[RPG Theory] Re-Rolls

[stevekenson]

It is resolved: a game mechanic that allows for a re-roll of the dice, especially if it involves expending a limited resource (uses per day, points, etc.) should never result in a worse roll than you started with.

Comments

[savageplanet.livejournal.com]

Amen. :)

If you're forced to take the second roll, no matter what, it ends up being a mechanic that will almost never be used except in hopeless situations where you need that "20" and anything else is a failure.

[bruceb.livejournal.com]

Intensely strong agreement.

[etherlad.livejournal.com]

I've never understood that mechanic.

"You can reroll, but you must take the second roll no matter what it is."

Sort of defeats the purpose of rerolling, no? What's wrong with "take the better of the two rolls"?

[adamjury.livejournal.com]

It depends on what the action *is* -- if you're picking a lock and the first roll fails, it doesn't matter if the second one fails worse, the door is still closed. "Going back" to the first roll won't help you.

If you're trying to do something less discrete like repair a car, then a crappy roll that's "good enough" would still take effect even after rolling a worse second roll trying to get the car *really* moving.

A lot of it hinges on how your gaming group [and the game] views rolls -- whether they are for specific actions ["picking the lock"] or entire events ["picking the lock, how close are the guards getting, is there a guard right on the other side of the door, and do we make too much noise while waiting for the lock to get picked?"]

[etherlad.livejournal.com]

Well, I view it less in the context of "make another roll," and more in the context of some power or ability, like an Action Point or whatever, which lets you "cheat fate" and thus reroll.

In terms of making a second roll after the first one fails, I agree with you. In character, you're trying again, with roughly the same chances of success or failure.

[adamjury.livejournal.com]

Well, I view it less in the context of "make another roll," and more in the context of some power or ability, like an Action Point or whatever, which lets you "cheat fate" and thus reroll.

You could then argue that the first roll -- even if it turned out to be the best of the two -- never even existed. ;-)

I'd generally agree that for "fun" purposes, giving the player the best roll of the 2 is the way to go. I don't think it meshes with every group's style of gameplay, though, nor every game.

[elissa-carey.livejournal.com]

Agreed. And most rerolls are looking for some kind of success, anyway, so mechanics that allow for that (like White Wolf's expenditure of Willpower for an automatic success, for example) seems the better way to go.

[nvdaydreamer.livejournal.com]

Agreed. We called such circumstances (where the reroll didn't help) an "ordained" result.

[rob-donoghue.livejournal.com]

I might submit an exception if someone is using a game with something other than binary (or simple progressive) results, but those are such an envelope case that it's a very small exception indeed.

[rob-donoghue.livejournal.com]

(though that includes the caveat that "Different" is not the same as "Worse", so even that isn't really a disagreement)

[shivianbalaris.livejournal.com]

I'd have to agree. A re-roll for me isn't a situation of doing the same task twice (like the lock pick example from adamjury) though I can see why people feel this way themselves. For me a re-roll is like going back a hair-second in time and doing something (hopefully) better. Hence only one option is used (whereas with a 2nd roll, both are "used" in succession).

I'd personally say that a re-roll uses the knowledge of the first split-second attempt that *didn't* occur (hence the re-roll) and you are trying it again. Same as seeing something play out in your head the first time, seeing your mistakes, and then when you "really" do it correcting for those mistakes. You might still do an equiviliant job, but the odds of doing a worse one are very low.

Which brings me to my thought of this: a re-roll should always be better, unless it's a 1 (on say a d20). This would enable a utter-failure rate for a situation where the first roll you tried *really* hard to see all your mistakes ahead of time, but then when you rolled the 2nd time (and did it "for real") you ended up over-fixating on the situation's problems *so* much that you totally tripped over your own two feet forgetting the basics of the situation and fucking it up royally.

Just my thoughts, though. ((grin))

[emessar.livejournal.com]

It also depends on the system and in some cases the specific task being attempted. If the system or situation is essentially pass/fail, then the only reason to use such an ability is if you fail. While the quantity shown on the die or dice may be lower, the result can never be worse than a fail which is equal to the original failed result. This changes if you are in a system or situation where results are more graded. If there is a chance for fumbles, criticals, or "degrees of success", there is a larger variety of situations where the re-roll might be used.

Possible alternate systems:

Reroll failed dice - In a dice pool system where each die is binary, reroll only the dice that didn't make the cut. The stronger your initial roll, the less incentive/effect there is for a reroll since there are fewer failed dice to reroll. This also caps your total to your character's natural potential.

Roll and add to the initial roll - In a dice pool system, such a mechanic might simply give you a set number of additional dice to roll. If the system is d20+number, you might get an additional d20 or alternately some lower die type (d10 maybe), to keep people from getting uncharacteristic results.

Alternately you could set the maximum result from the dice as being the maximum for a single roll so that if you rolled a second d20, the maximum of the two d20 rolls could never exceed 20. This would help cap a character to their maximum natural potential, much like rerolling failed dice.