Writing a ttRPG - Wyverns & Warrens (Pt 4 - Target Numbers) (Patreon)

(This article is not part of the content of this blog that is covered by the OGL 1.0a.)

I'm designing Wyverns & Warrens (WyvWar), an effort at a short, easy-to-learn, easy-to-play, fast-playing fantasy ttRPG that still has a depth of options. And I am doing that in weekly installments, exclusively for my paying Patrons!

Having decided I want a single core game mechanic, and I want it to be based on dice, and I want it to take only a single die roll in most cases, WyvWar uses 2d10 (which give a pyramid number distribution) as the core die mechanic... it's time to talk Target Numbers (which I currently plan to write as 'Target#' in the game). And to do that, I need to talk about where the decision to go with 2d10 came from.

I settled on this because flat distributions are way too random for me. For example, if a core mechanic uses a single roll of 1d20, then anything possible happens at least 5% of the time. Further, with each individual role, every result is as likely as any other. At the other extreme, bell curves are too unlikely to ever present the extremes of their range. On 3d6, for example, the highest and lowest results (3 or 18) each only happen less than half of 1% of the time. So, I want a pyramid distribution.

(An aside -- I ran the TSR RPG Alternity for a year or two in the 1990s. I loved it. My players did not. On a system where you wanted to roll high, it was possible to have die rolls of 1d20 + 1d6, or +1d8, +1d12, and so on. But you could also end up rolling 1d20 - 1d20, or even 1d20 - 3d20. In all the time I ran, no one ever succeeded on any check once they were subtracting dice from their d20 result. And when someone got a result of -10 or lower, even if it was mechanically no worse than failing a die roll by 1 or 2, the players FELT worse. As this system was also used for initiative, as players declared 'I go on 11,' and 'I go on 17,' I once had a player declare 'I go on DISMAL'.")

Taken from anydice.com (which is in my list of game design links I keep on my desktop), here's a table showing the chance that a roll of 2d10 will product a value of at least the total listed to the left.

This is the starting point for all my math. I like systems where you have to roll a given number or higher (so you always want to roll high), so my Target# will be a value that you want to meet or beat with your 2d10 + bonus. Using the chart above, I can figure out what Target# gives what chance of success when rolling 2d10 + 0. For example, if I want a single 2d10 die roll to have a roughly 70% chance of success, then I want the Target# to be 9. That also means if the Target# is 9, someone with a +2 bonus will succeed in one roll 85% of the time (since that's the odds of rolling a 7 or higher, and they then add their +2 bonus).

(Looking for an ambush? Casting a spell? Crossing a rickety bridge? In WyvWar, those are all an action check, all handled the same way -- rolling 2d10 + bonuses against a Target#.  Artwork (c) Dean Spencer, used with permission. All rights reserved.)

This directly impacts the size bonuses will be within the game, as will the fact they have a variable impact on the % chance of success, depending on the Target#. For example, a +5 bonus to 1d20 with a target number system gives a +25% bonus for any roll when the full range of d20 results matter. But 2d10 changes the value of the bonus depending on the Target#. With a Target 9, +5 gives a +25% chance of success, but against Target 20 it gives only a +20%. That limits the impact of bonuses at extremes, but not by so much players won't generally know how good a +2, +3, or +4 bonus is.

This base math means I can start building structures that calculate Target#s, and design systems to give characters bonuses... or penalties. The base math isn't game rules yet. It is, instead, the mathematical framework upon which those rules can be built. With an understanding of how likely any given dd10 + bonus roll is to succeed, I can start planning out how and when bonuses are gained, and what they apply to. And that begins with a simple question -- what do I want the odds of success to be for a typical character, using their typical skill, attempting a task of typical difficulty? And my answer is...

79%

Look, there ARE tons of essays and studies and theory pieces about how often a character should succeed at a typical task in a game. Those aren't all focused on ttRPGs, but they'd be great reading for someone who was building a ttRPG engine and wanted guidance on where to set the baseline success level. I've read a lot of game theory on this topic over the years. But I didn't look up any of that material when making this decision for this game.

Instead, I went with my gut.

I'm looking for a game where numerical bonuses are kept low because I believe that speeds play for a lot of people. Also, because I personally have ended up playing a great ttRPG session until 3am, at which point my brain is mush and simpler math speeds things up for me. Also, I want a single difficulty chart that applies to characters at all stages of play. If a GM would describe a task as 'Demanding,' it's Target# 12, regardless of how experienced the characters are. That means that as character bonuses improve, they become increasingly likely to succeed at more difficult task. And I'm fine with that. There are many ways for a character to advance beyond increases bonuses of course, but for the kind of game I want to write, having very experienced characters just be BETTER at more difficult tasks is exactly what I want.

The following are nine levels of task difficulty, from "Nearly Automatic" to "Nearly Impossible." Those names are subject to change, but having this chart will allow me to compare the effectiveness of various characters and NPCs at tasks of various difficulty. It also lets me see what the game is claiming is "typical" vs "heroic," which is important when figuring out defense value in combat, for example.

Each entry gives a Target#, its official difficulty level, and then the chances of someone hitting that on 2d10 with first a +0 bonus, and then a +2 bonus.

04 Nearly Automatic (+0, 97%; +2, 100%)06 Trivial (+0, 90%; +2, 97%)08 Easy (+0, 79%; +2, 90%)10 Typical (+0, 64%; +2, 79%)12 Demanding (+0, 45%; +2, 64%)14 Arduous (+0, 28%; +2, 45%)16 Heroic (+0, 15%; +2, 28%)18 Super-Heroic (+0, 6%; +2, 15%)20 Nearly Impossible (+0, 1%; +2, 6%)

That table also lets me define a critical success and critical failure system, if I want.; For example, I am leaning toward saying for any task that is Super-heroic or easier, a total result of 20 or more is a critical success. That makes characters with higher bonuses more likely to be critical successes. But unlike a system where you get a critical success because your result was 10 higher than the Target#, just because some task is trivial (hop over a 2-foot-wide ditch), characters won't all be producing critical successes. it ALSO means that you can't get a critical success on a nearly Impossible task... and I'm fine with that. And if I want critical falures, I can obviously do the same thing in reverse.

ALL of this is subject to change, of course. For example, maybe all Target#s will go up by one, so a Nearly Impossible task is Target# 21, meaning you CAN'T succeed without some kind of bonus. But this is a solid framework for me to start writing some actual 'rules' of the game. If this framework ends up unable to hold up the elements I want my game to have, I can tweak it, build separate subsystems, or even scrap it and start over. But I find having a solid idea of core mechanics and probability a great place to start when building a ttRPG from scratch.

Next Week: Damage and Benchmarks!