Chad Perrin: SOB

27 July 2008

10 blunders in RPG design

Filed under: RPG — apotheon @ 07:02

This is part of my RPG series of entries here at SOB. See the inaugural entry in the series for more details.

I have some pet peeves in RPG design — things that show up in games all the time, and annoy the heck out of me when I run across them. These are the sorts of things that I tend to eject from games I’m running, for reasons of smoothing out game play, unifying systems, improving suspension of disbelief, and so on.

1. calculations with division in them

Any game mechanic that requires you to use division is a broken mechanic. This is mostly the case because of the rounding problem — one has to figure out whether to round up or down. For some reason, it seems most game designers prefer rounding down in the majority of cases, which only makes things worse by making players feel like they’ve been robbed; the best handling I’ve ever seen of the rounding problem is to say that any time a higher number helps the PCs, round up, and any time it hurts, round down. Players like to feel like they’re favored over nameless bit-part NPCs sometimes.

There are other problems with the rounding problem than simple confusion over which way to round (which is of course compounded in games where for some stats you round up and for others you round down) and the idea that your players may feel put-upon by rounding down. For instance, division often takes longer than other arithmetic operations, especially when you’re dividing by some value other than two. It also introduces problems with complex arithmetic formulae, such as when trying to remember whether a bonus is added before or after the division (if the rules even specify it at all; I’ve seen some that don’t). Another problem is just the tedium of it, and the fact that if some part of the formula changes, the division needs to be recalculated.

Finally, there’s the metagaming problem: when you divide some number to get some other number, and that is the number used to determine success or failure, you create a situation where players are motivated to min/max their stats more. If you have a stat of 10, and another of 11, and both of them are divided by two to yield a bonus added to some common types of rolls, you’re subtly encouraged to increase the 11 rather than the 10 by the simple fact that improving the 10 by one point won’t make any difference if the rule is to round down. If the rule is to round up, the relative encouragement for increasing each stat is reversed.

2. unsmooth progressions

When some modifer progresses in a manner that isn’t smooth — such as in D&D 3.5’s attribute modifiers, where you get a +1 for every two points the attribute increases — you run into that metagaming problem again. You’ll be subtly encouraged by the way the rules work to favor improving odd attribute numbers, rather than even numbers. This problem often manifests at character creation, too, particularly if your game uses point-buy to generate starting attributes rather than random die rolls.

3. closed progressions

A point I made in a previous SOB entry, Open Ended Rolls, is that closed progressions are often problematic from a number of perspectives. Negative effects I might mention include:

  1. damage to suspension of disbelief because of the unnatural form of closed progressions and the effect they have on probabilities

  2. the frustration of being literally unable to succeed at some tasks (and similarly of someone literally being unable to fail) in some cases or, alternatively, two characters of wildly divergent skill levels being equally likely to succeed or fail in some circumstances

  3. the annoyance (in the case of closed stat progressions, as opposed to roll progressions) of not being allowed to continue improving a given stat

In the case of both rolls and stats, it is far better to use a system that makes progression past a “sweet spot” range very difficult than to simply apply an arbitrary limit to how far you can go. Rather than use a closed d20 roll the way D&D does, employ a system that allows the roll to be expanded in exceptional cases, such as my Nd9 critical system. Rather than letting a character spend one skill point to increase a skill by one rank, with an upper limit of 5 (for instance), increase the cost of skill improvements for each rank above 0 so that progression cost becomes prohibitive after 5, but a sufficiently motivated player can still increase the skill if that fits the character concept.

4. discontinuous progressions

If you have an open ended progression, don’t make it discontinuous. Rolling 1d10 and adding another 1d10 if you happen to roll a 10, as in the case of 7th Sea (for instance), is a bad idea. An even worse example is Rifts’ treatment of attributes that are 16 and higher. Such discontinuous progressions break probability curves in ways that can prove problematic for suspension of disbelief and smoothness of game play. I discussed this in the Open Ended Rolls entry as well, under the section heading “some problematic open rolling systems”.

5. fragmentary systems

To the extent that a game doesn’t tend to use a centralized system, it can cause problems. One of the best (worst?) examples of this is the scattering of unrelated systems for special combat maneuvers in D&D 3E. The grapple rules are the crowning achievement on that heaping pile of crap. Perhaps the single most important thing Pathfinder RPG has done, in terms of fixing D&D rules, is to unify most (if not all) the special maneuvers rules into a single system based on a Combat Maneuver Bonus. While the D&D 3E grappling rules are a broken mess of bad idea all by themselves — so much so that, as printed in the books, the grapple rules are literally unusable — the fragmentary nature of the special maneuvers rules in general are a huge problem.

6. overly complex systems

Here we go with the grapple rules from D&D 3E again. Aside from the fact that it’s just one of many fragments in the special maneuvers rules for combat, it is in and of itself so damnably complex that the guys who wrote up the rules for the D&D 3.5 Player’s Handbook seem to have lost track of what they were doing halfway through writing up the system and ended up with a system that doesn’t actually work. I don’t just mean that it works in a way that breaks game play; I mean there literally is not a complete system.

By making certain assumptions about what the writers must have meant, a system can be gleaned from the text in the book (and the SRD), but I’m not really 100% certain that the resulting system is exactly the system that was intended. I’m not even entirely sure the various writers and editors had the same thing in mind when they agreed on the final form of the grapple rules in the PHB.

The system goes on from that point to require many, many repeated rolls, and much of the time the result is that you have hurt your chances of achieving your aims, even if you succeed. Yeah, this system is far too complex.

7. arbitrary systems

Obviously, all systems are going to be “arbitrary” in terms of design decisions, to some extent. There’s arbitrary, though, and there’s Arbitrary, with a capital A, some bold letters, and some oblique emphasis. A mild form of “abitrary” is demonstrated in the choice of, say, a d20 to resolve contests of skill. A more severe, and problematic, form of “Arbitrary”, is the Vancian magic system used in D&D, where nobody can come up with any reasonable explanation for why you can cast three first level spells and one second level spell — and that second level spell can’t be given up to provide more first level spells, or vice versa. It’s stupidly arbitrary. It breaks suspension of disbelief for no good reason, unless you think “Well, I’m too lazy to think this through any better!” is a “good” reason.

8. dragon stuff

Dragons are supposed to be something special. They’re gigantic, fierce, mighty beasts, possessed of great power, shrouded in mystery, rare and frightening like nothing else in the world. They’re iconic, and slaying one is a Big Fucking Deal.

. . . unless you’re using one of the scores of stupid additions to D&D that have cropped up. Everywhere I look, I see dragon this, dragon that, all kinds of dragon things all over the place. You can team up with a Dragon Shaman, fight a Dracotaur, or even be a Dragonborn Paladin at first level. There are so many different types of dragons that, assuming there are enough of each type for them to be able to breed and propagate the species over the millennia, there must be literally millions of dragons in the world. It even started really early, with the Dragon Disciple in the DMG.

This isn’t really a complaint about all the dragon-wank in D&D. Well, not specifically. The dragon nonsense just makes for a handy example, and a particularly egregious example at that.

The real problem is that of utterly failing to capitalize on the real strengths of the core mood-setting tropes of your RPG genre. Dragons should be immensely powerful, frightful beings whose motives cannot be fathomed, creatures rare and mysterious, opponents only the greatest of heroes or of fools would ever dare face in combat; they shouldn’t be the parents of 50% of your adventuring party. Magic is similarly a thing of mystery, a powerful force, dangerous to wield and even more so to oppose when facing the evil wizard at the end of an epic adventure; it shouldn’t be a commonplace component of things you can buy and sell in the corner store, like the integrated circuits in an iPod or disposable cellphone today.

Most of the time, the “coolest” stuff in your game should also be among the rarest — otherwise it ceases to be “cool”, and instead becomes merely “common”.

9. incomplete explanations

In some of the later Monster Manuals, descriptions of the creatures ended up being mostly neglected, in favor of a picture and an italicized bit of flavor text that utterly fails to provide a generalized overview of the way the creature would come across. Many of the fire spell explanations in the game never touch on the possibility of igniting flammable materials. The actual relationship between the Spheres and coincidental uses of magic in Mage: The Ascension was very much undefined in the books, leading to in-depth discussions of the philosophical and game mechanics underpinnings of coincidental magic — and, while those discussions were interesting and often fun, they lost a lot of that fun when they devolved into flame wars, and they were only necessary because there simply wasn’t any “official” explanation to which the rules interpretation of one’s game could adhere.

Finish what you start.

10. playing favorites

We all have our favorite game concepts. I really like orks (when done right), for instance, and I’m not a big fan of mainstream D&D gnomes (or Dragonborn). If I include both orks and halflings as core races in a game or an expansion product for a game, though, I sure as heck am not going to do so in a way that does the Mary Sue dance with orks and makes a mockery of the very concept of halflings.

Sometimes, I feel like I’m the only person who thinks that way, though. In D&D 4E, dwarves are the gimpy red-headed stepchildren of the game while Dragonborn and the three (count ’em, three) elfish races available to players are obviously designed to showcase their kewlness factors. The fact that White Wolf basically buggered every single Vampire: the Masquerade player who liked Clan Ravnos just because Justin Achilli never liked the clan was petty and lame (and this is coming from someone who never really cared much for Clan Ravnos — I tried playing one once, and got bored rather quickly).

If you’re going to just love or hate something out of proportion with everything else, and can’t keep that separate from your game development efforts, don’t try to present it as an “equal” to the rest of the options. Don’t like dwarves? Make them monsters only, or just exclude them from the game. Think dragonborn are just the bee’s knees? Same thing — or just make a game where everybody plays a dragonborn, and stop pretending you’re trying to provide other options on equal footing.

Playing favorites like that can come across as amazingly passive-aggressive. Just don’t do it.


I might come up with some more peeves like these in the future. There are, of course, other things people often do when designing a roleplaying game that should, simply put, never be done (but often are). People seem to like lists of ten things, though.


  1. I’m glad to see you bringing up the dice mechanics again. That’s a really simple statistical thing to get right, so it’s such a shame that so many people get it wrong. Another thing I might add is to build up dice distributions with humps in the middle of them. For instance, the d20 mechanic adds too much variability to the game. Better to cluster the results a little bit more towards the median, by using 2d10 or the like.

    Comment by Cyde Weys — 28 July 2008 @ 08:00

  2. I have one! I have one! Well, more of a peeve than a blunder.

    I dislike RPGs that don’t include any mechanics for improving your character’s situation. Doesn’t have to be detailed. I always liked how… I think it was 2nd ed. D&D, fighters were capable of having followers, and eventually strongholds. Never followed it to the letter, but it gave a bit of a guideline.

    I think a lot of RPGs are missing a goldmine of opportunity when it comes to rewards for higher level characters. I see the ability to have an army march to war, or to influence the policy of a kingdom as the non-caster equivalents to those amazing high level spells.

    Birthright was neat in that way, if anyone’s ever played it. Your character really had a place in the world and wasn’t just an arbitrary hoover of gold.

    Comment by Anonymous — 29 July 2008 @ 03:01

  3. Cyde Weys:

    I thought about addressing probability curves more directly, but decided that wasn’t as big a problem for many game designs as the other things I listed — and since 10 is sorta the traditional list size, and prone to drawing in more readers, I decided I didn’t want to sacrifice any of the other points to shoehorn it in. In fact, in some games, having a flat “curve” for the common case isn’t necessarily a bad thing. It depends on the flavor of the game. Something like Toon wouldn’t suffer as much from that as a “gritty realism” sort of game where things like hydrostatic shock is factored into damage effects from gunshot wounds.


    When purely social things like that are incorporated into the game mechanics themselves, they need to be handled very carefully. I didn’t actually like the rules for followers in 2E at all — they were too arbitrary, cookie-cutter, and limiting. On the other hand, the very existence of those rules did provide me with some inspiration for how to handle that stuff in a more heuristic, off-the-cuff sort of way when I was running AD&D games back in the early ’90s.

    I think 7th Sea had better rules for that sort of thing. Fame systems tend to be better than minion systems as in the AD&D rules of the ’70s to the ’90s where it’s just assumed you magically acquire arbitrary numbers of devoted followers. Maybe that’s just me, though.

    Comment by apotheon — 29 July 2008 @ 11:12

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.

All original content Copyright Chad Perrin: Distributed under the terms of the Open Works License