Skill Challenges for Dummies

[two]

I don't really see what there is to explain. The DM designs the challenge, as laid out in the books. The players choose their skills, powers and items, as laid out in the books. They run through the skill challenge, and (according to AtomicPope, at least) it all works.

Well that's the thing. AtomicPope gets them to work by paying close attention to various skill boosts (and possibly misreading Beguiler's Tongue), and loading up on skill boost magic items. Cool.

Other players in the "Math Heavy" thread have indicated near continual Skill Challenge losses.

If the DM creates the challenge using the book numbers (which should apply to a generic average party), and the results are very tough for a party to "win" without major skill-maxxing, that's not the best situation.

 

[Tervin]

No, I am not, and neither is the coin-toss illustration of Gambler's fallacy. The illustration clearly acknowledges that the probability of flipping five heads in a row is 1/32. But where the fallacy lies is believing that if you succeeded an x number of times more than the math should dictate, the next result MUST be a failure to balance it out. And it just doesn't work that way.

It's not a matter of the math being wrong. It's not wrong. But it IS a matter of the math not telling the whole story.

Sorry, but no. Honestly I think you are a victim of the Gambler's Fallacy there... The Gambler's Fallacy is after all a subcase of people believing that the math is not telling the whole story - or simply misundertanding how the math works.

The math tells the whole story for general cases - which is what all that we were talking about in that thread. Your description of the Gambler's Fallacy as quoted above is correct, but has nothing at all to do with the calculations I and others made.

 

[mmu1]

No, I am not, and neither is the coin-toss illustration of Gambler's fallacy. The illustration clearly acknowledges that the probability of flipping five heads in a row is 1/32. But where the fallacy lies is believing that if you succeeded an x number of times more than the math should dictate, the next result MUST be a failure to balance it out. And it just doesn't work that way. Just because you somehow flipped heads four times in a row does not mean that you all of a sudden have a 1/32 (or less) chance of flipping heads again.

It's not a matter of the math being wrong. It's not wrong. But it IS a matter of the math not telling the whole story.

Sorry, man, but you're still not making any sense.

Who said anything about "balancing" things out? All I've seen is people calculate the actual probabilities. The math doesn't lie and it's not being misinterpreted.

 

[SableWyvern]

No, I am not, and neither is the coin-toss illustration of Gambler's fallacy. The illustration clearly acknowledges that the probability of flipping five heads in a row is 1/32. But where the fallacy lies is believing that if you succeeded an x number of times more than the math should dictate, the next result MUST be a failure to balance it out. And it just doesn't work that way. Just because you somehow flipped heads four times in a row does not mean that you all of a sudden have a 1/32 (or less) chance of flipping heads again.

Ok, that's an explanation of the Gambler's Fallacy. However, you have yet to show us how that relates in any way to the topic at hand.

 

[mmu1]

And as far as the OP goes:

Most of the relevant points have been covered already, but I think you're also hugely over-estimating the usefulness of Aid Another - the default assumption of how Skill Challenges work is that the PCs are making the checks alone, though the DM might use his discretion to allow Aid Another in particular cases. It's definitely not supposed to be a simple way to get a big bonus on every skill check in the challenge, though.

Finally, can someone post a link to the "other" thread? (I failed miserably trying to find it before. Watch it be on the front page now that I asked...)

 

[TimeOut]

Finally, can someone post a link to the "other" thread? (I failed miserably trying to find it before. Watch it be on the front page now that I asked...)

http://www.enworld.org/showthread.php?t=229919

 

[Dave Turner]

There is a complexity problem only if you conflate complexity with difficulty. The complexity of a challenge should be a function of its significance within the context of the adventure as well as general believability concerns. If it's vitally important that the players win this challenge, or if the basic objective is something that would reasonably take a long time, then it's a complex challenge.

But that's hardly intuitive from the RAW, is it? Without the benefit of the math that was crunched in the other thread, I don't think that this is a natural conclusion to draw. The hidden trap in the RAW is that a DM who assigns skill challenges based on how long it will take to perform a task isn't making a choice based on how hard the task should be. Ideally, I think that the DMG should help DMs choose tasks that are properly calibrated according to the likelihood of success. The RAW gives no indication that a Complexity 3 challenge is actually easier than a Complexity 1 challenge. So our poor newbie DMs (and many math-challenged veteran DMs) are trying to design challenges that won't frustrate their party and being mislead by the RAW.

No, I am not, and neither is the coin-toss illustration of Gambler's fallacy. The illustration clearly acknowledges that the probability of flipping five heads in a row is 1/32. But where the fallacy lies is believing that if you succeeded an x number of times more than the math should dictate, the next result MUST be a failure to balance it out. And it just doesn't work that way. Just because you somehow flipped heads four times in a row does not mean that you all of a sudden have a 1/32 (or less) chance of flipping heads again.

It's not a matter of the math being wrong. It's not wrong. But it IS a matter of the math not telling the whole story.

This is hardly an improvement in your approach to this topic. If you had bothered to read the math-heavy thread, you'd see that the mathematicians there took pains to carefully explain their reasoning. It seems that many math-challenged readers (like myself) were able to follow their reasoning without problem.

Your alleged explanation is a pale, pale shadow of theirs. You may very well be right, but you're killing your message. Why bother with trying to defend yourself ("No, I am not...") if you aren't really committed to it. There have been numerous, skeptical responses to your assertions in this thread. If you care about your criticism, you'll need to give a proper response which lays out your ideas in a way that non-mathematical folks can understand.

A good start would be to reference the statements by mathematicians in the other thread by quotation and then showing why the Gambler's Fallacy undermines them. I am, of course, being charitable by suggesting that you've already read the thread in its entirety.

 

[Tervin]

One might argue that the entire DMG is written from the perspective of ensuring that the DM is in as good a mood as possible, what with things like "say yes" and stuff.

True. Actually something I really like.

Why? You take the PCs' level into account, what you want the skill challenge to achieve, its general significance as a part of the adventure and ongoing storyline, and so on. Scaling up difficulty just because the PCs have items that make skill checks easier just means you're back to a red queen-style setup.

Normally I would never want to do that. But as it is the rules as written don't make me feel confident in that the difficulty level is set right. You are right that it is wrong to punish people for being especially good at something - it is just that right now I don't know what average means in this system, and the DMG doesn't help. To be more exact my plan is to set the difficulties quite a bit lower than the DMG suggests, but if skill boosting items should be taken into account, I might reduce the difficulty less.

There is a complexity problem only if you conflate complexity with difficulty. The complexity of a challenge should be a function of its significance within the context of the adventure as well as general believability concerns. If it's vitally important that the players win this challenge, or if the basic objective is something that would reasonably take a long time, then it's a complex challenge.

Those are very good points, but they are not quite enough for me.

The complexity problem is there because the rules state that a success in a complexity 1 challenge should give out as much XP as 1 monster of its level, while a success in a complexity 5 challenge should give out as much XP as 5 monsters of its level. It doesn't feel fair to the players to get a lot more XP for something that is easier to do, just because it meant rolling more dice.

Taking in your arguments, plus making the skill challenge hard enough that bonuses from clever play is what takes the probabilities up from a "likely" to a "very probable" success will probably make this work for me.

 

[Darth Cyric]

Who said anything about "balancing" things out? All I've seen is people calculate the actual probabilities. The math doesn't lie and it's not being misinterpreted.

No, what the math calculates is the probability that a certain sequence will take place before any part of it happens to take place. The math is not wrong in that context.

Where the math does not tell the whole story, what it is dismissing, is the fact that each die roll is an independent and unrelated occurance. The result of one die roll does not dictate the probability of the next die roll.

 

[Tervin]

Where the math does not tell the whole story, what it is dismissing, is the fact that each die roll is an independent and unrelated occurance. The result of one die roll does not dictate the probability of the next die roll.

Look at the formulas involved. Hey why don't I quote one from my Excel sheet to make it easier to follow? This is the formula used to calculate the probability of success for a Complexity 4 challenge:

=E4^14+E4^13*(1-E4)*14+E4^12*(1-E4)^2*91+E4^11*(1-E4)^3*364+E4^10*(1-E4)^4*1001

(E4 = the probability of any individual roll being a success.)

Where does this formula take into account what the result has been of any earlier roll? In fact the formula is set up like it is specifically so that it will look at each roll individually...