Thursday, September 30, 2010

Transparent Outcomes

Apocalypse World brought another bit of game design tot he to of my mind, one that I've always instinctively liked but never given much though to - transparent outcomes.

Games with transparent outcomes are one where the player can roll the dice, look at them, and know roughly what happened. At it's most obvious, success of failure can be read off the dice, but often it's a bit more nuanced than that. But before we get to that, let's step back a bit.

What makes transparent outcomes different is that there is no element of secret information in the exchange. In D&D, you roll a d20 and if you rolled high you probably hit, but you don't actually know for sure. The GM knows the number you need to hit, and that is revealed to you after the fact. If, as is sometimes the case, you knew the number you needed to hit before you rolled, then you would be able to immediately proceed, either mechanically (with rolling damage) or descriptively.

That would be a transparent outcome, and some game straddle the line on this with the assumption that the GM will always communicate the target that needs to be hit. However, in practice I have seen that to more often be a function of table best practices, and it is usually subject to specific vagaries. For example, a GM who usually shares target numbers might obscure them for perception checks.

For a game with truly transparent outcomes, the resolution is apparent in the dice. For example, in the Storytelling system, you only need one success to succeed at a task. This gets muddled somewhat with contested rolls, but as a baseline it's pretty spiffy and rewarding to players. For me, it's probably the single biggest mechanical improvement from oWoD (which often required a variable number of successes) to nWoD.

But where this gets interesting is nuance. Apocalypse world, for example, adds a third tier. Roll 2d6+stat (usually a value form -1 to +3) - on a 6 or less you fail and the GM gets up in your grill. On a 10+ you succeed. But on a 7-9 you get a complicated success. You might succeed with a cost or consequence, or you might manage only a partial success. This is nice, but only the beginning. Talislanta had Mishap, Failure, Partial success, Full Success, and Critical success.[1]

I think the finest grain probably came from the Dying Earth system (now in generic form as Skulduggery) where the range was:
1 - Total failure
2 - Failure
3 - Partial Failure
4 - Partial Success
5 - Success
6 - Exceptional success

Corresponding to a single d6 roll

Now, the fact that these outcomes are transparent does not mean that they do not require interpretation. Some of these may have specific mechanical meaning (Such as doing a certain amount of damage or offering certain choices) but even beyond that, the GM still has a role in spinning forward the outcome.

Yet despite that, what this does is remove a step in the player/GM communication. THe dice hit the table and there is a shared understanding regarding the outcome which allows play to proceed more smoothly. If you're very interested in the back and forth between players and GMs to proceed without a hitch, then transparent outcomes are probably something you should consider.

But they're not suitable for every game. While "secret information" may sound sinister, in reality it is a way to increase uncertainty, which is important in games where the sense of risk is important. When dealing with smart, math-oriented players, secret information is what keeps everything from degrading into a pure statistical exercise. As a rule of thumb, if the actual play of the game is part of the fun (rather than just the necessary vehicle to move back to the play you consider important) then transparent outcomes are less useful. To use a concrete example, much of the fun of D&D is in the actual engagement with the fighting mechanics. Lots of rolling, lots of uncertainty. Secret information makes things more exciting.

In contrast, in a lighter game where the mechanics are engaged for just moment, to answer a question as it were, there's a lot to be said for transparent outcomes.

Now, there are more variables than this. Margin of Success, for example, throws a whole other wrench into this thinking, so don't go thinking that the choice is black and white. But when the time comes to design your own system, take a minute to think about whether or not transparent outcomes are for you.

1 - Many games actually have 4 outcomes - Critical failure, failure, success and critical success - even if they don't actively view it that way.


  1. I think this is one of the reasons why the Basic Role-Playing system worked so well. Because your roll to succeed was based effectively on the ability of the character, not that of the opponent. Simple gratification, you roll the dice and know whether the action succeeded or failed.

    Of course, the action can still be blocked or parried or dodged (in the case of combat) or equalled (in the case of a contest), preventing victory from being achieved, leading to long drawn-out contests when both contestant/combatants are either unskilled and flailing around madly, or highly skilled and deftly turning aside each other's blows.

    But as soon as the dice hit the table, you knew that you had achieved something.

  2. This is also why Cortex throws some people, because it's a statistical maze. Not everybody likes dice pool games, and a dice pool game with different sized dice AND opposing dice rolls AND other things... instant headache for a statistician.

    But then again, those are the kinds of games I love. :)

  3. It took me a while to crawl back with the way-back machine, but I found a couple of posts where I did some thinking about this. The first is more interesting than the second, which is more of a rant. :)

    The TL;DR version is that I strongly prefer more transparency at the table. Players should have a reasonable opportunity to know how tough a task is going to be. This becomes even more important when the player has expendable resources they can use to improve their chances of success.

    (Put another way: if the system has a fate point mechanic that wasn't tied to the success check, it would be less important that the player know exactly how tough the task was. In other words, if spending a fate point turned any failure into a minimal success, is it really important that the player know the exact die roll they need to succeed?

    I'm reasonably certain I hit this in another post that I couldn't find, but two other ideas I've generally applied to tweak this:

    1. Transparency applies to numbers for a player's MINIMUM SUCCESS, but not necessarily for an extraordinary success. In other words, if a Mediocre success gives them some information, and a Great success gives them a lot more, I may not specifically reveal anything about a Great success.

    2. Sometimes it is okay to lie to the player. They may have misread the situation. The bridge looked like it might be tough to get across, but halfway across the bridge, the supports that the villain sawed through finally give way.

  4. I'm starting to toy with the degrees of success thing myself, and it seems to fit like a glove to systems using variations of the Otherlind kind.

    I particularly like the scale presented in the new Dr. Who game in which an exceptional success/failure is read as "Yes/No, and..." and partial success/failure is read as "Yes/No, but..."

    - TheRoleplayer

  5. I'm starting to toy with the degrees of success thing myself, and it seems to fit like a glove to systems using variations of the Otherlind kind.

    I particularly like the scale presented in the new Dr. Who game in which an exceptional success/failure is read as "Yes/No, and..." and partial success/failure is read as "Yes/No, but..."


Note: Only a member of this blog may post a comment.