Contests (Opposed Rolls)

[chrisbender]

The way the OP describes is how GURPS handles contests. This is also how I would assume it works in TFT.

From GURPS Lite:

Each competitor attempts his success
roll. If one succeeds and the other fails, the
winner is obvious. If both succeed, the winner
is the one with the largest margin of
success; if both fail, the winner is the one
with the smallest margin of failure. A tie
means nobody won (in the examples above,
both fighters grabbed the weapon at once,
or the knives hit the same distance from the
bull’s-eye).

[platimus]

Originally Posted by chrisbender

The way the OP describes is how GURPS handles contests. This is also how I would assume it works in TFT.

From GURPS Lite:

Each competitor attempts his success
roll. If one succeeds and the other fails, the
winner is obvious. If both succeed, the winner
is the one with the largest margin of
success; if both fail, the winner is the one
with the smallest margin of failure. A tie
means nobody won (in the examples above,
both fighters grabbed the weapon at once,
or the knives hit the same distance from the
bull’s-eye).

I came to quote the rest of the rule and suggest that you treat both contestants failing as a "tie" but this seems to be the most sensible thing to do.

Another alternative is handle contests with 2d6+stat+skill. Highest total wins.

[chrisbender]

Originally Posted by platimus

I came to quote the rest of the rule and suggest that you treat both contestants failing as a "tie"

I like the idea of treating both failing as a tie. Feels more correct.

[JohnPaulB]

Originally Posted by chrisbender

I like the idea of treating both failing as a tie. Feels more correct.

Perhaps treat both failing as BOTH FAILING. Both throw the knives and miss, both stumble and fall, both make the audience boo.

EDIT: Oops, that is what GURPS suggested. Sorry about that.

[Skarg]

Again, I think it's really valuable to appreciate the difference in the chances of "upset" results between counting auto-success/failure, or not, and learning to think about the situation represented and what is appropriate.

example: Gambling. Most games of chance retain more or less of a chance element no matter how different the skill levels are. In this case, I would at least allow auto-success failure at the usual thresholds, or likely even increase the odds of pure luck deciding the outcome of a simple game.

example: Arm wrestling. This is a pretty direct contest where there is very little that can happen to allow a much weaker person to defeat a much stronger one. Unless it is a very drunk and chaotic contest, ST 9 should not beat ST 18 even 8-9% of the time (as would be the case if you allow auto-success/failure to count). In this case I would not allow auto-success or failure to count, reducing the odds 9 beats 18 to under 1% (probably meaning the 18 was caught by surprise or slipped or something that should not happen often at all).

[Oneiros]

Originally Posted by JohnPaulB

Perhaps treat both failing as BOTH FAILING. Both throw the knives and miss, both stumble and fall, both make the audience boo.

EDIT: Oops, that is what GURPS suggested. Sorry about that.

It's actually not. The GURPS guidelines quoted above is the contestant with the smallest margin of failure wins, i.e., succeeds. Not sure if I agree with that.

[platimus]

Originally Posted by Skarg

Again, I think it's really valuable to appreciate the difference in the chances of "upset" results between counting auto-success/failure, or not, and learning to think about the situation represented and what is appropriate.

example: Gambling. Most games of chance retain more or less of a chance element no matter how different the skill levels are. In this case, I would at least allow auto-success failure at the usual thresholds, or likely even increase the odds of pure luck deciding the outcome of a simple game.

example: Arm wrestling. This is a pretty direct contest where there is very little that can happen to allow a much weaker person to defeat a much stronger one. Unless it is a very drunk and chaotic contest, ST 9 should not beat ST 18 even 8-9% of the time (as would be the case if you allow auto-success/failure to count). In this case I would not allow auto-success or failure to count, reducing the odds 9 beats 18 to under 1% (probably meaning the 18 was caught by surprise or slipped or something that should not happen often at all).

I mostly agree with this and is actually what I had in mind. For something like a poker contest, I'd use the 3/IQ vs 3/IQ method (modified by skills). For something like the arm-wrestling, I'd use my 2d6+ST vs 2d6+ST method. As an autonomous, self-aware being, I'm fully capable of deciding on the fly which method to use, what constitutes a critical success/failure for either, and what effects might come of a critical success/failure...or if any of that is even important.

And sometimes a tie is great and need not be resolved immediately. Suppose you had a two guys wrestling over control of a turnstile that controls whether or not a drawbridge get raised. Each time they tie, the bridge stays where it is and might give the invading side time to cross it.

[Skarg]

Originally Posted by Oneiros

It's actually not. The GURPS guidelines quoted above is the contestant with the smallest margin of failure wins, i.e., succeeds. Not sure if I agree with that.

In GURPS, there are rules to figure out the odds of hitting a target of any particular size with any particular weapon, so a GM could figure the chances to actually hit the target or not, so it'd be clear where the "you both missed the whole target" point was.

(There are also rules for slow contests versus quick contests, and various scattered optional rules going into more detailed ways to resolve different situations.)

[Skarg]

Originally Posted by platimus

I mostly agree with this and is actually what I had in mind. For something like a poker contest, I'd use the 3/IQ vs 3/IQ method (modified by skills). For something like the arm-wrestling, I'd use my 2d6+ST vs 2d6+ST method. As an autonomous, self-aware being, I'm fully capable of deciding on the fly which method to use, what constitutes a critical success/failure for either, and what effects might come of a critical success/failure...or if any of that is even important.

And sometimes a tie is great and need not be resolved immediately. Suppose you had a two guys wrestling over control of a turnstile that controls whether or not a drawbridge get raised. Each time they tie, the bridge stays where it is and might give the invading side time to cross it.

Yes, there are various ways where it's pretty important that the GM consider the situation and how it makes sense to resolve it, or else the results will tend to be weird, and possibly really inappropriately disastrous, depending on what is at stake.

It seems to me that while you can figure such things out, you're an experienced GM who thinks carefully about game mechanics, which seems to me a much higher bar than "an autonomous, self-aware being". I've seen some excellent GMs just roll 3 dice and handwave a result without thinking much about how appropriate the odds they're assigning to results for are, which has led to some silly and some excessively lethal situations (especially with Climbing rolls - oh, you failed a climbing roll, guess you fall to your death - AAAAAAAaaaaaaargh SPLAT!)

[platimus]

Yeah, that climbing example is not what I would call a "self-aware" entity! LOL

In cases such as that, I would try to make sure I have adequately described the dangerousness of the situation to the players. "Are you sure? It's a long way down if you fall..."

Or I would probably have them make a check for every 10 feet they have to climb. If they fail one, they slide down 10 feet and take 1 hit. Keep climbing? Ok, resume making checks every 10 feet (or every 20 feet or whatever if it's a really tall structure; no need to get monotonous with it).