Inspired by this comment from Lasgunpacker over at Jeff’s Gameblog, here is a simple rule for combat stunts.
Say you want to perform a combat stunt that is high risk but also high reward (such as doing a backflip over charging goblins and kicking them over a ledge). The referee will give you a difficulty modifier based on the specific circumstances (say, -2) and if you accept the risk you make an attack roll. If you hit, your stunt succeeds (exact effect is by referee ruling). If you miss, it counts as a fumble, as if you had rolled a natural 1. Either roll on a fumble table or do whatever it is your group does for fumbles.
Mathematically, this compresses the probability distribution. Everything above the target number is something like a critical hit, and everything below is a fumble. It is the combat equivalent of “all in” which captures exactly the flavor that I think a stunt should have. Also note that a penalty to the attack roll need not always apply as the increased probability of a fumble may sometimes be enough of a difference from a normal attack.
Many systems use auto-success mechanics for stunts (e.g., luck or action points) which create exactly the opposite of the desired narrative effect: certainty rather than suspense. Another common game mechanic is to call for something like a dexterity or athletics check. I also consider this suboptimal because it emphasizes ability scores and thus character engineering. My dislike of emphasizing ability scores should by now be well established.
I like it, because it allows the players to come up with something zany, the DM to say “yes, but”, and seems like it would be quick to adjuticate as well.
I’ve used a similar rule for a while. None of my players went along with it and there weren’t many stunts attempted in games.
Thus my new rules was a reaction to that and making stunts less risky to perform overall (well, most of the time), but still not always going to succeed.
Interesting. I will be curious to see how my players react. They tend to not be overly careful, but you never know. I probably won’t put it forth as a formal rule, I will just explain the consequences the next time someone tries something crazy and see what happens.
Failure needn’t automatically equate to a fumble, but the consequences of failure ought to be spelled out in advance. Perhaps like so: “of course, the danger is that if you fail to execute the flip, you may fall prone behind them and they get a +2 to attack you on their turn.” These consequences can be negotiated as well. Well that’s how i do it at least.
I like that method too. In this case, part of what I want to get out of a stunt is an increase in suspense, so I don’t think exactly spelling out the consequences is appropriate, though I do often employ that method in other situations. For example: the edge of the precipice is slippery with waterfall spray. If you go near the edge you will have a 2 in 6 chance of falling.
Perhaps a series of fumble tables for different contexts might work well, though it would be more work.
One thing you need to think about in a stunt system is the difference between a flourish and a bonus.
Is back-flipping over the goblins in line with the kind of tone you want for your campaign (e.g., swashbuckling, high action)? If not, then you want to discourage stunts overall. No matter the type, applying penalties is appropriate.
Does it actually provide the PC with any bonus over just running through the goblins? If not, then it is just a flourish. It’s just descriptive text. If you want high action with crazy stunts, you can either straight up allow it with no penalty (but also no additional effects), or you can actually provide a bonus/reward for excellent description (Exalted is original example of this style, but most action point systems recommend giving out points for cool descriptions).
If the stunt is intended to provide a bonus, whether that is avoiding attacks of opportunity, allowing for an unusual sort of attack (e.g., pushing them off the ledge), or just stunning the enemy with shock and awe, then it needs a risk/reward adjustment. The PC is looking to get more than usual mileage out of his action. I think that your system above is a good one. I especially like the idea that trying a stunt is more likely to cause a fumble, rather than simply being more likely to cause a miss.
I think some systems use the idea of stunts as a replacement for powers or abilities: something the character can “do” which is like a resource that the player can tap. That’s not what I’m going for here.
In general, I would expect that a stunt would have some dramatic mechanical effect in order to compensate for the much greater risk. For example, an attempt to run the goblins through would probably be a standard attack roll followed by standard damage. There is a certain ceiling to the possible effects. Your third paragraph is exactly it. I can imagine many stunts leading essentially to auto-kills or being able to effect many more opponents at once than is normally possible. My backflip example is unfortunately not the most creative, but in that case maybe the intent is a domino effect where several goblins are forced over the ledge rather than a single goblin being at risk of some damage.
Right now I don’t actually have an intended tone, though I think it might be interesting to try to work tone into the mechanics like this. I’ve been letting my players choose the tone by their actions. The only incentive system that I am explicitly using is XP (given for treasure, defeating monsters, and several metagame things like writing session reports).