on.  These typically offer specific tactical abilities or passive buffs for a unit, though there will be a variety of RNA Codes.  Because these are shaped specifically to fit into the DNA Codex in certain ways (there will be no rotation), there will be choices to make regarding which Codes are used in which combinations.  A 5-unit long powerful Code won’t cooperate with a 3-unit tall utility Code, for instance, as they would overlap in the grid, and that’s not allowed.  Players will choose which puzzle pieces are important to their particular game aims, and tailor their units to their preferences.

The hope is that this will allow for a variety of playstyles.  Those who prefer higher risk might load up their units with potent offensive Codes, but skimp on defense.  Players who prefer long, slow grinds through a level might prioritize defensive Codes and naturally wind up with a mediocre offense.  There should be a spectrum of possibilities between those, even including some tricky utility options for gimmicky builds.  (I’m reminded of Stasis Bubbles or Assault Shuttles in Master of Orion 2; quirky strategies in spaceship combat that can be employed to devastating effect with a little care, but it’s also possible to just outfit ships with big guns and go start a scuffle.)

It’s also probably worth noting that these RNA Codes are layered on top of a unit’s baseline abilities.  In fact, you could even take all RNA Codes out of your unit’s Codex and play through the game, it would just be harder.  Not impossible, just more difficult.  (See also “No Sphere Grid” runs in Final Fantasy X.)  While I actually like the ability to make “gimped” or dysfunctional builds, I’m going to assure that a RNA-free unit will be usable.  RNA Codes will be bonuses, not baselines.

With those in mind, what all this leads up to, the overall balance and pacing will be established around the autopilot DNA development.  Players who ignore that aspect of the game can play the purist game “as intended” without a fuss (though I’ll certainly include hints that the DNA system could be useful; there’s no reason to make the system totally irrelevant).  Players who want to flush their RNA will be able to make life harder on themselves.  Players who try to abuse the DNA system can make the game easier.  It’s on their head if they want to step off the dock and swim in weird waters, and the trick will be in making the default experience enjoyable as well as making the customization choices viable and fun.  (Notably, this means being able to change one’s mind.  None of these choices will be irrevocable.)

Anyway, this article is already more massive than any of mine in recent history, so I’ll dig more into some more specifics of the design in another article later.  I’d like to write a bit about some “Rock-Paper-Scissors” triangular balance elements that I’m using, and step back and show a bit more of the DNA Codex and some further implications of the design.

If you’ve gotten this far, apologies for my long-windedness.  There’s a lot to consider in game design, and it just all doesn’t fit neatly into 1000 words, even if there were One True Way to design games… which I don’t believe there is.

Read Full Post »

Following up on a comment from Spinks over in the Dual Wield Healing comments, I’ve wondered for a while why “players LOVE classes”.  I suspect there are a handful of reasons, and I’d love to hear what some of you think.  I’m not really disputing that assertion, since I’ve seen plenty of evidence thereof, but I am always questioning why that might be, and if there’s an alternate way (or three) to scratch the underlying psychological itches.  While thinking a bit about those itches, I’ve been thinking of other ways to approach the scratching.

One game that I’ve looked to for good ideas is Final Fantasy Tactics.  FFT has character “Jobs” that function much like classes:  The characters have a core Job that defines their gear permissions (weapons and armor, anyway) and their primary combat abilities.  Soldiers are melee fighters, Black Mages are ranged magic cannons, etc.  Characters can learn abilities from their active “main” Job, eventually Mastering the Job.  They can also use skills they have learned from other Jobs to customize their approach.

Overall, I like FFT’s system, as it allows you to build up a character with a wide variety of abilities that cross-pollinate and synergize, but filters them through the ability to only use a handful at a time.  It’s a nice compromise between learning everything and making tactically relevant limited choices.  Players can make characters specialists or generalists, and anything in between.  This works largely because you tend to field a handful of units in any given skirmish, rather than just a single character.  You can build a team that works well as a whole, rather than just try to do everything yourself.

Battletech works in a similar fashion.  There are several different ‘Mech chassis designs, and several weapons to put in those ‘Mechs.  Players are encouraged to customize their machines by swapping weapons, armor, heat sinks and such, trying to optimize their machine (or team of ‘Mechs in some iterations of the IP) for how they play.  Certainly, there are “stock” configurations of the machines, but half of the fun of the Battletech universe is tinkering with the delicate balance of heat, ballistics, energy weapons, range, mobility, size, and half a dozen other aspects, trying to build the most powerful ‘Mech for its weight.  The stock designs are not usually optimized for greatest potential, which I suspect was intentionally done to give an impetus to tinker, and a reward for those who master the tuning system.

The rough analogue to MMO class design is the Battletech ‘Mech chassis, and the “spec” for a class (minor tweaks to how the class plays) are the loadout of the ‘Mech.  Of course, a MechWarrior need not be tied to a single Mech for his career, which is where the Battletech variability wins out over a class design; it’s like the ability to change your class (chassis) at a whim (or limited by experience/story permissions/bankroll, whatever), allowing for a much greater gameplay variety over the course of a single character’s “life”.  This is also where FFT shines; it allows a single character to change their class/spec/loadout often and completely.

I really like this sort of customizability, as I love the freedom it offers, and I can get more invested in my characters since they really are mine.  Their progress is dictated by my choice, and ultimately, those choices affect how I approach the game as a whole.

Still, that depth does put off some people.  I suspect that it would similarly put off people in MMOs who LOVE their class and can’t imagine playing anything different.  It’s a lot to keep track of, and some people don’t want to bother with learning that much.  There’s nothing wrong with that.

*Quick tangent… I also see class distinctions as yet another way to artificially extend playtime, since you can’t take an existing character and just change their class like you would a Job in FFT.  You must start a whole new character and grind through the levels.  The ability to change your class completely in an MMO doesn’t rob you of identity any more than the ability to change your spec or gear.  It’s your character, and you can always just stick with one class, even if there are options to change.  When there are no options, though, the player interested in exploration of game mechanics is unduly forced to jump through altitis and grind hoops.*

One of the game designs that I’ve toyed with in the last few years is a Tactics-esque game that has a FFT/BT level of depth for character customization, but has what I’m calling Autopilot Character Development.  For those who don’t want to make those choices of how to build a character, there would be “templates” that could be assigned to a unit, automating that progress, allowing the player to just focus on the tactics and strategy inherent in a larger campaign/storyline.

For example, a unit might be given the Scout Template, which would automatically assign them to the Scout class for a while, as it learns some Scouting abilities, then later, assign it to a Ninja class where it can learn some greater evasion and attack abilities.  At any point, the player can turn off the Template and take control of the progression, but if they just can’t be bothered with the minutae inherent in the system, the Autopilot lets them get on with playing the upper-level game.  (Here “upper-level” meaning higher concepts, like tactics and strategy, not high unit level.)

Put another way, this sort of Template system could be overlaid on an open skill system to create a loose sort of “streamlined” class-based system.  UO could become Diablo, as it were.  The key here is that you would always have the option to go back and take the reins, mixing and matching to make your Scout dabble in magic or your Barbarian toy with bows.  This, of course, means that you would also be able to change pretty much everything about the character, from the most basic stats (the prototypical SRT, DEX, whatever) to skill levels to combat skillset (a limited set of usable abilities, like the FFT system).

Is it a lot to keep track of?  Of course it is.  Is it a lot to dig into and potentially have fun with?  If done well, definitely.  Is it good design?  I think so, largely because of the experience I’ve had with games.  (Of course, this mostly applies to those games that require a huge investment of time and character building.  Team Fortress 2 and Smash Bros. work because each round of playing with a class only takes a few minutes.  When that play session extends to hours, weeks and months, it’s onerous to think of “replay” as “rolling another class”.)

I played Titan Quest through as a Sage, a Hunter/Storm ranged DPS machine.  I used Hunter as my “main” class because arrows are infinite, and I could attack at range without burning through mana reserves.  I used Storm to augment that plan, buffing my offense with elemental punch, making my basic ranged attacks sufficiently powerful to kill all but the hardiest enemies long before they got to melee range to bother me.  Ranged enemies went down even quicker since I had great range and high damage… and they were typically slow casters with little defense.  I had a blast, but once I finished the game, I wanted to try another class build.

I didn’t want to spend the time grinding through the lower levels of the game building up a new character, though, playing old content just to see how another class would approach it.  So I found a little program called the TQ Defiler.  It let me edit my character, changing his class to anything I felt like.  I would not have played the game as much as I did without that freedom.  In my younger, stupider days I might have jumped back in with another character from the very start, but with life constantly intruding on my gaming time, I don’t have that luxury any more.  Of course, the TQ Defiler also allows for other sorts of hacks which make the game much easier or harder, but the part that interested me was the class swapper.  There is a “respec” option in the game, but it only allows you to change the way you’ve allocated your skill points, not change your class or secondary, and the cost in game currency increases with each use of the service.

Why?  What does that add to the game?  “Replay value”?  In my time-constrained world, playing through the same content with a different approach is pretty low on the replay value scale.  Yes, it’s technically “replay”, but the bulk of that sort of replay is just repetition, which never sits well with me.  (Mostly because DIKU design is very repetitious to start with; repeating the repetition just gets too stupid too fast.)

“Class identity”?  Thing is, if you have the option to change, you don’t lose that identity; those classes and builds are still there, you just gain the ability to make more choices in the game.  Remember, I like choices.  Purist players in a freeform system will always have the choice to stick with their initial choice, but it doesn’t work the other way; those who want freedom can’t drag it out of a class system without a hex editor.  (Which is effectively making the game behave in ways it wasn’t built for, but arguably should have been.  That sort of hacking doesn’t work in MMOs, since the admins tend to frown on it, banhammer in hand… understandably so, if disappointingly so.)

In a freeform system with Autopilot, you could let the Templates handle the minutae of maintaining a “class identity”, and just go ahead and play your class.  Those who want to do something more freeform could use the Autopilot a bit, or just go all in and do their own thing.

Guild Wars already has something somewhat like this with their Build Templates that you can save and load when you do your “free respec” thing in any town.  They are shorthand precooked “builds” that can be used at any time you would respec, so you can quickly change from a “farming” build to a “questing” or PvP build.  You can also change around your “attribute” numbers willy nilly, to accent your particular build of the moment.

I’m just extending the concept to push that freedom into more aspects of the game, all the way down to the most basic of character customization, the “class” choice.  I’ll reiterate, though, I’m talking about adding choices, and adding an Autopilot for those who want the more constrained experience.  This system wouldn’t destroy the ability to make a killer Rogue or buffalicious Tank, it would augment the game as a whole to allow for more variety and player ownership of one of the few things they truly can control; their character or team.  And yes, this design ethos would apply equally well to a Tactics team-based game as to an MMO.  Any game that uses classes or jobs could benefit from this sort of freedom.

I know, some people wouldn’t like that sort of freedom.  Some want strict predictability and/or relatively simple decision making.  That’s the point of the Autopilot, to let those players just get on with playing the game.  For those who want to dig deeper, though, why not let them do so?

Read Full Post »