July 7, 2011

12 ways to improve turn-based RPG combat systems

In my last opinion piece, I provoked a certain subsection of the world of RPG enthusiasts by slaughtering a particularly sacred cow: the D&D-style combat system. A surprising number of people wrote in agreeing with me. Predictably, however, others responded in one of two ways: (1) “So you think a real-time, action-centered combat system is better?” or (2) “Name an RPG combat system that’s better!”

The answer to (1) is easy. No, I don’t think real-time is better. Just the opposite: I prefer turn-based combat in my RPGs. Of the six games I’ve released since I started designing games, five use turn-based combat, and I’m working on two more with turn-based tactical combat for good measure. That should probably tell you something about my tastes.

The answer to (2) is more complicated. I don’t think that there is just one way to do a turn-based RPG combat system correctly. I’ll avoid naming particular games, since I don’t want to give the impression that all RPGs should employ combat in the style of any one particular game. I will, however, discuss the features that good turn-based tactical combat systems have in common, and cite games that successfully employ them.

The Four Virtues of a good tactical turn-based combat system

If you’ve read my last article, this list is going to look familiar. A good tactical turn-based combat system exemplifies the following Four Virtues:

(1) Emergent complexity. It creates complex gameplay out of a comparatively simple set of rules.

(2) Clarity. The immediate consequences of various tactical decisions are made clear to the player.

(3) Determinism. The system is sufficiently deterministic that skilled play using a proper strategy will nearly always result in victory.

(4) Tactical tools. If there is some randomness in the system (which there will be in most cases), the player has sufficient tactical tools at her disposal so that skilled play will almost always trump bad luck.

How can we employ these features?

The Four Virtues of a good tactical turn-based combat system are closely interconnected: with a handful of simple rules, a turn-based combat system can exponentially increase its tactical possibility space, thereby achieving the goals of both emergent complexity and skill-based outcomes to battles.

Clarity, in turn, arises organically if you do this properly: which is to say, if you don’t achieve complexity by overloading your combat system with arcane rules, the player should quickly be able to understand exactly how her actions will play out in combat, allowing her to plan ahead and strategize. (Clarity also depends upon good interface design and appropriate visual cues to the player, but those things are basic to good game design in general, and aren’t worth discussing here.)

So let’s get specific. There is a veritable cornucopia of techniques that game developers have used in the past to make their turn-based combat systems sparkle with tactical possibilities, and I want to see new RPGs start using them with greater regularity.

Perhaps the most powerful technique is simply to

(1) Use space. Adding a spatial dimension to combat increases its complexity exponentially without making it substantially harder for the player to understand. Most people have played  games like Candyland or Monopoly, to say nothing of Checkers and Chess. Everyone (even your mom) intuitively understands the concept of moving pieces between spaces.

By using space in your battles, you add a new dimension to combat both figuratively and literally: the concept of attack range comes into play, and the player gains direct control of actions like fleeing and protecting weaker characters behind stronger ones.

Of course, you aren’t required to have a grid-based (or hex-based) map with movable characters to create a good tactical combat system, but it’s an awfully effective way to introduce complexity using simple rules. This alone will put your game far ahead of most jRPG combat systems.

Then again, spatial combat is not exactly a huge achievement: virtually every western RPG has this in one form or another. Let’s be a little ambitious. Here are 11 other techniques for achieving the Four Virtues that have been consistently overlooked not just in jRPGs, but in western RPGs as well:

(2) Give the player at least six characters. This one is absolutely key, and yet most western RPGs of the past 20 years have missed it. Imagine playing chess with only four pieces–you’d be looking at a game with greatly reduced tactical complexity and far less interesting matches.

Putting more characters under the player’s control pays great dividends in terms of tactics. More characters means that the player can be expected to handle much more involved combat scenarios, and becomes responsible for keeping more characters alive. This naturally gives rise to dilemmas about how to balance multiple objectives with minimal losses, which in turn make combat more interesting.

Also, with more characters under the player’s control, each individual character can be much more specialized. Speaking of which…

(3) Specialize the characters. How dull would chess be if it were played entirely with knights, or bishops? Make sure characters of different classes serve different battlefield roles; don’t just make them all fighters with different hit points, armor and spells.

If you differentiate your classes successfully, your player will have to think carefully about which characters should perform which actions during battle. The Fire Emblem series, especially, does a great job with this.

(4) Specialize the enemies. This should be obvious: if every enemy is a melee bruiser with magical attacks, your player has no reason to prioritize one enemy over the others. Give enemies distinctly different capabilities, weaknesses, and battlefield roles.

(5) Variable distance. Do not always begin battles with enemies 1 turn or less away from melee range! Spacing them out a little will give the player more flexibility to try out ranged tactics, as well as pressuring the player take cover from or flank enemy ranged units.

(6) Directional facing. Make it so that characters are easier to hit and/or suffer additional damage when attacked from behind or the sides. Doing so amplifies the importance of positioning in close quarters, adding another wrinkle to the player’s considerations. It also has a side effect of making faster-moving characters more dangerous, as well as increasing the effectiveness of flanking and pincer movements.

(7) Variable terrain. Used properly, terrain can add new dimensions to player considerations about character positioning, creating natural choke points on the battlefield, providing cover, or even providing bonuses or penalties to the characters who stand on it (Advance Wars and the Disgaea games serve as excellent examples of the latter).

Terrain can also serve as a point of character specialization, with certain classes performing better on certain squares (or being uniquely able to move through them).

(8) Manipulable terrain. But why stop there? Let the player actually manipulate terrain during the battle. Temporary terrain creation occasionally appears in western RPGs in the form of walls of fire, walls of ice, and traps that players can place to form barriers or choke points. Destructible terrain in RPGs is rare, however, and most created terrain lasts for only a short duration.

We can do better than that. Give the player meaningful flexibility to shape the battlefield, creating new avenues of attack and closing off existing ones. This will cause your combat system to accommodate more creative tactical thinking. The player won’t just be thinking about how to place her characters to best make use of the environment: she’ll also be thinking about how to change the environment itself to create those opportunities in the first place.

(9) Resource management. This appears in nearly all western RPGs to a limited extent, mostly in the form of budgeting gold, conserving magic points and hoarding scrolls and potions. However, this standard RPG implementation of resource management isn’t usually too important at the combat level–it’s typically a broader challenge spanning a whole foray into a given area.

To liven up particular combat encounters, use a system that requires players to balance more powerful attacks against other priorities on a turn-by-turn basis. Action Point systems (e.g. the kind used in X-Com: UFO Defense and the first two Fallouts) are great for this, forcing the player to weigh a variety of factors all at once when deciding on character actions. Also good are systems where most powerful abilities use large amounts of magic points, but magic points regenerate over time (e.g. Tactics Ogre: Let us Cling Together).

(10) Give units multiple attack options. This dovetails nicely with resource management: giving each character the option of more effective but more expensive or risky attacks expands the player’s tactical options greatly. Should the player pin her hopes on her character landing a more powerful blow, or should she have him get in a quick jab and leave some resources for the character to defend himself afterwards?

These sorts of small-scale dilemmas are the bread and butter of a satisfying tactical combat system. (Fallout and Fallout 2 provide a great example of how to use this technique.)

(11) Support multiple objectives. I mean this not just in the sense that your combat system should challenge the player with different win and loss conditions; I mean this in the sense that multiple different objectives (not of the win/loss variety) should be  able to coexist within any given battle in your combat system.

This is a simple extension of the idea that a good RPG should have choices with consequences: there should be post-battle consequences for what happens during battles as well. Did she hit a town guard with a misguided arrow? Make it so the player is wanted for assault. Did she successfully protect the manor of the richest man in town? Make the rich NPC give the player a reward after the battle.

Aside from the obvious benefits such a feature creates for immersion, it also creates a richer tactical experience for the player. The goal becomes not just to win–the goal becomes to win while accomplishing as many side objectives as possible.

Because of the effort involved in setting up custom objectives for each battle, you might instead choose to set up your combat system with persistent side objectives such as maintaining character morale (X-Com) or gathering certain resources that can only be gathered during battle (treasure chests in Fire Emblem, captured majin in Eternal Poison).

In Telepath RPG: Servants of God, for instance, slain characters can only be resurrected through the use of soul charges, which in turn can only be obtained by using one character’s Soul Suck ability on a critically injured enemy during combat. The player is forced to periodically use the character who can Soul Suck in order to maintain a supply of soul charges; and within particular battles, she must weigh the risk of keeping an enemy alive long enough to harvest it against the reward of having a one-shot chance to raise a slain character later on.

(12) Allow delayed attacks. Delayed attacks add a new twist to the turn-based formula, allowing characters to attack even when it isn’t their turn anymore. Counterattacks are an effective (and, at least among Japanese strategy RPGs, common) form of delayed attack. By allowing melee units to only retaliate against melee units and ranged units to only retaliate against units at range, counterattacks further complicate unit match-ups and enforce character specialization. (Players will want to attack melee units at range, and vice versa.)

Attacks of opportunity are another form of delayed attack that encourage the player to pay extra attention to her own unit movement and placement, juggling the costs and benefits of expending all of a character’s actions during its turn versus keeping some in reserve in case enemies wander into attack range later on. These are, in short, an extension of the resource management concept we talked about in #9 above. (You can see the attack of opportunity mechanic at work with X-Com’s reaction fire system and with the guard command in Pool of Radiance.)

Interestingly, a couple of upcoming indie games (Frozen Synapse and Fray) actually use delayed attacks as the centerpiece of their entire combat systems. In these games, combat focuses around the player’s ability to guess which units will be where, when.

Well, that’s twelve. I know of no RPG combat system that uses all twelve of these techniques at once, and to be honest, I’m not entirely sure that I’d want one to! That would be one heck of a complicated combat system. But in an environment of RPGs whose combat systems err on the side of simplicity, that’s the sort of problem I really wouldn’t mind having.

UPDATE: Part three has now been posted.

  • Pingback: Game review double-header: Styrateg and Age of Fear: The Undead King @ IndieRPGs.com()

  • Pingback: Weekly Roundup | Line Hollis()

  • Battlejack

    Great article, sure it’s a year old, but it’s points still resonate true. I have worked hard while not knowing of this list to make sure that my own game Battlejack does these things. While my game is more of a board game, I know of one other RPG system (that isn’t D&D) the soon to be available Age Past that does a fantastic job of covering these aspects as well. Keeping these goals in mind when running a game that maybe doesn’t do the best job of covering them itself much more fun and interesting.

  • Pingback: 6 more ways to improve turn-based RPG combat systems « sinisterdesign.net()

  • stinky472

    I have to agree with Mark. If a tactical RPG applied all twelve suggestions, I’m not sure I’d enjoy it so much. I do agree that it would add complexity and require more tactical considerations, but when I think of the classical tactical RPGs that I loved most like Shining Force, AD&D Gold Box games, Fallout 1 & 2, Jagged Alliance, and, although not quite as much of an RPG, X-Com/UFO, then they’re all considerably simpler than the new generation of tactical RPGs like FFT, Gungnir, and Tactics Ogre.

    These games excelled so much IMO because their rules were fairly simple and combat was quick, yet there was a lot of depth and challenge in spite of it. Imagine if we tried improving Chess by adding terrain factors into it or lots of dual-piece moves (knight + bishop attack in a single turn, e.g.). We probably wouldn’t end up with a very popular game. Chess, in spite of its simplicity, has virtually infinite depth. It’s a game that’s easy to learn but can take a lifetime to master, and that’s what keeps it so engaging.

    In addition to this, each of the games I listed above had some special ideas that made combat more interesting:

    1. Shining Force: Shining Force is, in my opinion, the antithesis of a lot of modern tactical JRPGs which often involve more generic characters with loads of customization options. Shining Force really made each character a very unique tactical piece in combat which made you truly appreciate that one character’s particular strengths and weaknesses. Whereas people are making whatever they want like a knight with dual monk abilities in FFT, Shining Force presents you with a one-of-a-kind character like Adam or Bleu whose abilities and attacks are not shared by any other character. It was also quite simple — it didn’t make terrain such an important factor, it didn’t have dual team abilities where two or more characters would combine their abilities to form a single attack, etc.

    2. Fallout 1 & 2: Fallout only put you in direct control of one character, yet it had pretty interesting tactical combat primarily because of its intricate targeting system and the character customization abilities which really allowed you to tailor your character to a particular fighting style. Fallout is an example of a fairly non-deterministic system with lots of random factors — a single critical could really change the tide of a battle and yet it balanced this in a way where the luck factor wasn’t overbearing.

    3. X-Com: X-Com really utilized line of sight in a way where it really made a significant tactical difference. Being able to do things like put a character on guard and automatically shoot anything which comes into sight changed the way we played the game considerably to a point where we’re not necessarily charging into the middle of a fray with guns blazing.

    4. Jagged Alliance: Like X-Com, Jagged Alliance really made terrain important and
    combined this with character stances (crouching, prone, etc). It made it
    important to seek cover to crouch behind, e.g. This is an example of
    terrain done properly IMO as opposed to a single tile that just gives
    you a 30% defense bonus when standing on it. Also like X-Com, it also made
    line of sight very important.

    5. AD&D Gold Box games: these games had some interesting concepts of their own like the fact that when a character is next to an enemy and turns to move away, the enemy gets a free hit. That kind of system made you think carefully about how you positioned your weaker characters since if an enemy even moves to an adjacent square, that weaker character can’t run away without risk of being hit: he’s locked into place. Combine this with the bandaging system to prevent wounded characters from bleeding to death and the system was quite interesting. It also made status effects like sleep a lot more useful (and dangerous when used against you) than most RPGs since a sleeping character could be dispatched with a single blow, e.g.

  • stinky472

    Real-time combat sacrifices control. It’s also rarely done right IMO. I hate the hybrid wannabes where it’s real-time but you can’t issue orders fast enough unless you PAUSE. PAUSE, cast a spell, RESUME, wait for spell to finish casting, PAUSE, cast another spell, RESUME… That PAUSE/RESUME style of play is choppy and distracting — I’d like to see a realtime engine that plays at a pace and offers intuitive enough, easy-to-access controls where pausing is never necessary.

    Role-playing is really about decision-making which inevitably comes down to a menu. You cited Dragon Age as a game you like, but in spite of featuring real-time combat, it presents you loads of menus and choices to make outside of combat. Some of us like being able to take our time to make such choices in combat as well and to really feel like we have 100% control of the situation. The real-time strategy game sacrifices control but lets you sit back and watch a lot of what’s happening and respond which is a bit more frantic IMO, and it’s not really faster if that reduced amount of control is causing you to take longer to win a battle.

    As others mentioned, realism doesn’t always equal fun. In fact, the overzealous seeking of realism often makes a game less fun (take the constant hunger mechanics introduced in Ultima 7 which basically made the player constantly have to buy and drag/drop food over his characters). Of course if I’m playing a flight sim, that’s an exception and it’s all about realism, but making a gun fun should always precede making it realistic otherwise. I wouldn’t enjoy a game that required me to stop at a toilet and relieve characters every 30 minutes because I have 6 characters in my party who all want to pee at different times.

  • stinky472

    @CragStern Have you ever played Shining Force? I like your ideas but I think if you haven’t, you have to check that one out. It’s an extremely simple tactical RPG yet it’s a beloved classic and considered by many to still be unrivaled. It might offer you a different perspective or at least give you an idea of how simplistic you can go without losing the player’s interest.

  • stinky472

    I think a trouble with permadeath is that most players would probably be tempted to reload. How many of us actually had enough integrity to avoid reloading if our best character in X-Com, after spending hours and countless missions to improve, dies in battle because of some really goofy mistake like a misclick or if you had one of those battles where that one last alien can’t be found and you drop your guard and run into a critical shot?

    I think permadeath only makes a lot of sense in games where the player didn’t devote so much time to developing a particular character or where the consequences of losing that character are interesting in their own right or as an option for hardcore gamers, or in a game where saving/loading mechanics aren’t available (ex: online game that saves automatically and does not offer the ability to undo a mistake). I think too many of us mortals would be too tempted to load a previous game otherwise.

  • CraigStern
  • stinky472

    Ooh, glad to find a fellow fan! Where I might differ a bit in opinion is that I think the simplicity of SF actually makes it a lot better than many tactical RPGs out today.. so much so that I was recently playing Gungnir (similar engine to FFT) and after a couple of days, I got bored and stopped in the middle and had this terrible urge to play Shining Force again.. and I did and played it all the way through with equal delight as I did decades ago.

    Of course we can’t help but love those games from our childhood, but it’s rare for me to revisit them and find the interest to play them all the way through, especially when I have a new tactical RPG right next to me that I could be playing instead.

    A lot of modern tactical RPGs get so deep into customization.. players have 6 separate pieces of equipment they can use (in SF it’s just one), the spells available to them span pages and pages (SF only support 4 max per character), the characters are often a lot more generic in terms of their particular abilities (the player customizes them instead to what he wants), etc. Characters have special dual combo abilities where two or more characters coordinate a single attack, we have scissors-paper-stone type systems where particular types of attacks are better suited for a particular type of defense… All that extra overhead can be off-putting at times and the simplicity of a game like Shining Force keeps drawing me back when these newer tactical RPGs start stressing me out and making me feel bored.

  • CraigStern

    That, to me, is the essence of my argument for elegance: achieve complexity emergently by using relatively few rules and having them interact in interesting ways, rather than by loading your game down with tons of unnecessary stuff. Shining Force errs on the side of fewer rules, and the result is a much more accessible game.

  • Pingback: Game review: Dark War Strategy Chapters 1 and 2 @ IndieRPGs.com()

  • Troika Moira

    The system doesn’t have to be complicated to use all twelve. Troika Moira uses ten out of the twelve – there’s no option to delay the attack (though there’s nothing stopping the narrator from allowing players to do so) and it doesn’t deal much with terrain. HOWEVER; the base of Troika Moira combat is five different attacks – you choose one depending on what you think the enemy is choosing. It’s a diceless system, with a few resources, but mainly based around (1) placing the character at a proper distance and (2) choosing the right way to attack.
    The styles mean that there are (broadly speaking) 10 different character types, though it’s easy to mix and match or choose multiple specialisations. If you have an interest in this kind of system I’ll e-mail you a free copy. I’ve just made this RPG, and I’m about to do a first print run – so I need all the publicity I can get.

  • Troika Moira

    Oh – while the terrain isn’t exactly a combat feature, I should have mentioned that the magic is entirely scenery based. The magic is innert until it interacts with the environment, so grasping vines (in the jungle), freezing lakes and crumbling taverns are at the heart of the magic system.

  • Pumpkinhead

    Final Fantasy Tactics has all of these. Provided there are only two occasions of #8, and #2 is only technically true of when you have “guests” in the party. (The Geomancer class might also qualify for #8, or maybe a combination of #8 and #10.)

  • Pingback: Interview with “Telepath Tactics ” Creator Craig Stern()

  • Pingback: Weekly Roundup | Line Hollis()

  • Pingback: Turn Based Combat – Ways To Improve? » Tim In Development()

  • Harold_Smith284

    Actually, Battle for Wesnoth already does all of these things:


  • Paladox

    A lot of these are impossible or entirely too complicated to use for those who are less experienced with RPG Maker, but it does make you think a lot about the possibilities of what is or isn’t possible. I’ve been struggling with ideas on how to make the combat more interesting in my RPG and this has actually given me an idea on how to improve it, so thanks for writing this. 🙂

  • Pingback: Devicezero Game Dev()

  • Dan

    These seem reasonable ideas for a tactical RPG.
    Anytime you add features to a game I think it’s important to reflect on what you hope the feature will achieve in-game. What will it incentivise the player to do? And are those actions enjoyable for the player or tedious (grinding, endless min-maxing equipment, inventory management, save scumming etc)?
    These techniques mentioned seem well reasoned in that regard, especially for the games you’ve made 😀

  • Pingback: ¿Está el combate por turnos obsoleto en los rpgs? – Laboratorio 32()

  • rdococ

    If you ask me, terrain is a bit of a stretch. What about a battle system that takes place in the overworld? Swinging a sword would involve facing the enemy.

  • CraigStern


  • Scott

    I believe Final Fantasy Tactics uses all of these and is a great example of them used right.

  • Visage

    I’d just like to point out that Rebelstar Tactics for the GBA uses all of these to a pretty strong extent, only really being weak on the two terrain-based ones… but not very.

    It’s a pretty great game, IMO.

  • Mark

    i know its been six years or more since this was posted but, I have to say, i wouldn’t hate turn based as much IF what you mentioned was implemented…. like final fantasy, and all the others were simply just repetitive button pressing, I had a controller for the ps1 that i only used for this (since i didn’t have two people to play with normally.)