Taking 20 dnd 3.5

[finaldooms]

Hi yall! So i cant find the rules on taking 20 for some reason so im here to ask yall

In the case of searching how long does taking 20 normally take?
What about climbing..a 20ft wall? Or is the time up to the dm completely?( no real guidelines)

Im asking because the rogue in the party...has no ranks in search ( he is the trapfinder) and i had him search a body and he rolled an 8 which was not enough to find something

I thought i phrased it in a way that he found " everything" but i might have said it to...obviously? So now he wantz to take 20 and look it over again..although the party does not want to wait long on him

[prototype00]

Taking 20

When you have plenty of time (generally 2 minutes for a skill that can normally be checked in 1 round, one full-round action, or one standard action), you are faced with no threats or distractions, and the skill being attempted carries no penalties for failure, you can take 20. In other words, eventually you will get a 20 on 1d20 if you roll enough times. Instead of rolling 1d20 for the skill check, just calculate your result as if you had rolled a 20.

Taking 20 means you are trying until you get it right, and it assumes that you fail many times before succeeding. Taking 20 takes twenty times as long as making a single check would take.

Since taking 20 assumes that the character will fail many times before succeeding, if you did attempt to take 20 on a skill that carries penalties for failure, your character would automatically incur those penalties before he or she could complete the task. Common “take 20” skills include Escape Artist, Open Lock, and Search.

You're welcome.

[meschlum]

Which answers the question as to the time it takes, compared to the base time.

The time you need for skills is usually in the skill description (search takes 1 full round action to search a 5x5 area, or goods covering the same). For climbing, you move at quarter speed with a successful check - so 5' per round (30 / 4 = 7.5, probably rounded down) for most characters. If you take a penalty on the roll, you can move at half speed (15' per round for most characters). Of course, if you fail, you may fall...

So a normal climb check gets you past a 20' wall in 4 rounds, if you make the 4 checks (one per round). It's a move action, so you could double move and do it in two rounds.

An accelerated climb check lets you go 15' in one move action, so you'd get past a 20' wall in a single round. Move action 1: normal climb check. Move action 2: accelerated climb to the top (best do it in the order, so if you fall it's only 5' and you take no damage).

If you have a Climb speed, you move at that speed instead with each Climb roll (or twice that if you use accelerated climbing), and you have significant bonuses too. So with a Climb speed of 20', it would take a single move action to get up the wall (and you'd be rolling with a +8 and the ability to take 10, so a minimum of 18 before skill ranks and attribute bonuses or penalties).

[finaldooms]

Thanks alot !!
And the example for,climbing helps a good bit too!! I feel like the rogue might start climbing things soon sincw he likes his bow so much and has been talking ooc about looking into rooftops

[Kaleph]

I believe you cannot take 20 on climb, since there's a risk the PC may fall.

[prototype00]

I believe you cannot take 20 on climb, since there's a risk the PC may fall.

Well, RAW, you can take 20 on climb, if you are willing to assume that you failed numerous times.

How is your fall damage mitigation?

prototype00

[Crake]

Well, RAW, you can take 20 on climb, if you are willing to assume that you failed numerous times.

How is your fall damage mitigation?

prototype00

You also need to take into account the loss of progress that comes with climbing. If you cannot make the climb in a single check, then you can only really take 20 on the first climb check, because beyond that, even if there's no damaging penalty for failure, there is still the loss of progress.

[DEMON]

You also need to take into account the loss of progress that comes with climbing. If you cannot make the climb in a single check, then you can only really take 20 on the first climb check, because beyond that, even if there's no damaging penalty for failure, there is still the loss of progress.

So 20 attempts to make the first check times 20 attempts for the second one. If you can safely make 400 attempts, you should be allowed to Take 20 on the second iteration, right? :smallsmile: 8000 attempts for the 3rd iteration and you're good... :smalltongue:

[OldTrees1]

So 20 attempts to make the first check times 20 attempts for the second one. If you can safely make 400 attempts, you should be allowed to Take 20 on the second iteration, right? :smallsmile: 8000 attempts for the 3rd iteration and you're good... :smalltongue:

Only if you have a chance of losing progress(must fail by 5 or more). Otherwise it would be 20 + 20 + 20 ...

Math question: If on a 1 you just barely pass or only fail by 1, 2, 3, or 4, at what height of cliff (1 unit = your movement speed in ft) does it stop being faster to take the -5 penalty for accelerated movement (x2 speed but 5%-25% chance of falling)?

[KillianHawkeye]

You cannot take 20 on any skill if there are penalties for failure. The penalty for failing a Climb check is that you fall (possibly taking damage), so you cannot ever take 20 on a Climb check. Not ever.

[OldTrees1]

You cannot take 20 on any skill if there are penalties for failure. The penalty for failing a Climb check is that you fall (possibly taking damage), so you cannot ever take 20 on a Climb check. Not ever.

If a wall has a DC of 15 and I have a +10 modifier I will never fall (not even on a 1).

[meschlum]

Math!

If you always pass at normal speed, you move at 1/4 your move speed (rounding down). You're making double moves and can go as high as you want, so you get:

Move Speed 20 (small size): 10 feet / round
Move Speed 30 (medium size): 10 feet / round (rounding hurts - it could be 15 instead)
Move Speed 40 (barbarian / monk): 20 feet / round

If you fail to advance 5-20% of the time, you only move 95-80% as fast on average

Move speed 20: 9.5 - 8 feet / round
Move speed 30: 9.5 - 8 feet / round
Move speed 40: 19 - 16 feet / round

If you fall 5% of the time, you also fail to advance 20% of the time, and would always advance at normal speed. Ignoring falling damage, we get:

In the long run, at any point in time, you're on the ground if you were on the ground last move and didn't manage to climb, or if you fell last turn.

P(H* = 0) = 1/5 * P(H* = 0) + 1/20, where H* is the long term height. This gives a long term probability of being on the ground of 1/16.

Otherwise, you're at height H* = N if you were at N-1 last move and managed to climb, or if you were at N and stayed in place.

P(H* = N) = 3/4 * P(H* = N-1) + 1/5 * P(H* = N), so P(H* = N) = 15/16 P(H* = N-1).

Fit in the probability of being on the ground, and you get an exponential curve, where P(H* = N) = (15/16)^N * 1/16.

Compute the sum of probability times height, and you get and average height of 15 'steps'. Meaning you'll probably fall fairly often before you get beyond that point.

So if you want to climb 10 'steps', in the long run, you are that high with probability (15/16)^10, so 52.44% of the time. Meaning you need (16/15)^10 moves to climb one 'step' in this case, so 1.9 moves - meaning you need 19 moves to advance 10 'steps'.

Move speed 20: 100 feet in 19 moves, around 5.2 feet per move, so 10.49 feet per round
Move speed 30: 150 feet in 19 moves, 15.73 feet per round.
Move speed 40: 200 feet in 19 moves, 20.98 feet per round.

If you want to climb 20 'steps', you get that high 27.5% of the time, so you need 3.64 moves to climb one step.

Move speed 20: 200 feet in 72.8 moves, so 5.5 feet per round
Move speed 30: 300 feet in 72.8 moves, so 8.25 feet per round
Move speed 40: 400 feet in 72.8 moves, so 11 feet per round.

Meaning that double speed is faster if you're trying to get up to 10-11 'steps' up (100-220 feet), and slower otherwise.

If your odds of failure are higher, the odds change faster, and double speed is worth it for lower and lower heights.

[OldTrees1]

Math!
Yay Math!


-snip for brevity, go read the post firsthand-

In the long run, at any point in time, you're on the ground if you were on the ground last move and didn't manage to climb, or if you fell last turn.

P(H* = 0) = 1/5 * P(H* = 0) + 1/20, where H* is the long term height. This gives a long term probability of being on the ground of 1/16.

Otherwise, you're at height H* = N if you were at N-1 last move and managed to climb, or if you were at N and stayed in place.

P(H* = N) = 3/4 * P(H* = N-1) + 1/5 * P(H* = N), so P(H* = N) = 15/16 P(H* = N-1).

Fit in the probability of being on the ground, and you get an exponential curve, where P(H* = N) = (15/16)^N * 1/16.

Compute the sum of probability times height, and you get and average height of 15 'steps'. Meaning you'll probably fall fairly often before you get beyond that point.

Interesting emergent behaviour.

If I understood the math: Someone that would succeed 10/20, halt 4/20, and fall 6/20 would have
PH*=0 = 6/20 + 4/20 * (PH*=0)
PH*=0 = 6/(20-4)

PH*=N = 10/20 *(PH*=N-1) + 4/20 * (PH*=N)
PH*=N = (PH*=N-1) * 10/(20-4)
PH*=N = (PH*=0) * (10/(20-4)) ^ N
PH*=N = (6/(20-4)) * (10/(20-4)) ^ N

So if I understand the math correctly. If a person would halt 4/20, pass C/20 & fall (16-C)/20
PH*=N = ((16-C)/16) * (C/16) ^ N
Edit: Added a correction



Meaning that double speed is faster if you're trying to get up to 10-11 'steps' up (100-220 feet), and slower otherwise.

If your odds of failure are higher, the odds change faster, and double speed is worth it for lower and lower heights.

Thanks!

[Calthropstu]

Ladies and gentlemen, this thread right here is the perfect demonstration of why games should NOT be realistic... you should not need to learn advanced math in order to play a game. Most people would take one look at this thread and think... "nope."

[Kaleph]

If a wall has a DC of 15 and I have a +10 modifier I will never fall (not even on a 1).

In this case you don't take 20, you take 10.

[OldTrees1]

Ladies and gentlemen, this thread right here is the perfect demonstration of why games should NOT be realistic... you should not need to learn advanced math in order to play a game. Most people would take one look at this thread and think... "nope."

Algebra I is not advanced math* and is not needed in order to play the game. It is merely an emergent behaviour resulting from the repeated dependant cases of arithmetic and RNG that one does need in order to play the game. You can understand it with algebra or you can just let it happen.

*Although the notation above is not very approachable. The notation "F(X) = 10/20 * F(X-1) + 4/20 * F(X)" might have been more so.


In this case you don't take 20, you take 10.

Fair enough.

[meschlum]

Yay Math!

Interesting emergent behaviour.

If I understood the math: Someone that would succeed 10/20, halt 4/20, and fall 6/20 would have
PH*=0 = 6/20 + 4/20 * (PH*=0)
PH*=0 = 6/(20-4)

PH*=N = 10/20 *(PH*=N-1) + 4/20 * (PH*=N)
PH*=N = (PH*=N-1) * 10/(20-4)
PH*=N = (PH*=0) * (10/(20-4)) ^ N
PH*=N = (6/(20-4)) * (10/(20-4)) ^ N

So if I understand the math correctly. If a person would halt 4/20, pass C/20 & fall (20-C)/20
PH*=N = ((20-C)/16) * (C/16) ^ N



Thanks!

Happy to help!

The equation you wrote isn't quite correct - you have 24/20 chances of something happening, which is a bit much. You should use (16-C) instead, and get:

PH*=N = ((16-C)/16) * (C/16) ^ N

The other extensions which result are:

Average height in 'steps' (remember these are at half speed, rather than 1/4 speed for normal climbing) = C / (16 - C)

Probability of being T 'steps' or higher = (C / 16) ^ T

Average height gain (in half speed 'steps') per move action = (C / 16) ^ T (same as the probability)


Now for logarithms! Fast climbing gives a move speed that is typically 2 times as much as normal climbing, except when half move speed is odd (e.g. 30' move). Set the ratio to R (2 for move speeds 20' and 40', 3 for move speed 30', as you get 15' with fast climbing rather than 5' with normal).

We want to find out when R * (C / 16) ^ T = 1, since that's the target height at which the gain from fast climbing is exactly balanced by the risk of falling.

Take the logarithm of the equation, you get log(R) + T * log(C / 16) = 0, so T = - log(R) / log(C / 16)

As a very coarse approximation when C is high, log(2) = 0.6 and log(C / 16) = (16 - C) / 16, so we get for R = 2:

C = 15: T around 9.6 (actual value 10+, details above)
C = 14: T around 4.8 (actual value 5+)
C = 13: T around 3.2 (actual value 3+)
C = 12: T is a bit more than 2 (at T = 2, probability is 9/16, and 27 / 64 at T = 3 is less than 1/2)
C = 11: T is a bit less than 2 (121 / 256 is less than 0.5)
...
C = 8: T is exactly 1, obviously.

Using the approximation that log(3) = 1.1, we get for R = 3:

C = 15: T around 17 (matches the 20 'step' estimate above)
C = 14: T around 9 (actually in the 8-9 range)
C = 13: T around 6 (actually 5-6 range)
C = 12: T around 4.4 (actually a bit less than 4 - the approximation is getting worse)
C = 11: T is 3, numerically (1331 / 4096) (approximation is 3.5, still in the right range)
...
C = 9: T is 2+, numerically (81 / 256)

[KillianHawkeye]

You cannot take 20 on any skill if there are penalties for failure. The penalty for failing a Climb check is that you fall (possibly taking damage), so you cannot ever take 20 on a Climb check. Not ever.


If a wall has a DC of 15 and I have a +10 modifier I will never fall (not even on a 1).

It doesn't matter if you can't possibly fail. It's still a skill that carries penalties for failure.




In this case you don't take 20, you take 10.


Fair enough.

Also that. If you can't possibly fail and still want to take 20, you're just wasting your own time.