d100 vs. d20

In the older roll-under d100 systems, there was this one-to-one relationship between an ability score and the target number. The target number WAS your ability score, and if you raised your score, life got better.

Things flipped and became more dice-centric with d20 based games. Rolling high became the new thing, with the only exception being the older game Aftermath, a d20 roll-under system that worked a lot like the percentile systems. With the new d20 paradigm, rolling a natural 20 became the cool thing to do, and bang, that was it, you've made it in life.

Except when D&D 3 came along, and you had to confirm that crit.

Spoilsports.

It's funny the new versions of Car Wars followed GURPS and became a roll-under system too, because the math was easier than an open-ended roll. I still prefer GURPS as roll-under and Car Wars as roll-over though, there's no arguing with a 9+ to-hit on a heavy rocket and rolling double-sixes for a massive crit.

The original SBRPG was a 3d6 roll-over system.

For some things, I like roll-under and for others, roll-over. Percentile systems are classically roll-under, because the focus is different. I do not think all games have to be the same. One of the traps in being a game designer is taking your personal preferences and shoehorning them onto every other game you design or redesign.

A lot of D&D 3 came from the Magic the Gathering game. D&D 4's designers ported in some ideas from Iron Heroes and other games. D&D 5 borrows from yet others. It's not wrong to expand upon a good idea, but you need to weigh any addition against how it would fit within the original game.

One point being "healing surges" in D&D 4. D&D has classically been a game where a cleric can heal someone for X points of damage with the expenditure of a spell. In D&D 4, a cleric's healing ability used one of the target character's healing surges to heal them. If the target character was "all healed out" for the day, guess what, no heals for you. How does this make any sense with how clerics worked in the history of the game? It felt too alien, ported in from something else, and it wasn't how clerics felt their magic should work.

Porting in ideas doesn't always work or feel right within an established context. Nor does changing things too much to where things feel alien and strange given what came before. Just ask the Windows 8 team.

Can people "get" that they have to roll under a target number to succeed at an action? Yes. Does it feel good? Let me answer that with another question. How important is that feel-good of rolling high matter to your game, or does it just matter to you? Here's another thing to think about, rolling high invites open-ended dicing systems, because there are always larger positive and negative modifiers to throw around. Let's make our target number 913% and add +850% to the d100 roll!

With roll-under, you literally have 0% as your floor, and you do not want to go into negative numbers. You also get into this with d100, a 913% target number makes little sense other than being some obscenely high number you have to number-stack to reach. It means nothing percentage-wise. A 15% roll-under chance to-hit is simple, straightforward, and easily understood from anybody with a basic understanding of probability. A +10% chance to-hit modifier is also easily understandable.

Let's flip this around, and do this in a d20 roll-over system. Target number 21, and your modifier is a +3. What is your probability of success? Sure, rolling high is great now, but you have taken most people's understanding of odds and straightforward success and risk calculations out of the game. At least the average player's without some mental arithmetic. You get a further +2 to-hit, now what is your chance?

Hint: both examples are the same.

So as a game designer you ask yourself the question, do I want the "roll high" thrill, or do I want people to be immediately aware of their odds with roll-under? For certain types of games your answers will be different. For a hard-science sci-fi game, we may want percentage roll-under, because this game is supposed to be about science, odds, and the raw hard-probability chance of success. For a competitive game, like a PvP Car Wars game, roll-high is cool, since players around the table are now competing for the best and highest rolls.

If you choose roll-over, you are now opening yourself up to the unbalancing aspects of open-ended rolls. If you choose roll-under, your maximum ability scores and success chances are capped at or near 100%. There are trade offs to both approaches.

You tailor your game design decisions based on the types of behavior and feelings you want to promote at the table. One size or dice does not fit all, and roll-under or over is really a per-game decision. Never accept game design zeitgeist, and you make the right decisions for your game based on math, feel, and user experience.