Tales of the Rampant Coyote

Adventures in Indie Gaming!

Roll a D20 to Hit: Using Dice-And-Paper Rules in a Computer RPG.

Posted by Rampant Coyote on July 7, 2011

In his email to me, Craig Stern referred to his article called “The Battle System I Wish RPGs Would Quit Using” as flame-bait. He argues that the old Dungeons & Dragons rules model has long outlived its usefulness. Left at that, he’s probably in good company – even Dungeons & Dragons itself has moved pretty far from its starting point.  (And yeah, dude, THAC0 – “To Hit Armor Class Zero” – has been dead and buried for well over a decade…)

Virtues of a Good RPG System

He points out the reasons he isn’t happy with a D&D-style game system. All of his points are valid ones. If I could turn his negatives into positives, here are the virtues that a CRPG system should possess:

  • Streamlined – A good CRPG system should be elegant and well-organized, not ad-hoc as the old D&D systems were.
  • Player Understanding – the resulting effect of the (hopefully streamlined) rules should be easy for the player to deduce. The exact formulas don’t need to be comprehended in their entirety, but the player should have a feel for what happens if you tweak the variables.
  • Not Too Random – Too much randomness and it becomes a game of luck, not skill.
  • Not dependent upon the mercy and creativity of a living human game master

Historical Virtues of the D&D System in CRPGs

I agree with all of these points in principle.  I do take some exception to the contention that the D&D rules were somehow the antithesis of these principles, however. Let’s go back to why D&D style rules were used in the first place…

1) Player Familiarity. Most CRPG players in the 70’s and 80’s played D&D. They understood about rolling stats on 3D6 and understood the 3-18 range pretty intuitively.  They understood that a +1 to hit meat about a 5% better chance of hitting something. They knew what a cleric did, and what a fighter did. While it was still critical to RTFM, the familiar rules system was as important a baseline for RPG fans as WASD controls is first-person shooters are today.

2) Simplicity. The original D&D system – the “White Box” era that most early RPGs were based on – was actually pretty dang simple. Actually, in many ways it was too simple, and the rules didn’t foster the kind of emergent complexity Craig talks about. It spawned a thousand home-brew rules sets, of which the best ideas were incorporated (if often poorly) into Advanced Dungeons & Dragons years later.

3) Designer Familiarity. Let’s face it, it’s far easier to build on existing, well-known foundation than it is to come up with a new system from scratch. This is no less true today – why do you think the stores are overwhelmed with games that adhere to the same basic gameplay? The D&D rules were known to designers – both in its strengths and weaknesses – and almost as importantly, it had been tested and known to work. Not perfectly, maybe not even well, but tens of thousands (eventually, even millions) of gamers had put it through its paces.

Sloppy? Moi?

So those are historical reasons for using the D&D system. But let’s take a look at some of his other criticisms:

Is D&D sloppy and inscrutable? Maybe, but compared to what? Many RPGs I’ve played, with custom combat systems, are even more inscrutable, and aren’t transparent enough for me to judge whether or not they are sloppy. Maybe I know that when my strength goes up by X, my damage with my current weapon goes up by Y.  But really, while I like seeing how numbers change, I don’t track relationships and I don’t really understand the relationships between values. This is especially true with action-RPGs, where I may not even be sure that a particular stat does anything. In fact, I think it was Ultima VII where the Dexterity score literally wasn’t implemented to do anything in-game. But they left it in for the role-players.

It’s only because of the visibility into the D&D system – or ability to deduce similar operations in other CRPG rules systems – that we can perceive any sloppiness in the system. And then there’s the question of what really IS sloppy. You can argue that an exception-based ruleset is sloppy, but I think it’s the exceptions that make the game interesting. If every magical attack does exactly 5 points of damage per spell level, there are few interesting decisions to be made there. But if this one spell actually does more damage than that, and breaks the consistency of the system – well, that’s interesting.  Assuming that the extra power comes at a cost, I’d not call that sloppiness. I’d call it an interesting decision.

Chess has some significant exceptions to otherwise simple, straightforward rules as well. What about the pawn’s first move allowing 2 squares of movement instead of one? And the en passant rule needed to make that work? Pawns in general are pretty unusual compared to the other pieces. And castling? Those do seem, to me, to be departures from what was otherwise a very simple, streamlined set of rules. But I think they make the game much more interesting.

Sure, there’s a point in any rules system where too many exceptions could become a convoluted mess.  But most implementations in the past of D&D-style rules variants into CRPGs were fairly basic. I don’t know that many got too carried away. If anything, most really tried to simplify the D&D rules system (and add their own variations) rather than going overboard with the complexity.

Randomness  (In Moderation) Is Good

Now onto the biggest subject: Randomness. Craig’s game doesn’t have randomness in combat resolution. There are many games that do not. They can be a lot of fun.

But for me, a lot of the fun (and skill) in RPGs comes from manipulating the system to get luck on your side. And the chance of failure – no matter how carefully you’ve tried to work the odds – keeps things exciting, and demands risk management. Sure, you may be 90% likely to kill the dragon before it gets the chance to attack again… but what happens if you don’t? Can you survive another onslaught of its fiery breath? Is it better to plan accordingly, sacrifice your chance of a quick kill to reduce your vulnerability?

Doing this does require an understanding of the rules – one of those virtues listed above. Should the player use up a valuable spell point to cast bless at the beginning of the fight, or save it for a critical heal spell later in the fight? In a fully deterministic game, you may be able to predict the exact results in advance (if the AI is also predictable). In a game with randomness, you are playing the odds.  For me, the latter feels more “realistic” and less like a board game – we never understand all the variables going into a situation. A major reason I quit playing the Hero rules system in our dice-and-paper sessions in favor of 3rd Edition D&D was that the bell curve for Hero made things a little too deterministic for all but a narrow range of values. It got boring.

My most memorable and exciting game moments in RPGs are often when our group succeeded – or failed – via longshot odds. Like the time the party rogue managed to get really lucky on the first shot with an arrow of undead slaying against a powerful vampire wizard. Or when the monk, cornered by a large monster and unlikely to survive another round against it, shouted for the magic user to go ahead and fireball them both, because “he could take it.” A 20% chance of failure bit him, as well as a surprisingly high roll on 7d6 for fireball damage that took him down to exactly -10 hit points. Or there was the time when a series of brash decisions led me to the point where my survival depended on a 50/50 dice roll to see if I could jump to hyperspace out before the incoming missiles hit.

Sure, the failures due to random chance suck. But the awareness of the chance of failure is what makes success much more entertaining.

That doesn’t invalidate the point about D&D being too random. Some randomness is good. Pure randomness is bad. But the exact threshold of “some” is a fuzzy, subjective thing. Is D&D too random for CRPGs? Possibly. For some players, I think so. But I played a little Icewind Dale recently – using the Baldur’s Gate combat system that remains one of the most authentic implementations of the Advanced Dungeons & Dragons 2nd edition rules. In one fight, I found myself getting my butt kicked repeatedly until I buckled down and quit depending upon luck. I paused the game, cast the buffs, applied movement and positioning to my advantage, focused on range and line of sight issues, and otherwise got nice and tactical. Suddenly a combat that I couldn’t win without losing half my party became quite manageable. While I didn’t have perfect control over the situation – the randomness forced me to change plans in mid-battle – the impossible fight became merely challenging.

I think anybody who has played Knights of the Chalice to completion will be well-prepared to make a persuasive argument about the skill necessary in this game based on the core ruleset powering 3rd edition Dungeons & Dragons.  Is luck a major factor? Definitely. But so are the skill demands.

Is this rules style too random? I think for my own tastes, the D&D rules are probably a little too random. I think most gamers wouldn’t mind if the curve was bent a little closer to average. But I don’t see it being a big problem.

Optimal? No. Functional? Yes.

Craig’s concern seems to be that modern designers are throwing out the baby with the bathwater, eschewing not only a less-than-perfect older system that was fairly ubiquitous in RPG design 20 years ago, but throwing out turn-based gameplay and everything else associated with slinging the ol’ 20-sider around. One could argue that virtually nobody is actually using the older D&D ruleset he’s complaining about now — Eschalon, Knights of the Chalice, and some roguelikes are the only ones that even come close in recent years, to my knowledge. And no “official” D&D computer games are going to use anything other than 4th edition rules going forward, which IMO don’t have much in common with the older system anyway.

I don’t think the D&D system – especially older editions – are perfect for CRPGs, either. I would definitely encourage game designers to experiment and come up with fully original rules systems if they feel so inclined. But if all else fails, I think starting with a variant of a familiar rules system – even old-school D&D – isn’t a bad way to go.


UPDATE: Craig has a follow-up to his own post now up: “12 Ways to Improve Turn-Based RPG Combat Systems.”

Filed Under: Design, Dice & Paper - Comments: 12 Comments to Read

  • Adamantyr said,

    I reached many of the same conclusions with my vintage CRPG design.

    Dungeons & Dragons influenced a lot of designers in the day… and what’s funny is, it would have been more efficient for them to implement their own systems rather than copy D&D.

    The main reason why they didn’t get original is the same reason modern-day games aren’t original: you want to copy the design that works. In the old days, the designers/coders lacked the confidence to deviate from a known working pattern. These days, it’s the marketing team demand they keep things to pattern because it’s more stable for profit to use what worked before.

  • LateWhiteRabbit said,

    I have to say, after reading the article, I agree with Craig – I played AD&D for 15 years, but the points he makes are all sound.

    D&D did and does have a lot of weird and inefficient aspects to its rules. (I will be speaking mainly of 2nd Edition AD&D, since that is what I played the most and for the longest time.)

    THAC0 was monstrously illogical, causing a situation where suddenly lower was better on rolls and you now had a stat that could have a negative value – and that was supposed to be good!

    And stat values themselves are funky. Why 1-18 for stats? I understand it has to do with the 20-sided dice, but why not 1-20? 1-10 would be more logical and obvious and 1-100 is intuitive to anyone. So why not use those values and one or two 10-sided dice like Craig suggested? 10-sided dice were certainly called for elsewhere in the rules.

    And what about XP? The first few levels were logical – moving in blocks of 1000 XP for each level. Then suddenly each level is an arbitrary giant amount to reach. I know this introduces a curve where you gain levels faster at the beginning and slower later on, but it causes an arms race in XP that is ridiculous.

    And treasure equals XP? It makes no sense and means the richest people in the game world are the most strong and talented and powerful – physically and mentally. It’s the creator god of that universe was a diehard capitalist. I remember realizing how stupid the whole system was when one of our players PANHANDLED his way to 2nd level.

    And the complaint about rules and exceptions and a crap-ton of dice rolls for everything is incredibly valid. I can’t count the number of times we had to stop in the middle of playing and look up a rule or clarification in the Player’s Guide – no wait, it was the Dungeon Master’s Guide – oops, no, I guess it was in the Fighter’s Handbook.

    And combat took so long, with so many calculations and rolls outside of roleplaying that my gaming group decided not to play with more than 4 people, and preferred 3. It moved things along and kept everyone focused. My brother runs a gaming group with 10 people, and it was indeed a turn-based RPG when I played a game with them (making it 11 players). I would make a decision or take a swing in combat and then wait 15-20 MINUTES to get to do anything again. I know, I started timing it. Blah.

    Often CRPGs will implement nearly all the D&D spells, but most are useless outside of a PnP roleplaying session. Others are handicapped. So I can use Polymorph Self to become an ogre in combat, but I can’t use it to disguise myself and bypass guards? I can use Detect Alignment to determine an NPC’s alignment, but I can’t call out the NPC on being evil or inform anyone? I just have to wait for them to betray or attack me? Some of that is poor game design, but the fact remains that designer’s broke their own games by slavishly copying D&D rule books.

    And the lack of a DM in CRPGs is indeed a great reason why the rules are a bad idea to use. DM’s make things more interesting, they can fake dice rolls to increase drama or save a party from a complete wipe, they can decide that ‘no’ a dragon is not an appropriate random encounter, and they can tailor adventures and situations to the players’ skills and interests. And I mean, how fair are enemies with instant death spells like Petrify, Power Word Kill, any wizard with 9th level spells (which are only available to 18th level wizards!! Argh, what is this? Why not have a spell level for every wizard level?)? With no DM to “hold the reigns” and use these powers at dramatically appropriate times, the computer will just spam them, because it is doing the most tactically sound strategy and the game makers copied the monster abilities verbatim out of the Monstrous Manual.

    We always preferred roleplaying to combat, and CRPGs based on D&D rules tend to be more “kick-in-door” dungeon delve type adventures than roleplaying. It does tend to become all about the min-maxing of stats and collecting uber-loot and weapons.

    TSR itself went on to make better PnP systems than D&D. D&D is so very . . . primitive. A prototype of PnP gaming that being the first of its kind had a lot of bugs, flaws, and kinks to work out.

    I think a lot of CRPG designers where and are DMs of PnP games. They wanted to make their own adventures and campaign settings and didn’t think twice about reaching for the books from last weekend’s D&D gaming session, just like a lot of fantasy writers don’t think twice about reaching for Tolkien. Familiar and safe, doing things this way is a lot like writing fan fiction – all that hard stuff like rules is out of the way and you can focus on designing or changing what you consider to be the fun parts.

    Game design is all about understanding what rules DO and implementing them specifically for a reason or effect – not implementing them in every game because they are your favorite.

  • Greg Tedder said,

    I am a huge fan of managable randomness. I am currently playing FFT again, taking weak parties into tough battles taking chances with 26% chance of success. Failures can make for some amazingly complex battles. Too much certainty feels way too artificial.

    And I really agree that exceptions can make the game, having wild variants in options that says, I can nearly kill them with 95% success, or I can turn them into chicken with 65% success. The choice is fun, and even though I really want to turn a dragon into a chicken, yeah, it may kill me but that dragon’s gonna cluck!! 🙂

    Transparency in randomness does improve things for me. Even though I love ADnD rules, I find it hard to keep track of hit rates for 6 characters.

  • McTeddy said,

    I kindly disagree with Craig… at least with the severity of his preaching.

    I find most of his issues to be personal opinion and not awful sins of game design. I’ve always found randomization to be a great feature because it adds excitement to the usual fare.

    I know Craig loves his TRPGs… but not everyone does. You don’t see me attacking his style of games, do you? Of course not! Because there is room for all games dagnabbit!

    Yeah… Dungeons and dragons does have plenty of flaws. But that doesn’t change the fact that it is well known… both to players and developers. This prevents the imbalance that comes from a new game design… and forcing the player to learn a new rulebook every game. Good for both worlds!

    Now I have a question… did he mention one game that used those mechanics in the past 10 years? Call me crazy… but I haven’t seen much for D and D mechanics in any non-D&D game in a looooong time.

    I know he hated on “Balders Gate”, “Planescape:Torment”, “Knights of the Old Republic”… yet… those were all INTENDED to match the tabletop rules for their respective game.

    Having all your properties use the same mechanics is not bad design… because people interested in one form will easily learn the other. Learn the rules once… experience/Sell two separate games… sounds like good design to me.

  • Menigal said,

    Like McTeddy said, he seems to be about 10 years too late. It just comes across as being a bit bizarre and out of touch.

    Yeah, even the more current versions of the game are far from perfect, but you’re never going to make a system that works for everyone. In my experience, most homegrown CRPG rules systems are far more inscrutible, unbalanced, and random than D&D. Of course, these days everything’s an action game, so the rules are increasingly meaningless to players. How often do you get a thick manual that explains how everything works these days? Or even a thin manual that does more than say “Press X to kill monsters”? :p

  • Adamantyr said,

    @LateWhiteRabbit – The ability scores used 3d6 to determine range. This creates a bell curve so that the middle range of values are more likely than the extremes. I think the percentage chance of rolling a 9-12 is around 75%.

    If you roll a d20 for ability scores, you end up with really wacky statistics. It also severely randomizes the game in a way that makes it difficult to establish a baseline.

    Also, treasure = XP was required in 1st Edition, it was optional in 2nd Edition, and suggested only for rogue classes who used skills to obtain the SPECIFIC treasure.

    I think when you were referring to THAC0, you meant the whole “armor class low equals good” system? I totally agree there, Gygax himself admitted they could have reversed it a long time ago and it would have worked fine. The problem is you spend a half-second or more just trying to find out what AC you hit on a roll; some character sheets included a look-up table to make it quicker.

  • EHamilton said,

    I’m perplexed after reading Craig’s post. Apparently it was written in some alternate universe where the game industry is glutted with an excess of OGL-based 3rd Edition and Pathfinder RPG releases.

    If so, please, please, please, could someone kindly provide me with a portal to transfer to that universe? Thanks!

  • Rampant Coyote said,

    The lower AC = Better thing happened because Dave Arneson lifted the rule from a naval game that he liked. And in that one, smaller size = harder to hit. Still wouldn’t have taken much to change it, but it took on a life of its own. It wasn’t until 3rd edition, with so many changes to the system already in place, that they decided to correct it.

    And I agree… I’d love more OGL-based and *any* Pathfinder-based CRPGs. That would be awesome.

  • xenovore said,

    Agree 99% with Craig.

    D&D has always been entirely too ad hoc and arbitrary, and is just about the last RPG system I would choose, especially for a CRPG.

    There are (and have been for decades now) much better systems available, with better intuitiveness and consistency in the mechanics. Systems that are just more fun because you can actually role-play, not fight the rule-set.

    D&D, and all its various mutations, should be retired.

  • Keldryn said,

    I have recently gotten fed up with the complexity and rules bloat of modern incarnations of D&D (4e and 3e) and gone back to the “sloppy and inscrutable” game of the late 70s and early 80s, so I feel compelled to respond to a few things here.


    THAC0 did not cause a situation where lower was better on rolls. You still needed to roll high; THAC0 was simply a target number which got lower as you improved in ability. A 1st-level Fighter needed to roll a 20 to hit AC 0, and a 2nd-level Fighter needed to roll a 19.

    As an aside, Delta’s D&D Hotspot blog has an interesting article on combat algorithms, including THAC0 and 3rd Edition’s ascending AC: http://deltasdnd.blogspot.com/2009/07/what-is-best-combat-algorithm.html. He proposes a “Target 20” algorithm as the most efficient for the old-school descending AC games: d20 + fighting level + AC + modifers >= 20.

    Armor Class originally had meaning (9 was unarmored, 7 was leather, 5 was chain mail, 3 was plate mail). Magic armor bonuses in original D&D (1974) subtracted from your opponent’s hit roll rather than improving your AC and Dex did not give an AC bonus, so there were no negative AC scores.

    As Adamantry said, the range of ability scores (3-18, rolled on 3d6) was chosen in order to have ability scores distributed on a bell curve. For an adult human, 10-11 represents the average ability, and scores towards the ends of the distribution become increasingly less common.

    In the original D&D rules, ability scores were primarily descriptive and did not have many pre-defined effects on game mechanics (+/-1 to hit with missile weapons for high or low Dex, +/-1 hit point per die for very high or very low Con, max number of hirelings and their loyalty base for high or low Cha). The primary purpose of ability scores was to give the DM a basis for making a judgment call as to whether or not a character should be capable of doing something.

    As for XP, the advancement table for each class roughly doubles the amount of XP required to achieve the next level until around 9th level. At this point, the character is said to have reached “name” level (Lord, Wizard, High Priest, etc) and advancement dramatically slows down from that point onward. This isn’t creating an arms race in XP — the game assumes that at this point adventurers have mostly seen and done it all and have accumulated enough wealth to retire from active adventuring. Each class description discusses the sort of stronghold that can be built and the followers who have heard of the character’s great deeds and wish to pledge their service. 9th to 12th level is very high level in original D&D or AD&D 1e; few adventures were published above this range and there were very few threats described in the Monster Manual which would challenge such characters.

    The treasure for XP rule (1 gp = 1 XP) makes sense if you don’t take it too literally. Like most of the early D&D rules, it’s an abstraction for convenience. The assumption is that the most valuble treasures are hidden away in the deepest dungeons and guarded by the most fearsome creatures; thus, higher-value treasures carry a commensurate degree of risk in order to obtain and recover. PCs gained much more XP from treasure than from fighting monsters, so avoiding combat when possible was usually the smartest and most efficient use of resources.

    Many of these assumptions should have been made a whole lot clearer in the rules, but they are there. Treasure obtained at no risk to the player characters should grant zero XP. Sloppy DMing doesn’t mean that the system is broken or stupid.

    @nobody in particular:

    I once wrote off the “Basic” D&D rules (1983) as hopelessly simplistic and the AD&D rules (both editions) was clumsy, rigid, and inconsistent. After playing modern D&D for many years, I have come to the conclusion that it’s the modern rules that are rigid and complex. I used to see it as a flaw of the earlier rulesets that they required an experienced DM to actively make judgement calls and rulings and to be flexible in his interpretation of the rules. Now I see it as a feature, not a bug.

    Even after getting familiar with the rules, my group couldn’t manage to get 4th Edition combats resolved in much less than an hour of play. 3rd Edition isn’t much better. In my 1st Edition AD&D game (where none of the players are very familiar with the rules), most battles are over in 15 minutes or less and we’re back into exploration and interacting with the game world. We very rarely stop to look anything up, despite the fact that I haven’t run this edition in almost 20 years. If you were constantly stopping to look up rules in 2nd Edition, then I have to assume that you were using the plethora of splat-books that were released for that system. The game (all versions of it) run a lot better if you don’t bring all of that stuff into it.



    Despite all that I said above, I would not choose D&D as the basis for a CRPG. Basic D&D (the “Red Box”) is still the most intuitive introduction to tabletop RPGs that I’ve ever used, and I’ve never had a new player not grasp the basics of play within 10 minutes. It’s a terrible system for CRPGs because its enjoyment hinges on having an intelligent human being able to make judgement calls and changes to the rules if they don’t work for that particular campaign. The inconsistent resolution mechanics in D&D can be seen as unintuitive, but their modular nature makes it very easy to make drastic changes to one area without causing unforeseen effects elsewhere in the system. Also, while resolving all tasks on a d%, for example, is consistent and intuitive at one level, it also assumes that all tasks need that level of granularity. Why force everything into using a 1-100 or 1-20 scale when sometimes a d6 is all you need?

    I find that you are only fighting the rules set if you have players or especially a DM who takes the rules-as-written too literally. In that case, you’re better of with rules that are defined explicitly enough to not require much interpretation or judgement. I find that I end up getting very much the same results in-game but with a lot fewer headaches.

    When you have a DM who understands the intent of the rules and players who trust the DM to be fair and reasonable, old-school D&D runs smoother than any other RPG I’ve played, with the rules fading into the background while role-playing takes center stage. I do understand that groups who behave like this are somewhat uncommon… 🙂

  • Adamantyr said,

    @Keldryn – Good points on all of it!

    I agree with you, a paper-and-pencil RPG should not be used as any model for a CRPG. Inspired by, maybe, but not a direct copy. The two styles have mutually different requirements. Primarily, a CRPG needs rigid mechanics to define a finite number of potential actions, an RPG needs an open framework for an infinite number of potential actions, whatever the imagination allows.

    I found with my own vintage CRPG design, it was also partly influenced by the system itself. For example, it’s easier easier to generate a random set of values in the base 2 range. I can bit-mask a full range value to get any of those. Conversely, trying to generate a value from 0-19 would require division and other more complicated mathematics.

  • dcfedor said,

    You made a comment about exceptions making the game interesting, and include an example where spells do exactly +5 more damage each level. That reminded me of my first time playing Dungeon Siege II, where after a few hours and leveling up, I started tiring of the “+5 fire damage” or “+2 poison defense” attributes every magic item or spell seemed to have. Magic never seemed so boring to me in RPGs as they did while playing that game.

    When I started thinking about it some more, I came to a similar conclusion as you: exceptions make magic interesting in AD&D. The completely ad-hoc, often self-contradicting nature of AD&D’s spells is what made it seem like real magic. Reading about mages, spells, and artifacts was like dabbling in arcane arts. You had to cross-reference tomes of knowledge to know what you were doing, and you still might screw up.

    Plus, rules often couldn’t describe what spells and artifacts could do. Complete severing of limbs, polymorphing into other objects, decks of many things that could commit you to eternal imprisonment…it was inexplicable and even sometimes frightening what magic could do.

    Just thinking about it makes me nostalgic!