At AnthroCon, Matt Sowers handed me a small booklet that said in big letters, No Room For Pictures Role Playing Game Rules (Beta Copy). This, he informed me, was his attempt to meet the challenge of one of the designers at Steve Jackson Games “to present a fully playable game system, with a working rule set and enough examples to generate a complete role playing game in under 32 pages.” Matt wanted my input on his entry based on my frequent nattering here about my tabletop RPGs as well as my experience (however brief and long ago) in professional game design.
In an effort to be worthy of his trust on the matter, I’ve decided to give as detailed an answer as I can, having looked through No Room For Pictures (but not actually having tried running it). The short answer is, NRFP is not a game system I would be likely to use, partially due to systemic flaws, but mostly because there are so many other systems that already do everything it does, and some of them do it better. But there’s a longer answer as well.
I should point out, there are plenty of games on the market that already rise admirably to the SJG challenge. Savage Worlds, for instance, can be summarized in completely playable form with something like 16 pages, and is famous for its scenarios being one double-sided page or shorter. (It’s also supposedly brilliant at handling mass battles, but that requires another ten pages or so.) Savage Worlds works so well in this space, in fact, that any attempt to fill the same slot is going to have a really high bar to reach. The West End Games “Universal How-Much” (UHM System) created for Ghostbusters, that eventually evolved into the famous and well-liked d6 System, can be completely summarized and playable in something like 10 pages. There’s no shortage of rules-lite systems out there.
That said, what are the specific things about No Room For Pictures that would make me send it back to the drawing board? We’ll start with the core mechanics.
d20 Dice Pools
First off, NRFP is a “dice pool” game, and the dice used are d20s. The core mechanic is to roll a number of d20s based on your stats or skills and beat a difficulty based on the situation, starting at 10 and being modified by circumstances, then count your successes to determine how well you do at the given task.
This starts off hitting what I freely admit is a personal prejudice: I just don’t like dice pool mechanics, particularly with a sliding difficulty scale. My experience is that they tend to add a lot of math at the table and produce wild “always succeed or always fail” results. There’s always that one person at the table who can’t remember from round to round how many dice they’re supposed to roll for any given thing, and the GM has to learn what all the “standard modifiers” are. Add to that opposed dice pools, which NRFP uses for combat, and you’ve got tons of dice being rolled all over the place all the time, quickly escalating into a giant math-y mess.
The only way to minimize the piles of numbers being thrown around is by using small dice, probably d6′s or d10′s, because the math can be done quickly, but NRFP chooses a d20, I’m guessing largely because White Wolf kinda snagged the d10 already and everybody else was using d6′s. Unfortunately, d20 is pretty much the worst feasible choice. Although there is very little “mathematically” different between rolling d6 and trying to beat a 4, versus rolling d20 and trying to beat a 14, it’s still psychologically harder to grok quickly.
Combat is the biggest offender here. The attacker rolls a pool against a difficulty equal to the target’s defense score; the target then “can choose to dodge” (but I see no reason why they wouldn’t always do so), which is rolling a pool against a difficulty of 10+(the number of dice in the attacker’s pool). The likelihood of successfully dodging is low-to-nil most of the time, but the hit point pool is so low that anything to mitigate damage is worth trying… even if it bogs combat down. This might be fixed by having dodge become its own separate action, sacrificing a combatant’s turn, or possibly by having people split their combat pool into “attack dice” and “defend dice” or something– but right now the system more or less encourages just standing there and trading blows.
Healing also seems to be quite harsh, particularly for a game that has such rock-’em sock-’em examples. Health itself is a flat hit point total with no penalties for being injured, but you only regain 1 hp/week (with a typical hp total somewhere in the 4-8 range)… so one fight could easily put the kibosh on a whole party. That’s great for, say, Lovecraftian horror, but not so hot for kung-fu action. Some kind of distinction between “combat hits” and “longterm injury” or options for different genres might not be amiss.
Skills, Character Customization, Weird Talents
This part of NRFP is seriously under-developed. There is a framework for skills and a few samples in place, but no real skill list. The few skills presented have fairly arbitrary costs and and pre-requisites, and there’s no guidance on how the would-be GM should assign these other than a vague “more specialized skills cost less but have more pre-reqs.”
As for customization and/or weird talents, there ain’t none. It looks like they were intended– some of the sample foes have spells for instance, but there’s no description anywhere of what those spells are or how they work. “Fireball (5)” is presumably an attack spell with a 5d20 dice pool. But what is “Bellow (8)”? A damage spell? Deafening? Inspire/Intimidate? No idea. Extrapolating from what’s here, I’m guessing that weird talents (such as spells) would be bought as Resources, the same way money is.
The Cumulative Effect
Consider the provided example of picking a locked door:
Our hero, Bao Sing, has been locked in a storage room after being knocked on the head. He needs to get out, because anyone who wants Bao locked up isn’t someone Bao wants to meet.
Trying to be sneaky, Bao tries to pick the lock on the door. He, for some reason, has a set of lock picks, but no real skill in using them. On top of that, it’s a pretty good lock.
Bao needs to roll an unopposed test to pick the lock. Due to the fact that he’s defaulting to his Speed score, the GM decides that Bao’s penalty is +6, and the quality of the lock adds two more to his penalty.
So Bao is rolling his base Speed (4) against 17. 10 (Base) +6 (Unskilled) +2 (Lock Quality) -1 (Lockpicks).
Needless to say, he fails with no successes. So he’ll go for the noisy method, and kick his way out.
Let’s take it apart piece by piece:
- Why is he defaulting to his Speed score? There’s no “Lockpick” skill defined. One could convincingly argue that lockpicking is about Smarts instead of just nimble fingers. Presumably it’s a GM judgement call.
- Lacking a skill already means you don’t get to add dice to your pool with a task; presumably the additional difficulty is prevent players who just have ridiculously high stats from facerolling over every obstacle anyhow. But the suggested penalty range (up to +10) is crazy huge. I would propose instead that instead of modifying the difficulty number, being untrained should set a cap on the dice pool you can use for that task (say, 2d20 or the relevant stat, whichever is lower).
- Why does having a set of lockpicks reduce the difficulty of this task? I mean, shouldn’t it instead simply be that if you don’t have lockpicks (or something that can act as lockpicks), you can’t pick the lock? I could see having a really nice set of lockpicks providing a bonus, but I would suggest it came in the form of +1d20 to the pool (effectively making you more skilled) rather than modifying the difficulty number.
- Imagine for a moment that Bao Sing was skilled at this task… say he had Lockpicking (1). Suddenly he’s rolling 5d20 against a difficulty of 11– for what is supposed to a “pretty good lock.” I don’t have the math chops to work it all out off the top of my head, but even with 1d20, a difficulty of 11 is a 45% chance of success. What are the chances when you flip 5 coins that at least one will be heads?
These are the kinds of messes you let yourself in for with a dice pool mechanic. If you were set on it, tho, I’d alter it thus:
- Use a standard difficulty, which should be fairly high. Say, 15. Someone with 1d20 in their dice pool for a task would only have a 25% chance of success there.
- Make being unskilled at a task put a cap on the maximum dice pool you can use for that task as the default (say, half your relevant attribute). I’m a fairly brainy guy, but I have no idea how to pick locks. But being a fairly brainy guy, I could probably figure it out with time.
- Have some skills default to the value of the relevant attribute. I’m thinking things like “Brawl” and “Drive” here, which most people have a passable knowledge. Have other skills completely closed to anyone untrained (e.g., computer programming).
- Have beneficial conditions add dice to the pool. A really nice lockpick set gives you +1d20. A diagram of the specific lock you’re trying to pick gives you +3d20, etc.
- Have penalties come in the form of required number of successes. Easy task: 1 success. Moderate task: 2 successes, and so on. “A pretty good lock” might require 6+ successes, but with the mitigating factor being that you can try more than once and add your successes over time. (Or something similar.)
A Solution In Search of a Problem
But honestly, and I don’t want this to sound harsh ’cause Matt’s a cool guy and I don’t want to just dump on his project, I think it would be better to chuck the current setup all together and look for some other route. The gaming industry has dozens of perfectly feasible rules systems that can do this basic sort of resolution-based narrative. “Roll stat beat difficulty” is where tabletop RPGs have been for 20+ years now, and it’s a pretty mature concept. NRFP’s most fundamental flaw is that it’s re-inventing the wheel, when what Matt should be doing is trying to create wings.
Take a look at some of the past decade’s big new concepts. Gumshoe threw away the whole idea of rating your character’s Strength, Health, Height, Weight, Turn-Ons and instead uses your character’s drives (i.e., their motivations) as the engine of their success or failure. Dread is a very cool horror game that resolves everything by playing Jenga(!)– “Want to fail at the stated task? That’s free. Want to succeed? Pull a tile from the Jenga tower. If the Jenga tower collapses, your character dies horribly.”
To create something cool and exciting, I would suggest Matt come up with some specific problem that he wants to solve, and then create a game that solves that problem, rather than just making a different way to solve the problems of 20 years ago. Gumshoe came about because Robin Laws wanted to run mystery scenarios that wouldn’t get stymied by a missed Search check. Dread came about because the creators wanted to add real suspense into their horror game. Savage Worlds was written because the creators wanted a game that was “Fast! Furious! and Fun!” and could handle enormous combats quickly while still supporting roleplaying. The only problem NRFP solves right now, is fitting under 32 pages.
PS: In Matt’s defense, I got a kick out of the sample scenario provided. Who doesn’t love demon apes? You should totally write that up as a Savage Worlds one-sheet.