2013-10-21 : The Magic Trick: Otherkind Dice

The magic trick, recall, is to make combat feel genuinely risky without killing PCs off all the time. Recall also that while it's probably not possible to have your cake and eat it too all the time, the magic trick is to have your cake and eat it too most of the time.

Otherkind Dice
In Otherkind*, a combat roll is a pool of 5d6, and it decides four things. It decides (1) whether you advance toward your objective, (2) whether you hurt your enemy, (3) whether your enemy hurts you, and (4) whether your enemy hurts any of your friends and allies. Each of these is on a scale, like so:

Do you advance toward your objective?
1: You lose ground.
2-3: You hold ground.
4-5: You gain ground.
6: You seize your objective.

Do you hurt your enemy?
1-3: No.
4-5: Yes.
6: A lot.

Does your enemy hurt you?
1: A lot.
2-3: Yes.
4-5: No, but your enemy puts you off-balance or on the defensive.
6: No.

Does your enemy hurt your friends and allies?
1: Yes, badly, all who are exposed to danger.
2-3: Yes, but not badly, or only a few.
4-6: No.

To make the combat roll, you pick up 5d6, roll them, throw away the lowest number, and then assign the remaining four numbers one each to the four categories of outcome.

For example: you roll 1 1 3 4 6. This means that you throw away the first 1, and then choose how to assign the remaining 1, the 3, the 4, and the 6. Maybe you seize your objective with the 6 and hurt your enemy with the 4, but that means that now you have to choose whether your enemy hurts you badly with the 1 and hurts your friends not-so-badly with the 3, or vice versa. Make sense?

So here you are holding the dice in your hands, right before you roll. The roll is a genuine risk and it feels like one. The range of outcomes possible with this rolling method is pitched in your favor, what with discarding the lowest die and all, but it's not safe, and when the dice come out low they're absolutely unforgiving. It's possible that you'll get all high dice, and that's what you're hoping for. It's most likely that you'll get a mix of high and low dice, so you're thinking about what you're risking - particularly, what you're willing to sacrifice if it doesn't go purely your way, and what you'd really prefer not to sacrifice. And you know that it's possible that you'll roll 1 1 1 1 1, it's always possible, so you're hoping you don't!

The magic trick is the same as in Dogs in the Vineyard, even though the mechanism is very different: before you roll you're focused on what you're risking, how bad it can go, how much you might have to lose, but after the roll it's not as bad as you feared. Even on a terrible roll**, like a 1 1 1 1 2, you get to mitigate the disaster! Though the costs are high, you get to preserve a measure of whichever thing you value most. You still have a say. It's a bloodbath for your side and your enemy is unharmed, for instance, but at least you held your ground.

I want to emphasize that the magic trick isn't that terrible outcomes are unlikely. Terrible outcomes do happen, and bad outcomes are quite common. These dice are unforgiving! No, the magic trick is that when a bad outcome happens, or even when a terrible outcome happens, you get to make it the bad or terrible outcome you can live with.

1. On 2013-10-21, Vincent said:

This is part 2 of 2 of my answer to George's question down here. Part 1 is 2013-10-17 : The Magic Trick: Damage in Dogs in the Vineyard.

* There is no such game as Otherkind that actually uses this set of outcomes. It's just a good example.

For real live games using this dice method, check out Psi*Run, Mobile Frame Zero, and/or my most casual possible writeup here: Salt River. There are others!

** Yes, there's always the possibility of the 1 1 1 1 1 roll, but it's less than one fiftieth of one percent.


2. On 2013-10-21, Josh W said:

Interesting rule! Looking at the probabilities, this is actually a quite friendly one, despite the way it shows you all these negative possibilities:

Assuming you want to gain ground and hurt your enemy, and you're willing to be hurt yourself if you can avoid your friends being in danger.

The last criteria can actually be handled outside of this dice mechanic; attack the problem in such a way that they are not put in danger but you are.

Doing this effectively means you effectively drop the bottom two numbers, which means you're looking at the 3rd highest number among those dice.

The probability of that being higher than 1 is 96%, and the probability of it being higher than 3 is 50%.

So if you go ahead of your allies to protect them from danger, you have a 50/50 chance of being able to avoid being harmed, and a 60% chance of achieving your objective while doing it.

Having allies in the mix changes things dramatically, the probability of you all getting out unscathed is only 19%, although conversely, if you don't care much about their welfare, the probabilities are the same. This suggests to me a situation in which heroic figures brave danger alone, whereas less heroic characters try to get their community to shelter them.

Also, because of how in dangerous situations you can dump bad dice into your goal's outcome, it suggests that when in danger it's useful to be ambitious in unexpected directions, seeking to make something of the situation, as then that roll will be less likely to negatively effect the other outcomes if you dump it!

On the other hand, the objective outcome interacts a little weirdly with the off balance/defensive harm result. I'm guessing that acts like some kind of penalty dice for future conflicts? It's odd that you could gain your objective, say capturing and disarming your foe, and yet also be on the defensive.

I like the idea of changing the 4-5 harm result on damage to encourage seeking help in some way, so that there's a bit of tension between going off alone and getting backup when you need it.


3. On 2013-10-21, Vincent said:

Yep! All true.

I made these four categories up off the top of my head, just as an easy example. They're more-or-less infinitely adjustable. In Mobile Frame Zero, for instance, you roll for movement, defense, attack, and surveillance/communication between units. In Psi*Run, you roll for your goal, for whether you trigger any memories in yourself, for whether the sinister forces pursuing you come a step closer to catching you, and then for other things you swap in and out at need, like whether you're hurt, whether you're captured, etc. In Salt River we made a new and unique set of categories up for every new situation.

You can be more creative with the scales. In Psi*Run, One scale is 1 2-3 4-6, one is 1-3 4-6, one is 1-2 3-4 5 6, one is 1 2-3 4-5 6, and so on. You can really fine-tune the outcomes you want people to have to choose between that way. Like in the above example, a 4 is just as good as a 6 for keeping your friends unharmed, but for everything else, a 6 is better than a 4. This creates more opportunities for interesting choices than doing everything on the same scale does.

You can fiddle a lot of different ways with the dice, too. You can lose the spare die if you're disadvantaged (as you say), or not have a spare die in the first place. You could add a "no die" result to each of the outcomes and roll fewer dice than you need! You could use dice of different colors that you can assign only to some of the categories. All kinds of stuff.


4. On 2013-10-21, pigeon said:

One cute thing about this system that you just made up, by the way, is that on a 1 1 1 1 2 you can still choose one goal to partially achieve—unless it's hurting your opponent. Controlling the 2 and 3 results becomes a major lever here for expressing how amenable specific problems are to risk management.


direct link

This makes...
VB go "+1"

5. On 2013-10-21, Judson said:

There's an different question though: how much impact does how the categories are presented matter to the outcome? If, for instance, you arrange the categories in a grid: 1-6 down the side, and categories across the top, and boxes for each outcome so that a player can immediately see which results are e.g. more amenable to risk management, I think you kind of expose better and worse applications of the general trick.

More generally, if the text includes strict percentages of outcomes, is the overall magic trick hurt? In other words, does knowing how the trick works destroy the illusion? I kind of think it doesn't, so long as there actually is a real chance of real harm. 20 = instant death has to be a result, or the trick doesn't work.


6. On 2013-10-22, George said:

Thanks again Vincent. This example was also very interesting. I had to read it a few times, and also went back to your earlier description about the Dogs mechanic. In otherkind, it sounds like:
player: this is how i instictively react.
GM: this is what happens as a result.
A question I have regarding Otherkind, which I realised applies to Dogs as well, is: what signifies the end of the fight? It sounds like we're leaving behind hit points as the flag notifying us that the combat is over (which is probably a good thing). Would it be correct to say that this is part of the trick?

What I'm noticing in both cases is that you're creating an abstraction for an exhange between combatants (or even the entire combat).
I guess that escaping the confines of having to narrate every single result (and it's physical effect) after every single blow, allows you to accentuate certain elements (e.g. what's at stake) which is vital for the objective of the trick. You said: "No, the magic trick is that when a bad outcome happens, or even when a terrible outcome happens, you get to make it the bad or terrible outcome you can live with." This made me realise what's the secret of "failing forward" in combat. In combat, it always seemed to me that failing forward would be contrived.
The examples you gave achieve two things:
1. you ensure that the outcome is not a physics/reality/real combat outcome but always a "Story outcome" first and foremost.
2. the players don't feel protected by the hand of God.
And if i'm reading what you said correctly you do this by:
-emphasising what they risk
-you hand over some control to the players in terms of what they can afford to lose i.e. what is the next story outcome that will move the story forward. "Indy, will you allocate that die to stay on the tank? Or not?" So it's not the GM desperately looking for "the angle". He lets the players sweat it - this is how it was always meant to be >:-)
-you don't cheat, but keep the possibility of death there, as Judson pointed out.

Judson: I fully agree with what you said "20 = instant death has to be a result, or the trick doesn't work." The fact that this possibility is there, and the players KNOW that you can't change the outcome gives validity to all of the other outcomes. And when you have a system like Vincent's when it's not immediately obvious that the odds are stacked in the players' favour, then you're in business!


7. On 2013-10-22, Vincent said:

Thanks, George! You've got it exactly.

In Otherkind, the fight's over in the one roll, unless one party or the other decides to push it into a second roll. I imagine that if I were to use these precise rules I'd add an accumulating advantage or something, so that if you push it and lose again, the odds of winning it on a second push dwindle to desperate, and the odds of winning it on a third push after losing three times are nil.

Dogs in the Vineyard has a whole system for how fights escalate and how long they last. You have the game, I think you said? I can explain it if you want or you can just read up on it.

But your insight there is very, very good: in neither game does the fight end because somebody's run out of hit points. Neither of these fights are fights to the death. Something else decides when the fight ends and who's won, and that's an important part of these rules.


8. On 2013-10-23, Josh W said:

That is great! And it suggests a fix for AD&D actually; every time you get hit, roll damage as normal in hp terms then put the dice to the side. When your hp hits zero you surrender/give up etc.

Then at the end of combat use those set aside dice to roll for the fallout like in dogs!

Every healing spell cast removes one of the dice under its size as well as giving hp back.

You can also cause damage without leaving the dice around as damage via taunts and threats, non-lethal attacks etc. These use exactly the same rules as normal.

Then random poor conditions when travelling can also cause hp loss without causing damage, and good conditions can cause it to replenish as given in the normal rules. Also, once fallout has been worked out, healing spells also work directly on injuries according to their effects ie. cure light wounds gives you hp back, but doesn't deal with serious life threatening injury, whereas cure serious wounds does.

Of course, this system would only work when fighting intelligent creatures; loosing the will to fight when facing a giant monster is no great difference from the old interpretation of hitpoints!

I think it works though, so long as you set up the maths for the new fallout table right.


9. On 2013-10-23, George said:

Vincent thanks for clarifying about the abstraction in Otherkind. About Dogs, sure I'll check it in the book in more detail. It was good to hear your thoughts on the importance of moving away from a (story-agnostic) number.

Josh W: these all sound great. They all reinforce the genre. Rather than trying to force certain things into the narrative to try and patch combat. (things such as healing potions becoming an everyday item, the cleric class and the resurrection spell) Now if someone invented time-travel to tell my teenage self. Oh well, better late than never :)

As a general comment for these last two threads: I'm really looking forward to running my next game using these principles and rules. Thank you!


RSS feed: new comments to this thread