Dice do many different things in our RPGs. They are a crucial element of the Game part of it. They model the un-modelable, all the little things that combine to determine what happens. They offer the illusion of challenge - we know the PCs will win. They take the story in unexpected directions. But do the same dice do it all?

D&D's d20 is a prime example of what I've taken to calling a "goblin die". You roll high, a goblin dies. You roll low, a goblin lives. No one doubts the eventual fate of the poor goblin. It doesn't matter if it's killed this round or the next. But it's still fun to roll those dice, just as it is fun to fight the scrambling goblins. Hence, goblin dice: good for determining the fate of goblins. Not so good for determining the fate of heroes, or worlds. They are terrible for anything important.

And, immediately, a caveat: unless your game is just about rolling dice. In which case, good for you, and carry on. But I want story from my roleplaying, in addition to rolling dice. D&D and its ilk don't warn you that it's not what they were made for, and so many people still try to use them for this, because they don't know any better. I didn't know any better, and it's only now, after a decade of GMing, that I'm coming to realise it, undoubtedly reinventing the bicycle.

So why do I say goblin dice are unsuitable for anything more important than killing goblins? They are swingy. They offer no nuance beyond the binary pass/fail. They are meant to be rolled a bunch of times over the course of killing goblins, not once to see if you can survive/convince the king to help you/save the world. Failing a roll is anti-climatic, but that's ok, contingent on getting more rolls soon. Missing a goblin is fine... but even in 4e it's getting less so. Partly because the turns take so long, and party because everyone, not just wizards, have limited resources they can expend on that roll (encounter or daily powers). It's not fun to miss and waste a rare resource through no fault of your own. It's even less fun to die this way. Skill challenges introduced in 4e try to mitigate this, with mixed success.

So many issues I have with D&D stem from this simple observation, so much of GMing advice and wisdom accumulated over the years is aimed at circumventing it, that writing it feels like stating the obvious. And yet I haven't really considered it from this angle, so, perhaps, neither have you.

"Never have a single roll stop the party from progressing through the adventure." Social skills, unusable for anything meaningful. Save-or-die. GMs "cheating" to improve the game. Goblin dice are at the root of it all.

That last one deserves more attention. GMs and game designers have dedicated so much effort to making the games do what they want despite the goblin dice. In a certain class of game, you want the good guys to win in the end, not the guys that rolled higher. To put it another way, genre or GM or even players often know better than dice, especially goblin dice.  Of course they do!  And there are other games that acknowledge that. Yes, I'm talking about narrative control and players-as-authors or players-as-directors, and other Big Model-esque stuff.

While this post is not about solving the problem, merely about stating it, I'll offer a couple of ideas. The obvious solution would be to steal a conflict resolution mechanic of your choice, and use it in place of goblin dice where appropriate. Another idea, that has just occurred to me today, is to replace a single goblin die roll in non-goblin situations with a triple roll of the same dice. Normally, you'd use the middle roll, thus creating something approximating a bell curve, without actually changing any numbers in your game. You can opt to use the high roll as a "yes, but...", a success with a complication. And perhaps using the low roll (and still succeeding) would be a "yes, and", allowing for extra shiny outcomes. Or you could wager upon succeeding with the low roll. Not yet sure if this idea is any good, though.

There's more to be said on the topic of transplanting different narrative control approaches into "traditional" rpgs, but that's what next post is for.