... now with 35% more arrogance!

Wednesday, July 10, 2013

Criticals and Fumbles

There was some discussion about the criticals and fumbles on the modified Fantasy Trip mechanic post that is worth breaking out into its post, because it applies to more than just rolling Xd6 for ability checks. The features that SAROE and faoladh noticed in "roll Xd6 less than ability score, exact match = critical, all dice matching on failure = fumble" were that criticals become less likely for the highly skilled, and fumbles become less common for more difficult actions.

Now, the criticals becoming less likely for those with higher skill is disturbing, although maybe not as disturbing for me than for others, for reasons that may become clear in a bit. A possible fix would be to go back to the version of this mechanic I posted several years ago on a Microlite20 forum: all dice matching = special, a special that is also a success is a critical, otherwise it's a fumble. Another alternative would be the revision I later made: more than half the dice are 6s = special. This, however, may make criticals and fumbles a little too common.

But I'm not really worried about fumbles at all. This may be because of a crucial difference between the way I see criticals and fumbles, compared to everyone else. I see this as phenomenal accidents, good or bad, rather than being related to skill or difficulty. In my view, you can't "learn to do criticals", you can only learn to be more likely to succeed. Similarly, a fumble or critical failure is not, in my view, the fault of the individual, and trying something beyond normal human ability just makes you more likely to fail, not more likely for things to go horribly wrong. What increases the chances of things going horribly wrong is increasing the number of tasks (rolls) that must be done to succeed. You can't learn to avoid the unpredictable.

Ideally, the chance for criticals and fumbles should remain constant. That's one plus of the d20 method (or BRP d100 method) with its flat "roll of 20 = fumble, exact match is critical success". The odds are always 5% for either fumble or critical success on that d20 method. But if the odds change at all, they should probably decrease for criticals as you get more skilled, because you are improving your precision. And I'm thinking really difficult tasks should decrease the chance of a fumble, too, for more or less the same reason: they increase the predictability of failing right from the start, so you never even have a chance for things to go truly, horribly wrong.

But I have the feeling this is something most people just aren't going to see the same way.


  1. I quite like 'all dice the same = special, success or failure indicates which way'. It keeps the ideas of critical success/failure coming from unpredictable chances, but also models the fact that professionals will usually handle those curve balls better than amateurs. The expert takes an unexpected twist and turns it to his advantage, while the tyro is apt to freeze up and watch things fall apart. This will be the case if the character rolling has more successes than failures in the set of 'all dice showing the same pips.'

    It also handles the fact that the more complicated a situation already is, the less likely it is that a single new element is likely to drastically change things - really difficult situations aren't as prone to criticals in either direction.

    My only concern is that the dice-removal for experts would make criticals and autosuccesses come up for them far too often. Perhaps allow them to ignore additional dice for difficult tasks, or give them a 'reroll X dice' mechanic instead. That would keep both failure and critical results in play while still giving them insulation against it.

  2. Have any successful score with a roll of triples be a "no math" critical.
    Higher skill characters get a greater chance of success. Really difficult situations will create criticals slightly more often as more dice = greater difficulty but this has a drame enhancing trade-off and creates a "fun factor" for heroics where crasy outlandsih heroism can more likely have outlandish results.