Article Index

chartI know, there is the “What is all this?” box, where I explain that the aim is to provide a Second Chance (Második Esély) if the d6-based solution fails, and that apparently settles whether we should use d6 or d100. Does it really? In my previous article I pointed out that, in many cases, the commercial cost of d100 is an open question. Some people, of course, think bundling dice with the book is the answer. But that solution is no solution, because I can name at least three situations in which it does not and cannot work. The first is that the rules needed to try the game could be attached to a novel or short-story collection. A few pages make little difference if there is room left on the printing sheet anyway. Bundled dice would bother many bookshops here. But the even bigger problem is that they would not reach readers of library copies. Libraries still exist today. So do secondhand bookshops.

The other possibility is, of course, any free introductory rules booklet that can be printed anywhere and picked up in lots of places. You are hardly going to bundle dice with promotional material, because then many people would take it just for the free dice, despite already having a copy and not being the new players you are targeting. The third consideration: even well-known companies such as White Wolf have now moved largely toward print on demand. And with print on demand, you do not want to stock dice. Ordering dice, which in some people's fantasy world takes two clicks, is not just two clicks either. Let us start with this: if you give something away so that people can start playing the rules that came with the book or free booklet without spending money and immediately, your first sentence will not be “Just buy something from somebody else first and wait for the order to arrive.” Oh, and hold on: have a bank card. Shall I go on?

The other consideration is that, like it or not, a roleplaying game can succeed if you reach as many potential players as possible. Because, using the shared rulebook, they too can become game masters. Good game masters, too. Your first goal is to bring the barrier to entry down to nearly zero. One important feature of roleplaying is having a group you can play with, so excluding as few people as possible is essential. Of course there are plenty of homemade tools for generating random numbers. Drawing little chits worked even in D&D's early days. I mentioned cards myself in the previous article. I have seen “where does the homemade spinner stop?” used too, as well as opening a book and taking the last digit of the page number as a “d10.” The modern solution is a dice-rolling program. But I have even seen a target used, with the observed spread generating the random number. A system designer should be creative.

A substantial share of dice-using systems, however, currently fall into three categories. There are systems in which most rolls use a single die, to which we add modifiers and compare the result against another value, or a table of such values. Every result has the same probability. Then there are systems using several dice, where the sum of the roll matters. This usually means a variable number of dice, possibly with modifiers. We see a natural distribution in the roll. But when increasing a value means going from 1d6 to 1d6+1, 1d6+2, 2d6, and 2d6+1, each increase raises the expected value by 1 or 1.5 points. Although more dice point toward a natural distribution and predictability, replacing a fixed modifier with a die points toward uncertainty, and the maximum rises quickly and somewhat disproportionately. Finally come success-counting systems, where the mathematical background is quite different.

Even in those three main categories, some dice may involve a trick, such as rerolls. Some systems put things other than numbers on the dice, and swapping dice matters too. And of course many have even more complicated tricks, such as the roll-and-keep concept. The point, however, is that some people consider D&D modern with its many types of dice and thousands of modifiers. Others point out that modern systems do not use so many kinds of dice, and that its 5% steps, high critical chances, and many other concepts make D&D less than modern. Many even point to classes. But a flame war is guaranteed if someone modernizing a fantasy game wants something other than turning it into D&D. All those dice, thick core rulebooks, and so on do create a high barrier to entry in one respect.

What is all this?

There has not been a successful new M.A.G.U.S. system for a long time. Many of us play from twenty-year-old publications that are practically falling apart. We wait for new solutions, then the grumbling begins: why does the new system roll attacks that way?

But we have often heard that the rights holder would welcome homebrew and other attempts too. It has even been said that perhaps more than one could get a chance. They waited in vain. We did not trust that there really would be a chance. Delta Vision's developers are currently working on a new system. The fans are working on grumbling about the d6. 

But let us admit that four alternatives currently exist:

  • We get the d6-based system.
  • It fails and we get nothing.
  • We deliver a finished system and it gets a chance, either instead of or alongside the d6-based one.
  • A series of articles presents a method for producing a new core rulebook.

No developers of a homebrew system have been sued. And we have heard of at least four or five solutions. But there is official* permission for supplements and articles, and none for a homebrew system. 

Even if permission eventually comes, the articles will not form complete “core rulebooks,” only an SRD. Let us call this the Second Chance system.

* This does not make the resulting supplement official or officially approved. Only the permission for fan material and house rules is official in the legal sense. 

But that is not the only kind of barrier to entry. Money is not the only barrier: in a game system, most people become GMs only if they can understand its mathematics well enough to design a working adventure and estimate the likelihood of success beforehand. Here, however, there are at least two ways to improve the game master's chances. One is simpler mathematics: see the “single-die system.” The other is deciding that character performance should depend less on chance. They undertake things they will survive and accomplish. The degree and nature of success depend on the roll; there are stakes and something to get excited about, but it rarely decides between success and failure. There are, however, relatively infrequent “special events” that players talk about for days.

We have guessed correctly. Percentile systems have the enormous advantage of making probabilities very easy to calculate. In that respect, few systems are better than BRP, which underlies Call of Cthulhu and also handles character development sensationally well. But it too puts a burden on the game master: what do you do with the relatively frequent “special rolls”? Suppose a combat round contains ten people acting; you roll 1d100 at least ten times. Over five rounds you would expect fifty such rolls. On average, every fifty rolls there will be a special result: either 00 or 01.

An ill-timed roll—a 00 against a player, or their own 01, can easily prove fatal—is harder to handle as a game master, and it is a recurring topic. Moreover, M.A.G.U.S. was originally intended for campaign play following a character's life. On average, every hundred enemy attacks a player takes a 00, whose direct consequence in many cases, in the appropriate combat situation, is applying the moral rule. Many consider that damaging to the atmosphere. And if the goal is for stories to resemble novels and short stories, this flaw keeps resurfacing through long rounds of testing.

We can also see that, say, a character is usually capable of two attacks per round in most rounds. That is ten attacks in five rounds. Suppose they make no roll in the sixth, merely moving, so there are ten rolls per minute. On average a 01 comes up every ten minutes. Sorry, but we see fewer catastrophic mistakes than that at a junior fencing competition. 

There is something worth noticing about multiple-dice systems that count successes. Usually they offer more than ordinary and critical success: several things measure the degree of success, so an extremely good result is rare. By contrast, an extremely bad result is usually tied to rolling all ones. If the character rolls a single six-sided die, that happens frequently, with a probability of 1 in 6. But this is a tremendously risky attempt by a conspicuously clumsy, untrained person. Two dice are still not really good, and for someone who can muster only that many, the world itself also helps us understand why they will blunder so often. With three dice, the chance of trouble is already only 1 in 216, less than half the probability of a 01 on d100.

With four dice the chance is already 1 in 1,296. With five, which is perfectly plausible for a professional character, it is 1 in 7,776. And we can continue adding dice, because even eight to ten is not unusual. Catastrophic mistakes become believably rare. If we consider only the mathematical model so far, it seems obvious that a multi-die d6 system can be much better than a d100-based one. Until we ask ourselves what else we can do about rare results. 

Rolemaster's open-ended rolls come to mind. The mechanic worked by having us reroll on 01–05 and subtract the result from the roll, while on 96–100 we rerolled and added it. A lot depends on the order of operations, of course. Suppose I have an Attack Value of 37. My first roll is 02, my reroll 96, and my final reroll 28. It matters whether the result is 37+2−(96+28) = −85 or 37+2−96+28 = −29. Both interpretations of the rule could work. Now imagine the following rule: you can hit the wrong person if the roll's value is negative. In that case, compare the absolute value of the roll with the Defense Value of the “accidental target,” possibly yourself. And instead of 00, a multiple overhit is required, which takes several good rolls.

For professional characters with good scores, catastrophic results are suddenly rare, and the system becomes more believable. But suppose there are still too many situations where the enemy rolls too well and the moral rule would come in. In a life-spanning campaign, you do not want to “die” every two or three adventures, and you do not want to swallow the atmosphere-breaking moral rule either. You can either keep playing with the mathematical model or play a little with the interpretation of the moral rule. For example, the system can provide guidance on what saves you.

Even just the fact that M.A.G.U.S., or the Chronicles of Adventurers, refers to a work being written on Ynev. Whatever is written there passes into the world. A Ynevian hero has made use of this, and there is a literary example where it appears in connection with an apparently fatal event. Although Dragonlance has had something similar, it is rare for this to be central to a setting's concept, a fundamental premise of the game, and it forms part of M.A.G.U.S.'s flavor. Extreme rolls can be modified by a lucky amulet, special abilities, points awarded as rewards, and who knows how many other things. The effects of many such modifiers can be described in ways that enrich the game.

A conventional single-die roll, the sum of multiple dice, and an open-ended roll—even considering a single die in Shadowrun 3, for example—have completely different distributions. This also affects how the game feels to us and how we perceive its world. In many respects d20 is a lousy solution, but AD&D and D&D lovers associate it with many pleasant memories. They become excited as a conditioned reflex when they reach for the die, and this makes the game more exciting. Talking about a “natural 20” also gives them an internal language that symbolizes belonging. These “iconic” things strengthen the brand. Breaking with them for a system that “looks better” can damage that brand.

Bringing in cards instead of dice is much less extreme than switching to a different die. Of course, if many players “insist” on dice, you can manufacture special dice for them, with special dice replacing “deck building.” “Only from you, only for fans of your game”: 3D-printed dice with a special design. That just requires a very good 3D printer or some other equipment suitable for producing dice. A software-based solution that builds on the reflexes associated with a “dice-rolling program” is equally possible.