March 2007

"The world cannot live at the level of its great men." - Sir James Frazer

March 13th, 2007

D&D: Calibrating Your Expectations

I’ve been working and playing with the new edition of Dungeons & Dragons longer than most. Ryan Dancey sent me a playtest copy of the new Player’s Handbook back in 1999, almost a full year before it was released at GenCon 2000. I had been an outspoken critic of AD&D for several years at that point and, more recently, been involved in a number of heated debates with Ryan over the OGL and D20 Trademark License.  

By the time I was done reviewing the playtest document and sending my comments back to Ryan, I had basically done a 180-degree turn-around on both. Wizards of the Coast had assembled three incredibly talented game designers – Jonathan Tweet, Monte Cook, and Skip Williams – to rework the system, and they had succeeded brilliantly. They stayed true to the roots of the game and captured the best parts of it, while shedding decades of detritus and poor design. There were still a few quibbles here and there, but they had taken advantage of the largest and most expensive design cycle for an RPG ever conceived and used it to deliver an incredibly robust, flexible, and powerful system.  

One of the most impressive things about 3rd Edition is the casual realism of the system. You can plug real world values into it, process them through the system, and get back a result with remarkable fidelity to what would happen in the real world.

Some people will consider this to be a remarkable claim. It doesn’t take much experience with the roleplaying hobby before you're familiar with dozens of vehement diatribes on the lack of realism in D&D and the resulting shortcomings in the system. Whole laundry lists of complaints (aimed at hit points, the encumbrance system, falling damage, or attacks of opportunity, for example) have been generated. In fact, such claims are so prolific that making the opposite claim (as I have done) is practically a heresy of sorts.

But, in my experience, these complaints largely originate either from people carrying over their criticisms of previous editions (where many of the criticisms were true) or from people failing to actually look at the facts and run the numbers.

So what I want to do, rather than just making my claim, is to take a look at a few rules, actually run the numbers, and demonstrate how effective D&D really is at modeling the real world.

Before we do that, though, I want to make one disclaimer perfectly clear: D&D is a game. Its systems are abstracted and streamlined in order to keep things simple and, more importantly, fun. So, yes, there are compromises. (You’ll see a graphic example of the types of compromises which are made when we talk about the Jump skill.) The game is not a physics text. Nor is it without flaw.

It’s just really, really good. And part of what makes it really, really good is the fact that it does this simulation casually. It doesn’t make you do the math. It’s worked the math into the system. All you’ve got to do is roll the dice and handle some basic arithmetic.

This essay should also be understood as something more than a defense of the game from illegitimate critique. That defense is, in fact, almost an unintentional consequence of what this essay is actually about: Providing a useful resource for those who want a deeper understanding of what the numbers really mean. If a character has a skill bonus of +15, how talented are they? If they have a Strength of 14 how strong are they? And so forth.



Breaking down a door in D&D requires a Strength check. So the first thing we need to understand is the distribution of ability scores in the general population.  

According to the DMG, defaults NPCs in D&D are built on one of two arrays: The elite array and the average array. The elite array is used for exceptional individuals. The average array is… well, average.  

Elite Array: 15, 14, 13, 12, 10, 8

Average Array: 13, 12, 11, 10, 9, 8

Using the point buy system, the elite array is built on 24 points. The average array is built on 15 points.

John Kim demonstrates that the elite array is, in fact, the statistically typical result of rolling 4d6-drop-the-lowest (the default character generation) if you round down fractional results. Similarly, although John doesn’t show it, the average array is the statistically typical results of a 3d6 roll.

The DMG doesn’t tell us how common an elite character will be, but they are supposed to be a “cut above the average”. I think it’s safe to say that they’re supposed to be rare. There are a number of approaches you can take to figuring out exactly how rare. For example, I did some back-of-the-envelope calculations using the DMG demographic information regarding how many characters are of a level high enough to suggest exceptional accomplishment in the real world (as we’ll discuss later). I also ran the numbers from MENSA (which accepts only those who intellectually qualify in the top 2% of the population) and then compensated to include the non-intellectual ability scores. And so forth. But, consistently, these inferences got me a figure right around 5%.

What does all this mean? It means that the vast majority of people you meet will be lucky to have a single +1 bonus in any of their ability scores. Most of them will, in fact, have straight 10’s and 11’s across the board.

So, with that in mind, let’s go back and talk about breaking down a door.

Breaking down a simple wooden door – like the doors you might find inside a typical house – is a DC 13 check. This means that the average person (with a +0 Strength modifier) will succeed at breaking open the door about 40% of the time. This means that one or two strong kicks from just about anybody will kick the door open.

This matches our real life experience: Interior doors just aren’t that sturdy.

Next, let’s take a look at something sturdier. For example, a well-made front door with its deadbolt secured. This would be a DC 18 check in D&D (for a “good wooden door”). This is a lot harder to bust open: The average person will only have a 10% of knocking it open on the first attempt. It’s going to typically take five or six really solid kicks for the average person to get through such a door.

Again: This matches our real life experience. Front doors are strong, but the fact that they’re not impervious to breaking-and-entering is evidenced by thousands of burglaries every year.

But once you take a thick wooden beam and use it to bar the door shut with solid iron construction (Break DC 25), it becomes impossible for the average person to simply throw their shoulder against the door and break it open.

And, again, this matches the real world. Breaking a six-inch thick beam would be nearly impossible for all but the strongest among us. Breaking such a beam without hitting it directly (instead diffusing our impact through a door) is essentially impossible.

Breaking down doors is a simplistic example, but it shows how much thought has gone into make the system consistent with the real world, even when it comes to the small details.

It’s also interesting to look at various magical effects in the system and seeing what they mean in real world terms. For example, a hold portal spell adds +5 to the Break DC of a door. So a hold portal spell is basically equivalent to adding a deadbolt to a door (make sense). An arcane lock spell, on the other hand, adds +10 to the Break DC of a door, so it’s basically the equivalent of barring the door shut.

So if the necromancer rushes through a portal and magically seals it behind him, what does it feel like when the party’s fighter throws himself against it? Now you know.

(And knowing is half the battle. Go Joe!)



One of the things which directly led to the creation of this essay was a mini-rant by Shamus Young on encumbrance in D&D:

 “Now, I like to travel light: I don’t check baggage unless I really need to. For my five-day trip I managed to get everything into a single reasonably-sized carry-on bag. It was just the bare minimum of items for five days: I wore a few clothing items twice to save space, and only carried a couple of books and a laptop for entertainment. Nevertheless, the strap of this bag bit into my shoulder as I walked, and the weight threatened to pull me off balance. A full-out run was nearly impossible, and a light jog caused the weight to bounce all over the place, slam me in the leg, and generally make the simple task of walking a bit more tricky than it normally is. It wasn’t just the weight that was a problem: the volume made the stuff difficult to manage as well.

Note that I was not wearing any metal armor. I wasn’t carrying enough food for five days in the wild. I didn’t have a sword, rope, grapple hook, spare dagger, or any other items D&D characters seem to keep handy. Try lugging five days of food and a few metal weapons a half-mile or so and you’ll quickly see that the D&D rules for carrying capacity are pure comedy.”

Pure comedy? Well, let’s run the numbers.

What was Shamus lugging around with his carry-on that day? Sounds like about three outfits of clothing. A laptop. A couple of books. A complete outfit of clothing in D&D is rated right around 4-5 pounds (unless you’re wearing royal regalia, which we’ll assume he wasn’t), so let’s say 13 pounds of clothing. A quick Google search turns up this page which suggests that his laptop probably weighed about 7 pounds. Poking around on Amazon suggests that books can reasonably weigh anywhere from about half a pound to about three pounds, depending on size and format. Let’s assume an average and call it 3 pounds for two books together. The luggage probably weighs another 7 pounds. Plus another 5 pounds for any miscellaneous stuff he didn’t mention explicitly (like toiletries).

Adding that up quickly we can see that Shamus was carrying somewhere in the ballpark of (13 + 7 + 3 + 7 + 5) 35 pounds.

Now, I’m guessing that Shamus isn’t a professional weightlifter. And, by the same token, he probably isn’t a 90-pound weakling, either. My guess is that he falls solidly into the D&D average: Strength 10.

For a D&D character with a Strength of 10, the 35 pounds or so that Shamus was carrying constitutes a medium load. A medium load hits you with a check penalty on physical actions, reduces your speed, and stops you from running full out.

Which, when we compare this to Shamus’ experience, sounds like a pretty accurate mechanical representation: “…the weight threatened to pull me off balance. A full-out run was nearly impossible, and a light jog caused the weight to bounce all over the place, slam me in the leg, and generally make the simple task of walking a bit more tricky than it normally is.”

And he was also carrying the weight in the worst way possible – as an off-center load. D&D doesn’t try to model how you’re carrying a load, but it’s reasonable to assume that the rules are designed with the assumption that characters are carrying their gear in a way which minimizes the inconvenience.

(To understand the difference, grab a 25-lb. weight and try jogging around the block with it held in your hand. Then load up a backpack with 25 pounds of stuff, pack it tight, and strap it across your back. Do the same run. It’ll be a lot easier. There’s a reason backpacks were invented.)

This is one of the way in which D&D chooses to make a playable compromise rather than trying to slavishly model every aspect of reality. Sure, you could try to create a system which attempted to model how awkward it was to carry a particular item in a particular way. Maybe items carried in the hands carry a x2 encumbrance penalty. Maybe carrying more than one or two spears is more difficult than carrying an equal weight of a smaller and more compact item.

But such a system would be a nightmare. It’s already enough of an overhead headache to try to keep an accurate tally of just the weight being carried. Most groups will only do the occasional “update” of weight carried, rather than trying to account for every time picked up or dropped as it occurs. And this is fine.

But if you tried adding a whole new layer complexity in which players had to figure out exactly how many standard rations will fit into a small, medium, or large backpack or how carrying them in a bag on your hip will affect their encumbrance you’d almost certainly create a system which would never be used (because no one would want to use it). It wouldn’t increase utility and accuracy, it would reduce it.

But with that acceptable layer of abstraction in place, how else does the system demonstrate its accuracy?

Well, let’s look at the military. How much does the equipment they carry into combat typically carry?

Historically, militaries have tended to have kit weights for their infantry soldiers in the 40-60 pound range. We’ve seen that, for average people, this would constitute a medium load: Not so heavy as to be a serious impediment, but definitely inhibiting compared to wearing winter clothes.

Of course, most soldiers aren’t of average strength. Boot camp is specifically designed to build strength (among other things). It wouldn’t be unreasonable to expect an average infantryman’s highest ability score (13) to be in Strength. And with a Strength of 13, anything up to 50 pounds is considered a light load (with no penalties to action).

And take a look at this recent article: The amount of gear carried by an American infantryman is creeping up into the neighborhood of 70 pounds or more. That’s a heavy load for a person with average strength (10), and even our physically fit soldiers with above average strength (13) who were previously wearing light loads are now being weighed down with a medium load.

Our soldiers aren’t too happy with all that extra weight. And D&D shows us why.

(On a slight tangent, another common complaint about D&D’s encumbrance figures go the other way: They claim that things like longswords should be heavier and that plate armor should make it impossible to do anything physical. Most of these beliefs are either built on shoddy replicas, urban legends, or simple misunderstanding.

To pick on Shamus Young again, he also jumped on this bandwagon: “The system is even more messed up than it seems. A quick glance at the item weights in the player handbook will reveal gems like the following: A longsword weighs 4lbs. Even using lightweight modern metal alloys, I think you’d have a very, very hard time getting an adult-sized longsword that weighs only 4lbs. Even if you did somehow have a sword that light, it would feel like a toy in your hand.”

But even casual research quickly reveals that 4 lbs. is almost exactly what historical long swords weighed. The essay “What Did Historical Swords Weigh?” by J. Clements is an excellent resource for this. People didn’t make slow, heavy weapons and awkward armor because their lives depended on not making weapons and armor like that.)



There’s a common fallacy when it comes to D&D, and it goes something like: Einstein was a 20th level physicist. So, in D&D, Einstein – that little old man – has something like a bajillion hit points and you’d need to stab him dozens of times if you wanted to kill him. That’s ridiculous!

The problem with this argument is that Einstein wasn’t a 20th level physicist. A 20th level physicist is one step removed from being the God of Physicists. Einstein was probably something more like a 4th or 5th level expert.

This can be a little bit difficult for some people to accept, so let’s run the math. At 5th level an exceptional specialist like Einstein will have:

  • +8 skill ranks

  • +4 ability score bonus

  • +3 Skill Focus

In the case of our 5th level Einstein, that gives him a +15 bonus to Knowledge (physics) checks. He can casually answer physics-related questions (by taking 10) with a DC of 25. Such questions, according to the PHB description of the Knowledge skill, are among the hardest physics questions known to man. He’ll know the answers to the very hardest questions (DC 30) about 75% of the time.

And when he’s doing research he’ll be able to add the benefits of being able to reference scientific journals (+2 circumstance bonus), gain insight from fellow colleagues (+2 bonus from aid another), use top-of-the-line equipment (+2 circumstance bonus), and similar resources to gain understanding of a problem so intractable that no one has ever understood it before (DC 40+).

(This 5th level Einstein can also be modeled with as few as 5 hit points – 1 per hit die. Even if he rolled an average number of hit points on each hit die (3 each), as an old man his average Constitution of 10 will have dropped two points. With the resulting Constitution penalty, he still only has 10 hit points. This is the other reason why the hit point argument holds no water.)

You’ll see this same fallacy trotted out whenever someone insists that the local blacksmith “must” be at least 10th level in order to be competent in their profession. In reality, the typical village blacksmith is probably only a 1st level character. At 1st level the average blacksmith’s Craft (blacksmithing) skill looks like this:

  • +4 skill ranks

  • +1 Intelligence bonus

  • +3 Skill Focus

  • +2 from an assistant or apprentice helping them

That’s a +10 bonus on their checks. This bonus allows them to take 10 and craft masterwork-quality items. By 3rd level an experienced blacksmith can do that without the help of an assistant.

Even less capable 1st level blacksmiths (without an assistant or the Skill Focus feat) still have a +5 bonus to their skill. This lets them take 10 and craft high-quality items (the only things they can’t handle are exotic weapons and complex items).

And what does an exceptional 5th level blacksmith look like?

  • +8 skill ranks

  • +4 Intelligence bonus

  • +3 Skill Focus

  • +2 masterwork tools

  • +2 from an assistant or apprentice helping them

That’s a +19 bonus to the check. When taking 10 he can essentially triple the speed with which he can make common items like iron pots and horseshoes. He can easily create work far surpassing masterwork quality and can every so often (when he rolls a natural 20) create a work of essentially legendary quality (DC 39).

What does all this mean?

It means that the most extraordinary blacksmiths in the real world top out at 5th level. Amakuni, the legendary Japanese swordsmith who created the folded-steel technique? 5th level.

Arachne, the legendary weaver who challenged Athena herself to a duel (and lost)? She might be 10th level.

Does this mean you should never throw a 10th level blacksmith into your campaign? Nope. D&D is all about mythic fantasy, after all. But when you do decide to throw a 10th level blacksmith into the mix, consider the fact that this guy will be amazing. He will be producing things that no blacksmith in the real world has ever dreamed of making. And a 20th level blacksmith is one step removed from Hephaestus himself.

(Coincidentally: Why do dwarves have such a reputation for mastery of the forge? They have a +2 racial bonus to Craft checks. That means that, unlike human blacksmiths, the average dwarf doesn’t need to be 3rd level in order to single-handedly create masterwork items – they can do it at 1st level. Basically, due to their natural aptitude, dwarves are master craftsmen before they ever leave their apprenticeships.)



How well do these numbers hold up when compared to other skills? Well, let’s take a look at the Jump skill.

Based on our analysis of the Knowledge and Craft skills, we know that a 1st level character has professional competency in their chosen field. We also know that a 5th level character represents the most legendary levels of skill – the type of people who come along once in a generation.

So, when it comes to jumping, a 1st level character probably represents a typical college athlete. A 5th level character, on the other hand, represents the small handful of jumpers who challenge and break the world records. It would make sense then, that Olympians would probably fall in the 3rd to 4th level range (better than your run-of-the-mill specialists, but not quite at the level of once-in-a-generation).

Let’s take a look at a 4th level Olympian jumper:

  • +7 skill ranks

  • +3 Strength bonus

  • +3 Skill Focus

That’s a +13 bonus.

Now, back in the original 3rd Edition (3.0), the result of a running long jump check was:

5 ft. + 1 ft. per 1 point above 10

This can be more easily paraphrased this way: The distance of a long jump is equal to your check result minus 5 feet.

Our Olympian’s jumps will range from 9 feet (stumbling all the way on a roll of natural 1) to 28 feet. But a typical Olympic event involves three jumps in which the best distance is recorded. That means that roughly 80% of the time, our long jumper will be jumping between 20 feet and 28 feet in competition.

Looking at the 2004 Olympics, the top forty men’s long-jump results during the qualification round range from 24 feet to 27.25 feet. Those types of results will be posted approximately 60% of the time by our Olympic long-jumper.

With out 5th level jumper we can bump the ability bonus up to +4, add a +2 synergy bonus from Tumble, or a +4 bonus from the Run feat. The result would be a the ability to achieve jumps in the 29-35 foot range. The world record is currently set at 29.35 feet.

So, once again we’re finding that 5th level is right at the dividing line between legendary real world performances and the impossible realms of the superhuman.

And you’ll find similar fidelity with the high jump rules. (In fact, the 3.0 high jump rules are even more accurate than the long jump rules.)

The jumping rules, however, are perhaps the most visible victim of gameplay compromises in D&D. When the system was revised for D20 Modern, the distance of a long jump was revised to simply equal the DC of the check. This change was later picked up in the 3.5 revision of the D&D rules.

This rule is a lot simpler to remember, but it makes jumping significantly easier than any other skill (compared to real world performance). Under the new rules, 1st level characters can now trivially perform Olympic-level jumps and our typical Olympians will be routinely smashing the world record. (The high jump rules, on the other hand, remain fairly accurate.)



So what have we learned so far? Almost everyone you have ever met is a 1st level character. The few exceptional people you’ve met are probably 2nd or 3rd level – they’re canny and experienced and can accomplish things that others find difficult or impossible.

If you know someone who’s 4th level, then you’re privileged to know one of the most talented people around: They’re a professional sports player. Or a brain surgeon. Or a rocket scientist.

If you know someone who’s 5th level, then you have the honor of knowing someone that will probably be written about in history books. Walter Payton. Michael Jordan. Albert Einstein. Isaac Newton. Miyamoto Musashi. William Shakespeare.

So when your D&D character hits 6th level, it means they’re literally superhuman: They are capable of achieving things that no human being has ever been capable of achieving. They have transcended the mortal plane and become a mythic hero.

This requires a shift of perception for some people, but I’ve found it valuable, when crafting my own campaigns, to keep it in mind: Even though the PCs inhabit a world where there are many higher level characters, once they’ve gotten past 5th level or so, they are truly special individuals. They will be noticed. Their accomplishments will be (and should be) things which would enshrine them in the legends of our world. It’s OK for them to excel.

To help put this in further perspective, let me pop another popular canard:

People love to stat up their favorite heroes from fantasy literature as 20th level juggernauts. Fafhrd? 20th level. Elric? 20th level. Conan? 20th level. Aragorn? 20th level Luke Skywalker? 20th level.

I mean, they must be 20th level, right? They’re the biggest, bestest heroes ever! They’re the greatest warriors in a generation! Some of them are reputedly the greatest swordsmen who ever lived in any universe EVAH!

But when you stop and analyze what these characters are actually described as achieving, it’s rare to find anything which actually requires a 20th level build.

Take Aragorn, for example. He’s clearly described as one of the best warriors in Middle Earth. But what do we actually see him do? Let’s take THE FELLOWSHIP OF THE RINGS as an example:

  •  He leads the hobbits through the wilderness with great skill. (The highest Survival DC in the core rules is DC 15. A 1st level character can master the skill for non-tracking purposes. Aragorn, as a master tracker, would need to be 5th level, have at least one level of ranger, and have spent one of his feats on Skill Focus (Survival) to achieve all of this.)

  •  He drives off the ringwraiths at Weathertop. (It’s difficult to conclude anything from this because it’s one of the more problematic passages in the book when subjected to analysis. If the ringwraiths are truly impervious to harm from any mortal man, why are they scared off by a guy waving two “flaming brands of wood”? Are they vulnerable to fire in a way that they’re not vulnerable to mortal weapons? The point is, the true strength of the ringwraiths is obscure, so it’s impossible to know how tough Aragorn would need to be in order to accomplish this.)

  • Aragorn treats Frodo’s wound, unsuccessfully. (The highest Heal DC is 15. As with Survival, Aragorn could have mastered this skill at 1st level.

  •  In Moria (fighting orcs): “Legolas shot two through the throat. Gimli hewed the legs from under another that had sprung up on Balin's tomb. Boromir and Aragorn slew many. When thirteen had fallen the rest fled shrieking, leaving the defenders unharmed, except for Sam who had a scratch along the scalp. A quick duck had saved him; and he had felled his orc: a sturdy thrust with his Barrow-blade. A fire was smouldering in his brown eyes that would have made Ted Sandyman step backwards, if he had seen it. (Aragorn slays no more than six or seven CR 1/2 orcs in this encounter. A trivial accomplishment for a 5th level character.)

Even if you follow Aragorn all the way through The Two Towers and The Return of the King, you’ll find that this is fairly representative of what he accomplishes. The only other notable ping on the radar is his ability to use athelas, and even if we don’t assume that’s merely an example of him knowing athelas’ properties (with a Knowledge (nature) check), it’s still just one ability.

So what can we conclude form this? Aragorn is about 5th level.

And since Aragorn is one of the most remarkable individuals in all of Middle Earth, this would imply that Middle Earth is a place largely like our own world: People who achieve 5th level are uniquely gifted and come along but once in a generation.

Does that seem like a proper description of Middle Earth? It does. Tolkien was crafting a false mythology – a forgotten epoch of our own world. Thus the people in it are much like the people we know, although they live in a world of heroes and magic.

(For the record, I’d probably model Aragorn as a Rgr1/Ftr1/Pal3. That gives you the tracking, lay on hands, and quantifies his ineffable ability to instill courage in those around him. Use one of the feat selections for Skill Focus (Survival) and you’re still left with another three feat selections for the final tweaking.)

Why do people make the mistake of modeling characters like Aragorn as 20th level characters? I think it arises from several factors.

First, there is the assumption that the fictional world of the novel is a typical D&D world. If someone is described as “the best in the world”, therefore, they must be 20th level. Otherwise there would be people better than them and the description wouldn’t be accurate, right? But the reality is that, in Middle Earth, there aren’t any 20th level characters. (At least, none of mortal stature.) Even the most exceptional of the immortal elves are most likely no more than 8th level or so (and that’s pushing it). Gandalf is a demigod cloaked in mortal form and I’d have difficulty statting him up as even a 10th level character.

Second, people can be thrown off by some contortion required by D&D in order to get a very specific set of abilities. A character is described as having one very specific ability that only a 5th level druid can have and is simultaneously described as having another ability that only a 12th level ranger can have, so clearly they must be a 17th level character, right?

Well, no. Authors don’t design their characters around the class progressions of the core D&D classes. Take, for example, a character who can assume an ethereal state without casting a spell. The only way to do that in D&D, using only the core classes, is to be a 19th level monk. But if that’s the only special ability the character in question has, it would be completely nonsensical to model them as a 19th level monk – they don’t have any of the plethora of other abilities such a monk possesses. What you’re looking at is a character with a unique class progression or possibly a prestige class. Or maybe a racial ability.

Finally, you’ll get into an arms race of expectations which just reinforces the whole thing: Aragorn must be 20th level. So the orcs who posed such a challenge to him must be 15th level or higher. And since those were elite 15th level orcs, Aragorn must have been 20th level in order to face them.



The problem with having false expectations about what “Strength 20” or “15th level” really means is that it creates a dissonance between what the rules allow characters to do and what you think characters should be able to do. For example, if you think that Conan should be modeled as a 25th level character, then you’re going to be constantly frustrated when the system treats him as a demigod and allows him to do all sorts of insanely powerful things that the literary Conan was never capable of. From there it’s a pretty short step to making pronouncements like “D&D can’t do Conan” (or Lankhmar or Elric or whatever).

The other problem is the expectation it brings to your campaigns. If you believe that epic adventures are only possible for characters who are 20th level, then your players are going to have a long, hard slog through lower levels of utter tedium before they can get to the “good stuff” that resembles the fantasy stories they love.

I’m seen people spend countless hours trying to tweak various rules so that, for example, 20th level characters (who are basically mythological demigods) can’t fall off the Cliffs of Insanity and survive because “no one could survive a fall like that”. Well, that’s true. No one in the real world can survive a fall like that. But that’s because no one in the real world is a demigod. You might be missing the forest for the trees here.

The fact that D&D can handle a range of powers from the subhuman all the way up through the superhuman and into the demigod-like is actually one of the system’s strengths. People will rightfully point that, beyond 20th level, the system begins to break down. But compared to virtually every other RPG ever designed, D&D’s performance across that wide range of powers is still amazing. Nothing else can really compare.

(For example, HERO was originally designed to model superheroes, and it does that exceptionally. It’s also very good at handling cinematic action heroes. But when you ask it to do normal human beings, it starts developing some serious clunkiness. On the other end of the spectrum, GURPS is great at handling human beings near the normal end of the spectrum, but gets progressively more awkward the more powerful the characters become.)

But what frustrates some people is that D&D assumes that you’re going to move from one level of power to an extremely different level of power. So they spend a lot of time tweaking the system and trying to get it to perform at a more uniform level from 1st to 20th level.

I think this is the hard way of doing it. Instead of fighting the system, I’d rather try to work with it: Target the precise range of levels which form the “sweet spot” for whatever campaign concept I’m working on, and then tinker with the character creation and advancement rules to keep the campaign focused in that sweet spot. Those changes can be as simple as “XP awards will be 1/10th the normal size and everyone should create a 5th level character”, but more complicated variants are more than possible.

The point is that you find that “sweet spot” and then you tinker with one aspect of the system, rather than trying to redo the whole thing.

And the first step in finding that “sweet spot” is to recognize what the numbers really mean. Which neatly takes us back to the beginning premise of the essay. (I wouldn’t suggest going back and re-reading it like a literary ouroboros, though. It’s long enough as it is.)


To Eve Forward, for giving me the impetus to collect a variety of my scattered thoughts into this single essay.

To Rupert Boleyn, Malachias Invictus, Doug Lampert, and Sea Wasp (Ryk E. Spoor), who have all contributed to the development of these thoughts in online discussions over the years.

  | | Link

March 15th, 2007

Shamus Young's blog is provoking me to all kinds of random thoughts. Then I had similar thoughts over at Giant in the Playground. So I thought I'd mix the two together and share 'em with you:

Random DM Tips: Running Combat

1. ROLL INITIATIVE LAST: Have your players roll their initiatives at the end of combat. Use this initiative for the next combat. (Initiative modifiers essentially never change, so it doesn't really matter when you roll the check.) When it looks like the PCs are about to encounter something, roll for its initiative and slot it into the order. If they don't encounter it for some reason, no big deal. If they don't actually encounter it, no big deal.

Using this method, by the time combat starts, initiative is already completely resolved. As a result, there's no delay while you ask for initiative, the dice are rolled, your players tell you their results, and then you sort the results into order. This allows you to start combat off with a bang and keep the ball rolling with that same high intensity. It means that when the players are ambushed, you can maintain that adrenaline rush of surprise instead of immediately undermining it with the mundane task of collecting initiative.

This method also means that initiative results are generally being collected at a time when other bookkeeping chores are being done anyway: After the heat of battle, wounds are being healed; corpses are being looted; equipment lists are being updated; and options are being discussed.

2. INITIATIVE CARDS: Write the names of each PC on an index card. Do the same for each group of NPCs. When initiative is rolled, jot down the result on the PC's or NPC's card and then sort them into order. During combat you can just flip from one card to the next. If someone readies or delays an action, turn their card sideways so it sticks up. When they take that action, pull the card out and put it back in the rotation.

3. INITIAL PRESENTATION: The other thing which can deflate the initial tension of a battle is the set-up for the battle. Sketching the scene on the battlemap and positioning the miniatures takes time. Learn to work your set-up into the presentation of the battle:

"Suddenly you hear the mournful baying of wolves and the battle horns of orcs!" <as you say this, start sketching the scene on the battlemap> "Aragorn, what are you doing?" <finish sketching the map as you go around the table and find out what the PCs' first action is going to be> "Over the top of the ridge a dozen warg riders suddenly surge into view!" <begin placing the warg rider miniatures as you say this> "Aragorn, you're first: What are you doing?" <finish placing the warg riders as Aragorn declares his action, then more smoothly into resolving his action>

The point is that you're putting out the enemy figurines (even if they're not dramatically painted miniatures) can be a dramatic moment in itself, if you make it the moment of revelation. On more than one occasion I've had players murmur "oh shit" or "how many of them ARE there?" as the number of miniatures comes out onto the table.

It's also not a bad idea to get an erasable battlemap and use it more often than not. Even if the players aren't going to be fighting in a particular room, the visual reference isn't going to kill anybody. And it helps to have a firm visual reference of positioning even when you're not in combat. (When a trap goes off, for example.)

And the pay-off when you can reach down into a drawer; declare, "Suddenly, out of the black abyss, a black dragon emerges!"; pull the miniature out and, with a dramatic swoop, place it on the battlemap is totally worth it.

Also: When you're drawing a battlemap, don't sweat the details. If you're off by 5 feet, your players are never going to know.

Of course, if you don't use a battlemap (which is more than possible in D&D, despite what some would tell you), none of this makes the least difference to you. Get initiative preemptively out of the way and you're good to go.

4. GLASS BEADS: If you're fortunate enough to be able to always have enough miniatures at hand to cover every single monster in the scenario, then you're a lucky bastard and the rest of us resent you terribly.

While we're resenting you, however, we're going to head down to our Friendly Neighborhood Gaming Store and buy some of those glass beads that people use as counters for CCGs and the like. For $10-15 you can pick up more than a hundred beads in fix or six different colors. These are great: Not only are they cheap, but they can also be used to stand-in for miscellaneous battlefield scenery and the like. (Use the green ones for trees.)

Some people recommend dice for the same purpose. Dice have the advantage that you can use different numbers on each die to represent different creatures within the group. But the problem is that there are dice flying all over the table during combat as people make their rolls. The counters are easily distinguishable and, if they (or the table) are bumped, they aren't designed to roll away.

I’ve been known to start epic combats against a horde of mooks by literally pouring a handful of counters out on the edge of the map (”Suddenly, on the far side of the cavern, a horde of goblins — hooting and gibbering in their barbaric tongue — leaps up from their hiding places!”) and then rapidly shifting them to make sure they’re all in different spaces.

5. CHEAT SHEETS: I use three useful cheat sheets during combat:

  • A cheat sheet with key PC stats, particularly armor class. One of the biggest things that used to slow down my combats was having to ask, "What's your AC?" and then waiting for the player to look at his sheet. If your party has a lot of abilities which cause their ACs to move around (like a rogue using Combat Expertise a lot), get a whiteboard and have the players update their AC on the whiteboard whenever it changes.
  • A separate cheat sheet with the monster stats on them. I generally find that having all the stats for a monster on one page, in a large font (12 or 14 pt.), makes it a lot easier to find the information that I want quickly.
  • A cheat sheet briefly summarizing the mechanics for all the combat maneuvers in bullet-point style. My version of this is about two and a half pages long and it makes a huge difference. Sure, there's still the occasional need for a manual look-up when I need to know a particular detail, but generally glancing at the cheat sheet is enough to jog my memory.

Basically, you want to get rid of as many interactions in which you're just asking people for a number or waiting for someone to roll the dice as possible. You also want to get rid of any dead space created by simply flipping through pages trying to find the information that you need. "What's your AC?" is a big one. "I rolled X on my saving throw, did I save?" is another. Which brings us to:

6. BE OPEN WITH YOUR TARGET NUMBERS: The uncertainty of rolling against a DC you don't know can certainly increase tension for the PCs, but it gobbles up time. A good compromise here is to let the first few roles be made in the dark, but then be open about what the number you're looking for is. And only do this when you can use it to good narrative effect.

For example, it can be effective to say, "Your shot is true, but you're shocked to see it shatter on the creature's skin as if you had struck a wall of stone." That serves as a revelation of the creature's nature (high natural armor bonus). But after a couple of rounds you can simply announce the AC and move on (they're going to figure it out eventually in any case). And if they're just fighting bog-standard goblins in scalemail, there's really no reason not to let them know what the AC is from the get-go.

7. MULTITASK: Waiting for a player to make his attack roll? Ask the next player what they're planning to do. Did one of your monsters just drop a fireball on half the party? Ask them for their saving throws and then -- while they're rolling -- move on to the next monster and resolve its action. (In the case of a fireball you can also announce what the damage is, what the DC of the save is, and then let the players proceed with rolling their saves while you move on to the next set of tasks.)

You can also encourage players to multitask in small ways. For example, one common trick is to roll your attacks and your damage at the same time. If the attack roll results in a miss, you can just ignore the damage roll. But if the attack roll hits, there's no delay while damage dice are picked up and rolled.

8. ROLL A LOT OF DICE: Invest in a lot of d20s. Then, when the mooks are attacking en masse, roll all their attacks at once. I use a simple left-to-right metric: The dice which ends up furthest to the left after rolling is assigned to the monster farthest to the right in my field of vision. In a lot of cases, there'll be multiple mooks on one target -- as long as they've all got the same stats, I can roll all of their attacks at once without really worrying about exactly which die goes to which monster.

Some people suggest using color-coded dice instead: For example, the blue die might be monster #1, the yellow die might by monster #2, and so forth. I've found that this can work fairly well at low levels, but once iterative attacks kick in I find the color-coding is more useful to distinguish between iterative attacks. (All the black dice are attacks at the highest BAB, all the blue dice are attacks at the second highest BAB, and so forth.)

9. BE DESCRIPTIVE: Don't let a quest for speedy combat resolution reduce combat to nothing more than completely abstract and rapid-fire dice rolls. It doesn't matter how fast you roll the dice: If you've taken all the flavor and excitement out of the experience, even the shortest combats will still feel tedious.

So, describe the rapid flurry of blows. Describe the blood flowing. Describe the awesome wuxia swordplay. Describe the grunting and the sweating. Describe the amazing near misses. Describe the steel beating on steel. Be imaginative and remember that not every descriptive element needs a mechanical representation: A person can be staggered by a blow without necessarily being dazed. A person can be knocked back a step without actually repositioning themselves on the combat grid.

Here's an exercise I do every once in awhile: Take a really great fight film -- like The Matrix or 300 or Crouching Tiger, Hidden Dragon -- and narrate the action as it happens on screen, as if you were describing it to your gaming group. It sounds corny, but it builds your repertoire and helps loosen up your description instincts.

One more tip: Don't feel like you need to narrate every roll of the dice. Sometimes its okay to just let the dice speak for themselves. Sometimes its better to resolve a couple or three actions and then weave them into a post facto description. The orc misses his attack roll, then the player makes his, but the damage roll is low. Thus: "The orc throws his entire body into swinging his axe. You easily sidestep the obvious and awkward blow and counter-attack with a dextrous thrust. But the orc manages to twist his body lithely to one side so that your sword only scratches his arm. The orc turns that twist into a save counterblow which <roll the orc's next attack> swings wildly over your head as you duck under it at the last possible second."

10. DESIGN INTERESTING SCENARIOS: Try to shake things up every three or four combats by adding some sort of "spice" to the encounter. This might be an unusual tactic by an opponent (for example, pulling some of the PCs onto the ethereal plane in order to split their forces). This might be unusual terrain (fighting on a cliff-face or a rain-slicked street). It might be a location with interesting props (a bazaar where rickety tables are destroyed or a building that catches on fire).

Just a little something extra that breaks up the monotony, possibly disrupts the normal PC tactics, and makes the combat a little more memorable and unique.

Not every interesting scenario will require a new mechanic. But if it does, try to keep it simple. A rain-slicked street can be modeled by requiring a Balance check (DC 10) whenever the PCs run or charge. A bazaar full of rickety tables might require a Reflex save (DC 12) to avoid hitting a table (hardness 5, 6 hp). Don't ask your players to learn complex new mechanics just for the sake of the encounter -- you'll spend more time explaining the mechanics than you will enjoying them.

11. BE JOHN WOO, JACK KIRBY, AND STEVE DITKO: Steve Ditko co-created Spider-Man. Jack Kirby basically co-created every other classic Marvel superhero of note. And John Woo essentially created the modern gun-fu genre. What these creators have in common is their ability to create dynamic and exciting action sequences without putting their stories on hold. With most comic books and action movies, you can literally remove the fight scenes and insert a placard which reads "They Fight" and lose nothing except some cool visuals: All of the character development and important plot points happen when people aren't throwing punches. 

But you can't do that with a story by Woo, Kirby, or Ditko: The fights scenes are an integral part of the story. They are a crucible in which characters are revealed and developed. The plot continues to flow and evolve. The fact that they're incredibly awesome fights is just sauce for the goose.

Try to find that same quality in your own fight scenes. Ask yourself what purpose a fight can serve other than just straight-up violence.

Even a properly designed random encounter can serve as a source of information or character development: For example, in my Rappan Athuk campaign one of the frequent random encounters in the dungeon was with gelatinous cubes. The PCs had been wondering what was responsible for cleaning out the corpses they left behind them, and when they encountered the gelatinous cubes a light bulb went on over their heads. It was a minor thing, but those gelatinous cubes still stick out in the memory. 

12. TRAINING SCENARIOS: Quite a few people have annoying sections of the combat rules that they're constantly having to look up. Tripping, for example, is usually high on the list. Grappling and mounted combat are other common examples.

Here's a suggestion: Design a few combat scenarios in which a particular set of rules will be used over and over and over again.

For example, maybe the villains of your next adventure can have their combat tactics developed around tandem tripping combos (one villain trips the target and then the others beat up on the prone character). Familiarize yourself with the rules for this one specific combat maneuver, and then use the game session as a way to cram both your players and yourself on them. By the time you've resolved the twentieth trip attack in less than an hour, you'll have those rules down pat. (Which means that it will be a lot easier to throw a few random trips into your normal fights, which will help you spice them up with a little variety.)

Ultimately, we learn best by doing.  This is why we know how to resolve a sword swing so well: We do it several dozen times a night. 

| | Link

March 18th, 2007

Sarah and I celebrated the one-year anniversary of our relationship today. We treated ourselves to a weekend up at a bed-and-breakfast in Duluth, MN (where Sarah went to college) to celebrate.


Look at that gorgeous girl. She beams with beauty! I'm such a lucky guy.

  | | Link

March 20th, 2007



I have a simple rule of thumb: If you're designing a mystery for your PCs to solve, you should include at least three clues for every conclusion you want them to reach. More often than not they'll miss the first clue and misinterpret the second, but the third will do the trick. (And sometimes they'll spontaneously jump to a conclusion without even being given a clue, which is always a pleasant surprise.) If you further design the adventure so that they can complete it even if they don't reach every single conclusion that you want them to, then your adventure is probably robust enough to withstand actual play design.

This design methodology not only sidesteps the common problem (where the PCs miss or misinterpret some vital clue), but it also leads to a more robust scenario: All those clues give you a much firmer and deeper understanding of what's happening, making it much easier to improvise on your feet if the PCs suddenly go haring off in a random direction.

The classic adventure Death in Freeport doesn't quite honor this design principle: When running the adventure out of the box there are a couple of choke points where PCs might find themselves facing a brick wall if they make they turn the wrong way or make the wrong assumption. But way back in 2000, when I first ran this adventure, I buffed it up with a few additional clues and alternate investigation methods. And I not only ran it with great success in 2000, but I ran it again in 2002 to launch a fairly successful mini-campaign, and then I ran it again in 2003 as a one-shot. It was pretty much foolproof.

Then, in 2004, I discovered that I had never known true foolishness.

This is the story of the worst experience I have ever had as a DM. I had gathered together a gaming group with the intention of playtesting a mega-adventure that, sadly, was never published. In order to lead the group into this adventure (which started at 6th level), I decided to go with some tried-and-true material: The original Freeport trilogy that I had run to such great success before. After five or six sessions of material I was completely confident about, I would have a firm baseline for judging the success of the original material in the mega-adventure.

Instead, the campaign lasted only three sessions and never got beyond Death in Freeport.

I say it lasted "only" three sessions, but the reality is that these sessions were grueling and painful affairs. It was not just that the party ineptly blew off, ignored, or blatantly misinterpreted even the simplest of clues -- it was the inept bungling of their every attempt to carry through on a good intention and the utter incompetence of their exploits. A quickie adventure that generally takes about four hours to complete dragged out for more than twenty hours of gameplay, by the end of which I, as the DM, was struggling to find any way of bringing the scenario to a close.

Here are a few of the more memorable and (in retrospect and from a safe distance) hilarious exploits:

1. They were given a "To Do" list that the priest had apparently made the day before he disappeared. On the list there was a specific person mentioned. They tracked this person down and discovered he was a ship's captain. They proceeded to concoct an elaborate scheme in which they would pretend to have a cargo they needed to ship and then offer it to the captain's closest competitor! The competitor accepted the cargo. When this failed to elicit a response, they sat down with the captain and said, "Hey! We just gave your competitor some business! Whaddya think of that?"

The captain said, "He's a liar and a cheat and a swindle, but who you choose to do business with is your own affair."

They concluded from this that the captain had never heard of the priest they were looking for. (You'll notice that they never actually asked the captain whether he knew anything about the priest. They never even mentioned the priest.) Then they spent about an hour of game time acquiring the cargo they had pretended to have so that they could actually give it to the competitor and pay him to ship it. (Why? I never found out.)

2. Assassins were sent to kill them. They killed the assassins and discovered a note on one of their bodies describing where and when they were to meet the person that had hired them. The party went to this location several hours before the meeting was scheduled to happen and discovered it was a tavern. They stayed there for about half an hour and then left... still several hours before the meeting was scheduled. The next morning they went back, broke into the tavern, and tried to kill the bartender.

3. After missing or blowing off several other clues, one of them finally managed to get himself killed. So a replacement PC was brought in, and I seized the opportunity to give this new PC a "clue" which basically consisted of him saying: "Hey, I know the guy who's behind this. We should follow him and find out where their hideout is."

So they follow this guy for a couple of minutes... and then one of them steps out of hiding and stabs him to death.

4. So the bad guys kidnap another priest, and this time I connive to have one of the PCs see it happen. (I'm getting desperate at this point.) The PC follows the kidnappers for several blocks and then... shoots at them with his crossbow. He's outnumbered 6-to-1 and, after getting hit once, announces that he "only had 1 hp left" and is now dead.

This same group also had another memorable moment: At one point the party's wizard was hit by a silver dart which had a note wrapped around it, "You die at midnight." The party concluded, rightly, that this was a threat! So they head back to the inn where they were staying and resolve to all stay awake in the common room so that they can't be surprised...

... all of them except the wizard, that is, who instead specifically gets his familiar drunk enough that it's unconscious and then goes upstairs and falls asleep himself.

Oddly enough, when the rest of the group came in the next morning, they found the wizard dead with a knife sticking out of his throat.

And that's basically what happened to the campaign, too. It was a mercy killing, really. The PC who had gotten himself killed by launching a "cunning" ambush with only 1 hp left to his name was revived inside the bad guy's secret hideout while he was being prepared for a ritual sacrifice. With a little prompting he managed to escape, putting him in the perfect position to grab the rest of the party and lead them back to the secret hideout! This would start the straight-out dungeon crawl portion of the adventure, which would presumably negate much of their bumbling ineptitude!

... only that's not what he did. Instead he fetched the city guard, who moved in and secured the hideout. This was almost certainly the most competent thing any of them had done in the course of the entire adventure, but it also assured that the PCs never actually managed to accomplish anything at all.

It may have ended with a whimper instead of a bang. But at least it ended.

  | | Link

March 28th, 2007


One of the nits that seems to get perennially picked in D&D is the Tumble skill. Specifically, the uses of the skill which allow a character to avoid attacks of opportunity: The DC 15 check to avoid attacks for moving through threatened areas and the DC 25 check to tumble right through an opponent's space.

There are generally understood to be two shortcomings to these rules:

1. The game already makes it relatively difficult to control territory. For example, there is no effective way for a single person to guard a 10' wide hallway -- no matter how skilled they are and unskilled their opponents are. The Tumble skill exacerbates this because now you can't even control the space you're standing in: Say the PCs want to prevent someone from reaching the Lever of Doom at the end of a hallway -- it doesn't matter how many demigods you cram into the hallway, a 1st level tumbler can still move past them like water through a sieve.

2. The idea that some skill in tumbling would allow you to dextrously move past a slower and clumsier opponent is not problematic in and of itself. The problem is that the DC for check is flat: It would be fine if the 5th level rogue could tumble past a whole brigade of 1st level warriors, but it shouldn't be possible for that same rogue to tumble past Cyrano de Bergerac or Benedict of Amber.

Essentially, the Tumble skill needs to somehow take the skill of the person you're tumbling past into account.


Over the years I have seen several attempted solutions (and attempted many myself). These include:

1. Opposed tumble checks.
2. The tumble check is opposed by a Reflex saving throw.
3. The tumble check replaces your AC.
4. The tumble check is added to your AC.
5. Add your tumble bonus (sans Dex) to your AC (either with or without a flat-DC check).
6. Add the level of the character you're tumbling past to the DC of the check.
7. Add the BAB of the character you're tumbling past to the DC of the check.
8. Add the melee attack bonus of the character you're tumbling past to the DC of the check.
9. The tumble check is opposed by a melee attack roll. (If the target succeeds at the opposed check, a new attack roll is made to resolve the AoO.)
10. A flat-DC check grants you a flat +4 dodge bonus to AC. (The Mobility feat either makes the check unnecessary or stacks.)


None of these are entirely satisfactory, in my opinion. The reasons include:

1. Tumbling is not a required skill for being a highly-skilled swordsman. Not only doesn't it make sense for my ability to hit your with a sword to be dependent on my ability to (literally) jump through hoops, it also doesn't solve the inherent imbalance you're trying to correct: The tumbler will still be able to tumble past the finest swordsmen in the world (since it's unlikely they've taken ranks in Tumble).

2. This mitigates the problem better than an opposed Tumble check (since Reflex saves improve automatically). And although Reflex saves still aren't tied to melee prowess, the conceptual match is slightly better: It makes sense that quick reflexes would allow you to react quicker to a tumbler. But this solution doesn't actually fix the game balance issues: Saving throws simply don't advance as quickly as skill ranks do. The tumblers still outstrip the abilities of the fighters, it just takes them slightly longer to do it.

3. Tumbling should never make you easier to hit than if you just casually strolled by the person you're attacking. It is relatively trivial to come up with situations where replacing your AC with your Tumble check would result in precisely that.

4. One interesting facet of the 3rd Edition rules is that a character's AC is, essentially, a special case of the central resolution mechanic in which you take 10 and then add your various bonuses. (In fact, many variants exist where you roll a d20 instead of effectively taking 10 to determine your AC against any particular attack.) But when you think of AC in this way, the problem with this solution immediately becomes apparent: In addition to your own tumbling skill, you're also adding a second d20 roll to your total. This is obviously not balanced.

5. But simply adding your Tumble bonus doesn't work, either. It eliminates that second d20 roll, but you're still faced with the fact that this would become a huge bonus. Consider the fact that the game is obviously balanced so that two characters of the same level both have at least a decent opportunity to hit each other. Tacking on a +20 bonus to AC obviously throws that out of whack. If that doesn't convince you, simply consider the fact that a magic item conferring a +10 bonus to Tumble costs roughly the same as a +3 bonus to AC.

6. Adding the level of the character you're tumbling past to the DC of the check at least takes some measure of the skill of the opponent you're facing. But level is only proximate to combat, and the system begins suffering some real problems when you try to use it with monsters (many of which have HD higher than their CR).

7. Adding the BAB, on the other hand, is a much better solution. I would recommend lowering the base DCs slightly to compensate in the tumbler's favor here (so the checks would DC 10 + BAB and DC 20 + BAB). The only problem with this is that BAB, while a better proximate of combat prowess than level, is still only proximate: There are many, many things which improve your ability to make an attack. But this is definitely a workable solution.

8. Adding the full melee attack bonus, on the other hand, doesn't work. The problem simply becomes that the resulting DCs end up being far too high to be reasonable for a character of the same level. You can mitigate this somewhat by stripping away most or all of the base DC of the check (so the DCs become equal to the melee attack bonus and the melee attack bonus + 10 for the two checks), but this only mitigates the problem. And, at low levels, it results in check DCs which are too low.

9. This is the solution proposed by Monte Cook in ARCANA EVOLVED and then later picked up by Mike Mearls in IRON HEROES. Cook makes the basic "avoid AoO" a straight opposed check, while effectively giving the tumbler a -5 penalty on a check to tumble through someone's space. Cook and Mearls are both savvy game designers and, as one might expect, this is probably the best solution we've looked at so far: It turns out that attack bonuses and the skill bonus of a specialist tend to stay within reasonable distance of each other at any given level. And by making the check a gatekeeper for the actual resolution of the AoO (the check doesn't determine whether the AoO succeeds or not, it determines whether the AoO can be attempted), Cook makes sure that tumbling never makes it more likely for the tumbler to be hit.

The only criticism of this method is that it essentially doubles the amount of time it takes to resolve the action. This is not necessarily the end of the world, but whenever you add a die roll to the game you're slowing it down. Slow it down enough and it's no longer fun to play.

10. This is a fairly elegant solution, but suffers from two shortcomings. First, it fails to address the "move through their space" element. Second, the lack of scaling with skill has simply been moved from the person being tumbled past to the person doing the tumbling: No matter how skilled you are at tumbling, you still get nothing more than that flat +4 bonus to AC. Tying the size of the AC bonus to the result of the Tumble check can solve the second problem, but only clumsily or through the use of a chart look-up (neither of which, in my opinion, are desirable).


So, taking all of that into consideration, is there a solution which works? For me, a successful rule would need:

1. To take into account the skill of both the tumbler and the person being tumbled past. Highly skilled swordsmen should be tougher to tumble past than neophyte warriors; highly skilled tumblers should be better at tumbling past people than amateur acrobats.

2. Never result in the tumbler being easier to hit than if they hadn't tumbled.

3. Be simple to use and easy to remember. And, to that end, consistent with other skill checks. (In general, if three or four different tasks use the same mechanic, it's easier than if those tasks each use different mechanics.)

4. Minimize the number of rolls needed to resolve the action.


I'm going to make the rather radical suggestion that part of the problem in trying to solve this problem is that there are actually multiple actions trying to be resolved simultaneously. In reality, there are three things these Tumble checks are attempting to handle:

1. The ability to move around the battlefield quickly and nimbly (minimizing the risk posed from people taking shots at you as you run by them).

2. The ability to dextrously move through someone's space.

3. The ability to nimbly avoid a specific attack aimed at you.

I'm going to sugggest that the solution is to split these different actions up and resolve them independently of each other.

TUMBLING MOVE: By making a Tumble check (DC 15) you gain a +4 dodge bonus to Armor Class against attacks of opportunity caused when you move out of or within a threatened area. You can move up to half your speed without penalty. You can move at your speed by accepting a -10 penalty to this check and you can run by accepting a -20 penalty to this check.

TUMBLE PAST: You can attempt to tumble through an opponent's space as part of normal movement. Because you are entering an opponent's space, this provokes an attack of opportunity from the opponent. You must make a Tumble check (DC 25). If the attack of opportunity is successful or the Tumble check fails, you move back 5 feet in the direction you came, ending your movement there. Otherwise, you move through the opponent's space and can continue your move normally.

AVOID ATTACK: When you are hit in combat, you may use an attack of opportunity or swift action to attempt a Tumble check to negate the hit. The hit is negated if your Tumble check result is greater than the opponent's attack roll. (Essentially, the Tumble check becomes your Armor Class if it is higher than your regular AC.)


This solution keeps most resolutions to a simple skill check vs. a flat DC (the easiest of all possible skill checks), but it never negates the opponent's ability to interfere with the tumble. In one key regard it borrows Cook's solution of opposing the Tumble check with an attack roll, but by borrowing from the Mounted Combat feat mechanic of negating a hit it simplifies Cook's solution: Instead of two separate resolutions, there is essentially only one resolution point (nobody is ever asked to roll more than a single check to resolve the action).

We've also made this ability more widely useful, allowing it to negate any hit in combat. As a result, the character must burn a limited resource (either an attack of opportunity or a swift action) in order to perform the attempt. If a character wants to be able to really dance through a mob of opponents, they should pick up the Combat Reflexes feat.

Those flat DCs also hide another useful design feature: Eventually the tumbler won't have to actually make those checks. At that point, the only check which becomes important is the opposed check to negate the hit. This further speeds up gameplay while satisfying our design goals.

  | | Link