Vn8}nìm $_b;\86IO%J7|Mܴ+) gًd^"J|I&cڿ(\n,!K$, ϗ)г [5+@$en6uwY*i*ͻEN kH:ZpADfv:RPטR s! Y}69ӕ38!_~v4}U|iXէiB.b*"Jo.(Yl6<[KxP06Қ/w>_>"fHX7, cj |QĚp?B{LIP)ф/6b;NʏW`?Bp#'\@P>e>-\ I*Fg\ 4:2|blzpzl}Zxq9Ol 8O/|F8m#ʽ@μ[0N}IR#F ۞[K<,5S.FΩ=?5/cH ̀U,XYqxBrCm@Lݢ9cUԇDj[4rlwcƱȉc>ZK;=m4]ѕ M6D3%xg]ga+eq:~L3~%,%!!Vx&~OHHB66rC 醿eu8a{?{' SE+BZPE(Rr7v:L>J6ڎV6as /'@ Oÿ D9 ^uڶ}?mq51e)1X sitvRia:e== YΡZ/íQEH$'/YyLGHÿ/W5he/U\6-m*N1AȀE/'2Ȧ喫ZU*׍G)lG<ᚥsILݬT.>vӿ**em7*}Y~m7yY+eIrc"kdi82:{cV07IR VvYz= ;;O%=Ce眊V?f9c9$3"$Ir|W<WDYZoX: =„neZ|\e2WۘZ[cu)Bk*Zi>ۑ&Zo]WⶮMP>?#Qij#֬tGA`8ݹt4ucSq#p

The Travels of Mendes Pinto: Team Game

Can You Give Us a Descriptor of the Subject? (Part II)

Sergio Mascarenhas
Edited by Stephen Barringer
February 21, 2002  
Editorial Prologue: More Thinly-Veiled Excuses Disguised as Exposition This column, Part II of Sergio's most recent, should have been out in January. Alas, Your Humble Copy Editor has something resembling a life, including such overwhelming concerns as the planning of a wedding (his own), an insanely busy work schedule, and the relocation of his home prior to said wedding for sake of economy. (Yes, at the age of 31, I am moving back in with my parents. Thank God it's only for three months.)

To paraphrase King Claudius of Denmark, so much for that. Back to what you're really interested in: the point-and-counterpoint commentary of Sergio Mascarenhas and Steve Darlington on character creation and ability valuation. In December, they argued about describing their subject with open, English-language descriptors in an attempt to render as transparent and integrated as possible the game vocabulary; now, the second half of that process is revealed, discussing how to evaluate an imprecise and non-arbitrary comparison system. As before, Steve's comments will be inserted throughout the column in (bracketed italics), with Sergio's counter-comments following in [square bracketed italics].

Let the mayhem begin!

--Stephen J. Barringer, Toronto, January 2001


Last month, we spent a long time talking about the logic and methodology of creating descriptors for RPG characters, but we didn't really get into why. Descriptors aren't created for descriptivitis' sake. They're created to be associated with a comparative scale of performance, which allows us to discriminate between different characters or entities when action is called for. This is done by assigning a level to the descriptor in a performance scale. Which means there is a previous question that must be asked and answered: how to handle performance scales?

Performance scales can take countless shapes. From the levels of D&D to the 0-100% of BRP; from cumulative dice (like D&D's hit dice) to dice pools; from binary scales (for instance, ads/disads that you either have or don't) to open-ended ones.

Yet, I think there's a more basic issue when we consider scaling: should we opt for absolute scaling or for relative scaling? And what does this mean?

Absolute Scaling

How long can you jump? How tall are you? How heavy? How fast is a falcon? If you answered "7m, 1.87m, 95 kg, and 200km/h," you scaled all those things in absolute terms.

["And you're an olympic-level long jumper." -- Steve D.]

Absolute scaling means that some abstract scale is created against which all instances of the situation subject to analysis must be measured. The scale is supposed to allow the measurement of all such instances; otherwise, it will be faulty. The absolute scale is the baseline norm of measurement. When things are handled in absolute terms, there is no direct comparison between instances of the set of entities subject to the scale. Comparison is mediated by values assigned to those entities in terms of the scale.

Defining absolute scales is a tricky business, especially in RPGs. What makes it tricky are two things:

First, since most RPG settings are works of fiction, they include entities that fall outside of the familiar dimensions of our real world. (How do you compare the muscle power of a dragon and a Dhar Corondon when they're both far stronger than any elephant?) Furthermore, there's always the possibility of creating an entity that falls outside of the range of entities that were taken into consideration at the time of creation of the scale.

For instance, D&D introduced a way to handle scales of attributes where the human norm is the reference (human minimum was 3, human maximum 18). It was devised to accommodate common fictional humanoids that diverge in moderate ways from the norm (orcs, ogres, giants). Yet, it is harder to accommodate other creatures that diverge to a great extent from the norm (dragons, titans, gods). The early editions of D&D greater gods with a paltry STR 25 is probably the most memorable example of this.

Second, scaling is always subject to mechanical considerations: how will the scale be handled by the system? How will it map into character creation considerations and resolution mechanics? This is especially important in cases where task/contest resolution uses a specified-range randomizing device like dice or numbered cards. RuneQuest has a neat percentile scaling mechanic. The problem is that it also introduced provision for characters that go past 100%, and it did this in such a way that the scaling system began to fall apart -- because, after all, you can't roll more than 100 on percentile dice, and what fun is it to play something that can't fail?

Third, the spread of reliable and widely shared absolute scaling is a recent phenomenon in history. It's more or less three hundred years old. It required the development of finely tuned instruments, shared norms, and common mathematical education. To some extent, the very act of using fine numbers, precise arithmetic and absolute scaling puts you into an indisputably modern mindset, where roleplaying atmosphere is at best compromised and at worst ruined.

People did use absolute scales of a sort, true. Take for instance Mendes Pinto: when referring to time he usually mentions measures like the Credum (the time needed to say a particular Catholic prayer -- according to several practicing Catholics I've questioned, it takes more or less one minute), or hours and half-hours (but not minutes and seconds). When mentioning distances, he uses measures based on missile weapons shots (crossbow, arquebus, different types of cannons). But all this is highly unreliable. It's used more to impress on the reader an approximate idea of time and distance than to provide an exact measure.

(Notice that I will use these measures in my game, because the setting uses it. But I will not bother translating it into exact equivalents in the measuring systems we use today -- other than saying, like I did above, that a Credum is more or less one minute, or that a crossbow shot is approximately 300 meters).

Furthermore, absolute scaling doesn't correspond to the way we handle comparisons in most situations. The reason is simple: seldom do we have the ability to measure things according to an absolute scale. Look around you. Are you able to tell how high (in meters, feet, yards) is the building down the block? How fast (in km/h or mph) is that person walking? Can you tell how heavy (in kilos or pounds) are the eggs you're buying just by looking at them, or even weighing them in your hand? Or (to pick a really sensitive issue) how much higher is your IQ than that of a fellow roleplayer?

You can, I can, anybody can... provided we have the right instruments. If we live in a modern world and are performing some type of technical activity, we will be provided with the right instruments, and use absolute scales to measure the entities we have to handle. Say, if we are driving. Or if we work in a factory. Or if we are buying something in the supermarket. Otherwise, you're back to approximation.

Yet there is no absolute norm that says that all things should be handled with absolute scales. There are plenty of instances where this does not hold true. What absolute scale do you use to measure the joy you take in music? To say how much the stink from the garbage can is disturbing you? To locate a person in a crowd based on his size? Or to use more game-oriented examples, do you really care if the backpack you're carring weighs fifty kilos? You probably care more about the pain it's generating in your back. How important is it to know that Excalibur is 1.4 meters (or 4'8", if you prefer) in length? Is this information what makes that sword magical and wonderful?

The conclusion? Absolute scales in RPGs create more problems than those they solve:

They don't correspond to the way we think and how we compare things in most situations.

["Actually, they do, because in truth, our absolute scales are actually comparative. Nobody says they are "smarter than Bob," they say "I am smart." But what they mean is "smarter than average." 50kg means "50 times heavier than a bit of metal in Paris." Likewise, Superman and Conan have absolute strengths, not comparative ones. Superman is more powerful than a locomotive, but not more powerful than the Hulk. Conan has the strength of twenty men, but not the strength of Red Sonja." -- Steve D.] ("Yes, they say, "I am smart," but if we ask them "How smart?" they will not say "I am smarter than average" (who knows what's average?). They will probably say, "How smart? Well, I'm smarter than Bob, or Pete, or...." The examples you give of Superman and Conan are clear examples of relative scaling. Who is stronger or more powerful (for argument's sake, let's say that they mean the same thing): Superman, Hulk, a locomotive, Conan, 20 men or Red Sonja? We can only say that Superman is more powerful/stronger that a locomotive, and that Conan is more powerful/stronger than 20 men." -- Sergio)

They commonly default to artificial, system-oriented scales. BRP size scaling is based on the dice, not on the conventions we use to measure things in absolute terms. That diminishes meaning.

They tend to fall victim to numeric quantification syndrome -- making in-character decisions based on out-of-character degrees of fine distinction. Steffan O'Sullivan, when explaining why he wrote FUDGE, noted that in practice many RPGers decide who will attempt a task based on ability comparisons like, "I've got STR 75, you've only got STR 72 -- let me try." This is clearly unrealistic and atmosphere-destroying. They tend to break away when we need to go above, below or inside their normal range. That's why people complain about so-called generic or universal systems.

They usually are created for a particular setting or group of settings. The moment one attempts to go in other directions, the scale starts to fall apart. They contribute in no small amount to the difficulty of porting ideas from game to game ("Don't" #7, from the previous month's article), since different games tend to use different measurement systems.

They tend to concentrate on the humph side of things, we want the wow. It makes the game harder, especially when we consider the GM.

["Actually, I find absolute systems so much easier, cos you only need one set of target numbers." -- Steve D.]

This last point requires some more explanation. Suppose you are a player. You come out with a nice concept for your character. It's in your mind, you can visualize it. What do you do next? Do you write it down? Possibly, but there's something more you are going to do. You are going to translate your idea into the terms of the game system you're using, assigning values according to the predefined scales of the game.

In the course of this two things will happen: you will change the concept of your character, both because there are things that are present in your idea that cannot be represented within the system (in this case you are losing meaning) and because there are things in the system that are not in your idea (in this case you are adding unwelcome meaning). You'll be spending a lot of time doing something you really don't want to do (1). The reason why you are spending that time is because the game designer needed to fit his relative, approximated setting into the assigned scales of his absolute mechanics system.

Now, suppose you are a GM. You've found a very nice scenario for game X [TM] and would like to play it, but you're using game Y[TM] instead. You may have to spend a lot of time making the conversion from X to Y, and may risk losing control of it, once more either because there is loss of meaning from X to Y or because there's new meaning in Y that was not in X. (A great adventure for GURPS Fantasy can be torpedoed in Dungeons & Dragons because D&D characters will often have magic items that GURPS characters don't, for example.)

This happens if the games have things "hardwired" both in terms of game entities and their values according to the power scales they use. Since hardwiring game entities happens most of the time when there's a need to assign a scale to it, ultimately it is the drive for scaling that is behind the difficulties in making conversions among games.

Furthermore, one is subject to a lot of factors that can change the balance the absolute scales were supposed to provide. You have made a scenario that is balanced... if it's played by the characters of your regular four players. What if one of them doesn't show up today? Or a new player appears at the last moment? What if players come up with that killer idea you didn't think about when designing the scenario -- especially in the moment you were laying down the stats?

(Of course, that will happen in any game, anytime, anywhere. But my point is that it's a lot easier to come up with a solution on the fly when you don't have to calculate a lot of numbers and their interaction in your head on a moment's notice.)

Relative Scaling

There is another way, and it's called relative scaling.

Who's taller, you or your brother/friend? Who's stronger? Who's faster? You know the answer. Not because you measured it against an absolute scale, but because you have the experience of direct comparison. When you do this, you are measuring things in relative terms. Relative to the two entities being compared.

How does relative scaling work? Simple. Let's see:

PC Creation: All the players have to do is to worry about their choice of descriptors. They concentrate on the key factors of their PCs. This can be done in the same way as Hero Wars, for instance. Just note what your character is, what he does, what he has, how he evolved. If you know this, you know what can reasonably be expected from him. Don't bother to express this in absolute mathematical measures.

OK, but how does your character compare to the PCs of the other players? First, if there are descriptors you used for your character that they didn't, that means your character will be special in the fields covered by that descriptor. How special, and in which sense? That's up to you to decide (with the help of the GM). I guess that most of the time you will want your character to be better at it than the other characters. So, he has an edge in that field. Once more, don't bother trying to measure that edge, just acknowledge that it's there.

Yet, there may be instances where two players assign similar capabilities to their characters. How do these two compare? There are various possibilities:

  1. The traits originally defined are similar but do not overlap completely. That means that there will be circumstances where PC A will perform better than PC B, and vice-versa.
  2. The traits are similar and overlap completely. Both players decide to define instances that make them different, actually changing their abilities to suit the situation above.
  3. Both players accept that one of the characters will be better in the trait than the other.
  4. The players accept that there is no difference in performance between the two characters.

Let's see an example (TToMP related): PC1 is a European soldier; PC2 is a Japanese samurai; PC3 is an Indian merchant with no fighting-related traits.

We assume that PC3 is the worst sword fighter of the three; in any fight between him and PC1 or PC2, he will almost certainly lose. European soldiers know how to fight with European swords, while samurai know how to fight with Japanese katanas. Who's the best fighter, PC1 or PC2? It depends on the circumstances. Do they favor the European or the Japanese style of fighting? Or are they neutral?

Then there's PC4, who is also a European soldier. Is he better at sword fighting than PC1? Well, the players may decide that PC1 excels when using two handed swords, while PC4 is an expert on sword and shield fighting. Or they may agree that while PC1 is a good sword fighter, PC4 is a master sword fighter. Or even reach the conclusion that there's no marked difference in the swordfighting skills of PC1 and PC4. In this case, external circumstances (like health, psychology, etc.) will dictate if one of the characters has an edge, otherwise they are on par.

(Steve says that I should mention Amber "since it made the purely comparative stat idea famous". Here is the mention even if I don't know Amber.)

NPC creation: Relative scaling makes things easy for the GM. It all comes down to a simple question: Are the NPCs better or worse than the PCs? Are the NPCs advantaged because they perform individually better/worse, so that in a one-on-one situation they are at an advantage/disadvantage, or because they rely on numbers to compensate for their individual weaknesses? What's the intended balance between different NPCs and PCs along the gaming session?

["I don't see how this is any different from absolute scaling. PC1 -- Good; PC2 -- Good. Want NPCs to be tougher? Then NPCs = Great." -- Steve D]

("What you mean is that relative scaling can be done even with games that have absolute scales. Say, NPC A will always be 10% better than the PCs; or he will have 2 levels more; or whatever suits the system. As Steve says, "This is one of the cool things about comparative scaling, but you can do it with absolute scaling. It's just comparative scaling encourages to you to think that way. Which is why it's great for epic games like HW or Amber." I would say it works for any type of games, not necessarily epic games." -- Sergio)

Let's have another example: The group of PCs above are to face the devilish machinations of Major NPC in the court of Ayhodia in Siam. He is a diplomat (which overlaps the trade skills of PC3 on negotiation, with an edge to Major NPC in political negotiation, and an edge to PC3 in mercantile negotiations). He is on par with PC3 when fighting becomes a problem, and seriously disadvantaged to PC1, 2 and 4. He has some important non-skill points to his advantage: he is better connected, knows court life better, knows some secrets of a couple of people in power, is richer than the PCs, and has fewer scruples. He also has a lieutenant who is a veteran of many wars (on a par with PC1, 2 and 4 on fighting skills). He has a hired assassin that is better than any of the fighting PCs if he surprises them, but is at a disadvantage if fighting eye-to-eye. And he has around 6 toughs that are better than the merchant PC3, but not as good as any of the fighting PCs. The fighting PCs have their advantages, especially in the case of war gear (they have good quality armour and heavy weapons, while only the lieutenant of the Major NPC has similar equipment).

I guess that by now you see where I'm going: the GM must know the PCs, and model his NPCs around the PCs.

There's a well known strategic management matrix called SWOT. It stands for, "Strengths and Weaknesses, Opportunities and Threads." It provides an interesting way to handle NPCs. Put the PCs in the centre; next, figure out what are the NPCs' strengths and weaknesses when compared to the PCs; then, what opportunities and threads the PCs present to the schemes of the NPCs (and vice versa).

As you can see, it all boils down to how the GM wants to handle the setting and the plot (2).

What About Character Advancement?

The way I'm presenting it seems to leave character advancement, one of the key aspects of roleplaying, in the dark. How does a game system based on the ideas presented so far handle it?

First, recall that I don't put too much emphasis on character advancement (3). There are other things to roleplay than simply trying to improve on functional abilities, things like the importance of property, relationships, knowledge, well being, psychology, goals. At the same time that I de-emphasize functional advancement, I put more stress in these aspects.

["Of course, and particularly given the freeform system you're proposing, these things can be stats, so they can be improved. And indeed may improve other stats. This is just semantics." -- Steve D]

("I would say that they can change. That's more accurate than to say that they can only be improved. Emphasizing this isn't simply semantics, though. Furthermore this is not new. CoC did just this when it introduced the regressive SAN stat. In a game that works this way, who will concentrate on improving skills? Yet CoC marked a clear departure from the mold set by D&D. And it was not just semantics." -- Sergio)

On the other hand, this approach does not give up character advancement completely, it just handles it differently and in a way that's setting-oriented. How?

The first point is for the GM to decide at which level of functional capability the PCs will start. For simplicity, let's suppose that there are four levels: Inexperienced, Layman, Veteran, Expert. If the PCs start at Inexperienced, there will be scope for progression. Yet, progression does not mean gaining an abstract value. It means a change of balance between PCs and NPCs.

Imagine a pyramid showing the distribution of capabilities in the population at large (such as is often seen in demographic studies). At the bottom are the Inexperienced, above these are the Laymen, etc. Of course, there will be many more Inexperienced persons than Veterans, and the Experts will be a minority.

To move up the ladder you have to gain experience and knowledge. Either you are taught by someone of higher experience, or you get experience in the field of action.

["This is the common absolute/comparative scale used by many games these days, where a task is Easy/Difficult depending on the scale of the character." -- Steve D]

("No, this is not what I mean. The task is Easy/Difficult in context of the scenario, not the characters. I mean, the GM wants the task to be easy, so it is easy. If he is playing the scenario with a group of hardened Veterans, the task is easy. If he is playing with a group of Inexperienced grunts, it's still easy. The ladder I'm mentioning is not a mechanics ladder. It's a setting thing. It's like a shade in the background. It influences the way one interprets the results generated by the system in accordance with the setting. Suppose it's a combat. In the case of the Inexperienced grunt, "Easy" means the adversary is a completely and helplessly ignorant grunt. So the GM describes the NPC accordingly. In the case of the Veteran, "Easy" means an Inexperienced grunt. Once again, the description is adjusted to the features of the setting (of which the characters are part) but the underlying mechanic is the same." -- Sergio)

In any case, gaining experience does not change the description of the character (other than the change the GM makes in his notes). What it changes is the way the GM handles NPCs. An NPC that was a challenge once is no longer a challenge. That expert that used to be a deadly foe is now beatable.

How do players know that their characters have moved up on the experience ladder? In pure roleplaying terms. Yesterday they knew they were no match for Expert NPCs, that they were on a par with Inexperienced NPCs, and faced a challenge to defeat Layman NPCs. Now they see that it's become easy to defeat Inexperienced NPCs, that they are on a par with Layman NPCs, and that there's hope to beat even a Veteran (but they're still no match for an Expert).

In this system stats don't change. It's the interaction between the PC and the surrounding NPCs that changes. Notice that this is still a way to do relative scaling. The difference is that it establishes the relationship between the character and average values for the members of a given population at different levels of performance within a certain setting. It does not attempt to define an universal scale that can be used for any situation.


I mentioned in the previous columns that I was aiming at a generic system that could cover any situation. I think that what I'm proposing achieves just this. Why? Because it gives up absolute scales and predefined descriptors.

You see, with this system all one has to do is to pick a setting -- any setting -- identify the basic features of the setting and devise characters (and other game entities, living or non-living) based on these features that are consistent with it. Since mechanic-wise everything is dealt with in relative terms, it's the plot that drives balances of power, not the other way around.

It also allows for balancing among different settings, no matter how different. Want to pit a medieval knight against a Jedi? Just assign the advantages and disadvantages to each party.


They were a long pair of columns. They had to be. I presented the principles that underly character creation in the system that will be used with TtoMP. What is not explained is how to link the proposed system of character description to resolution mechanics. We will get there, don't worry. But before that I'll turn the ideas discussed this month into a full fledged game resource. In other words, I'll present in the next column a character creation rule system that implements what I present above.


(1) Unless, of course, you play by the system and are so used to it that you can only think in the terms it provides.

(2) Here you may consider re-reading Shannon Appel's column.

(3) I suggest you check the Ruleslawyer for Free column I wrote specifically about this issue: Ruleslawyer_Sergio_9_1.html. Furthermore, this approach adjusts very well to the mood I'm envisioning to TToMP as I presented it before: ruleslaw28jun01.html.

TQo0~^DҒt< ek&Ǿ$\۵ZFȃuwݝIŃU QYir2HR2.u3MFoعq]4#A`pP5(b& )b)ⰾp7(i<[-2gL#5[f g?*rVGf8*)s'+20ϟ̑F}KB<7wSL\gbvm9WiRބYŜvd y0'p2I_Fc2>#o A )VL[Qk?3`)<У[(*W.JH ?tXCt谙 X:@ \0w ~LqĤE-rFkYœj4q 5AQ6[AxG [>w|?( fХθY䝛$c=_qNĦoǸ>O_|&/_Mi7"宥CЧk0dӷLh;TmuCGU-!Ul{ h<\bQX.~"O2*yPcz!ŠGg

What do you think?

Go to forum!\n"; $file = "$subdir/list2.php?f=$num"; if (readfile($file) == 0) { echo "(0 messages so far)
"; } ?>

All The Travels of Mendes Pinto columns by Sergio Mascarenhas

  • Not Dead Yet by Sergio Mascarenhas, 18jul03
  • Still Struggling With Section 4 by Sergio Mascarenhas, 07mar03
  • It's Time For Combat by Sergio Mascarenhas, 31jan03
  • The Criminal Always Comes Back to the Scene of the Crim by Sergio Mascarenhas, 31dec02
  • Tahahahahime Is On My Sahide! Yes It Isn't! by Sergio Mascarenhas, 07nov02
  • Character Creation for Dummies by Sergio Mascarenhas, 08aug02
  • Thus Do We Start by Sergio Mascarenhas, 01jul02
  • Can You Give Us a Descriptor of the Subject? (Part II) February 21, 2002
  • Can You Give Us a Descriptor of the Subject? (Part I) December 6, 2001
  • Time for a Hiatus October 23, 2001
  • System as Language September 27, 2001
  • Don't Forget the Damn Index! August 30, 2001
  • R-O-L-E-P-L-A-Y-I-N-G  G-A-M-E  C-O-R-E  R-U-L-E-B-O-O-K.&nb\ sp; UFF! July 26, 2001
  • Who is this Mendes Pinto, Anyway? June 28, 2001
  • Once Upon A Time ... (How It All Started) April 25, 2001

    All Ruleslawyer For Free columns by Sergio Mascarenhas

  • Experience: From Fiction to Roleplaying Games September 11, 2000
  • The Applied Experience Curve Concept June 26, 2000
  • Experience Curves May 30, 2000
  • Trait Curves March 28, 2000
  • A Change of Course November 28, 1999
  • Spotlight on Alternacy, A Roleplaying System October 26, 1999
  • Introduction September 21, 1999

    Other columns at RPGnet

    TQo0~^DҒt< ek&Ǿ$\۵ZFȃuwݝIŃU QYir2HR2.u3MFoعq]4#A`pP5(b& )b)ⰾp7(i<[-2gL#5[f g?*rVGf8*)s'+20ϟ̑F}KB<7wSL\gbvm9WiRބYŜvd y0'p2I_Fc2>#o A )VL[Qk?3`)<У[(*W.JH ?tXCt谙 X:@ \0w ~LqĤE-rFkYœj4q 5AQ6[AxG [>w|?( fХθY䝛$c=_qNĦoǸ>O_|&/_Mi7"宥CЧk0dӷLh;TmuCGU-!Ul{ h<\bQX.~"O2*yPcz!ŠGg