D&D 3E/3.5 Please Confirm 3.5e Tumble

[Pielorinho]

    I see what you're saying, but where do you stop if you fail only the third Tumble check? You made Tumble checks one and two, so you should get to move past those opponents. But if the opponents are in a 5' wide corridor, there is no where for you to go.

Jason

Other mechanics rule that if a failed action forces you into a specific space but that space is occupied, you end up prone in that space. I'd have you end up prone in B's space in this case.

Daniel

 

[Artoomis]

Other mechanics rule that if a failed action forces you into a specific space but that space is occupied, you end up prone in that space. I'd have you end up prone in B's space in this case.

Daniel

Right. Also, I think there is a short discussion somewhere that either states or strongly implies that the only way you can be in an opponent's square is to be prone (or grappling).

 

[jaults]

I'd have you end up prone in B's space in this case.

    Riiiiight. Forgot about that. Good call.

Jason

 

[iwatt]

Tessarel, you said:

That's also one of the things I dislike about opposed skill checks - if only a single d20 is rolled, I can work out my chance of success easily from the DC, otherwise it's a pain.

This is slightly OT, but the chances of succes in an opposed check can be found if you know your rivals check modifier. For the complete math background, you can check my post midway down this thread

What you have to understand is taht an opposed roll check between characters A and B can be characterized as:

d20_A + CM_A v/s d20_B + CM_B

CM_A (is character A's check modifier including circumstance bonus). The same goes for CM_B. Now the random variable is the substarction of two uniform variables.


Here you have to set some ground rules. What happens in cas of a tie. I decree that the initiator of the opposed check wins on a tie. So let's assume that A initiates the check. A will win when

d20_A-d20_B >= CM_B-CM_A

In the above thread I do the math if you want to check it, but believe me that the
OCD = d20_A-d20_B (Opposed Check Difference)
is a random variable with a triangular probability density function with OCD ranging from -19 to +19. The cumulative density function becomes:

P(OCD<= N)=.125*(20+N)(21+N) -19<=N<=0
P(OCD<= N)=100-.125*(20-N)(19-N) 0<=N<=+19

That means that you can find the chance to succed from the following:

P(OCD>=CM_B-CM_A) = 100-P(OCD<=CM_B- CM_A- 1)

I know this seems a bit complicated, but here you have a table.

if we define N = CM_B-CM_A-1
Or if you want you can simply define X= CM_A-CM_B (positive values indicates A is better than B.

N____% succes___X

-20_____100_____19
-19_____99,75___18
-18_____99,25___17
-17_____98,5____16
-16_____97,5____15
-15_____96,25___14
-14_____94,75___13
-13_____93______12
-12_____91______11
-11_____88,75___10
-10_____86,25___9
-9______83,5____8
-8______80,5____7
-7______77,25___6
-6______73,75___5
-5______70______4
-4______66______3
-3______61,75___2
-2______57,25___1
-1______52,5____0
0_______47,5____-1
1_______42,75___-2
2_______38,25___-3
3_______34______-4
4_______30______-5
5_______26,25___-6
6_______22,75___-7
7_______19,5____-8
8_______16,5____-9
9_______13,75___-10
10______11,25___-11
11______9_______-12
12______7_______-13
13______5,25____-14
14______3,75____-15
15______2,5_____-16
16______1,5_____-17
17______0,75____-18
18______0,25____-19
19______0_______-20

Which basically means you need to have a +19 better than the character youa are trying to beat to insure 100% effectiveness. But with "only" a +11 you are beating him 91% of the time. If you are evenly matched (X=0, N=-1) you beat him 52,5% of the time (because ties favor you).

You can easily make an excel sheet to calculate your succes rates, or just print out the above table.

 

[Tessarael]

Hi iwatt,

Thanks for the info on the cumulative density function. I generated the same N vs % chance success table by churning through the numbers.

My point was that it's not intuitive (at least if you don't have such a table ) what your chance of success is.

Let's look at Tumble. Say I have a Tumbling skill of 10 and want to Tumble past four opponents:

1. DC 15, I have an 80% chance of getting past one opponent.

2. DC 15, then DC 17, my chance is 0.80*0.70=56% chance of getting past two opponents.

3. DC 15, DC 17, then DC 19, my chance is 0.80*0.70*0.60=33.6% chance of getting past three opponents.

4. DC 15, DC 17, DC 19, then DC 21, m chance is 0.8*0.7*0.6*0.5=16.8%.

So versus one opponent, it's probably worth the risk, more than that, and I'm in trouble. But how many player's have an idea of these probabilities on the fly?

What skill do you need to have an 80% chance of tumbling past 4 opponents?
It turns out to be skill 16!

 

[iwatt]

What skill do you need to have an 80% chance of tumbling past 4 opponents?
It turns out to be skill 16!

Actually it is +17
Opponent 1 (100%)
Opponent 2 (100%)
Opponent 3 (95%)
Opponent 4 (85%)
=1*1*.95*.85=0.8075

Skill +16 is:
Opponent 1 (100%)
Opponent 2 (100%)
Opponent 3 (90%)
Opponent 4 (80%)
=1*1*.9*.8=0.72

I'm just been nitpicky though
I actually agree with you that it is not intuitive, but you shouldn't be trying to tumble past 4 opponents with only a +10 in tumble anyways. This change rationalizes the uberness of tumble without going for the opposed mechanic check (which is a lot more uncertain by the way).

The number you should check though is the +14 to tumble (which let you tumble all over the place before):
Skill +14
Opponent 1 (100%)
Opponent 2 (90%)
Opponent 3 (80%)
Opponent 4 (70%)
=1*.9*.8*.7=0.504

Remember though that this is the probability of tumbling past "all" opponents. Besides, this adds some newfound value to Mobilty as a feat (+4 AC is not anything to sneer at)

 

[KnowTheToe]

But how many player's have an idea of these probabilities on the fly?

On the fly, I was lost and I just read it

 

[DrSpunj]

By the way, only tumbling "through" carries the penalty of being stopped before entering the area. Failing a tumble "past" only means you get targeted by AoO.

That's how I understood it, and that's how I tried to relay it. I suppose I could've split that sentence to make it a bit clearer though! Here, I'll go back and bold a couple things!

Thanks for clarifying that this IS 3.5e Tumble everyone, and to drnuncheon for the "by taking a -10 on your check you can Tumble at full speed" bit. That's another thing I like about it!


Tessarael, I have to disagree with your first post. While it would shorten gameplay to have only one roll, how would you adjudicate a failed check?

If trying to tumble past 4 opponents, even if you used a DC of 21, how would rule?
1) All 4 opponents get AoOs?
2) Only one does? (or maybe 1 for every 2 pts you missed the DC?) If so, which one(s)?
3) Some other way?

If trying to tumble through 4 opponents, which ones do you get through and where do you stop on a failed check? Do you again base it on the check result vs the DCs? (i.e. You need a 21 to get past all 4, if you only got a 16 on your check you make it through the first enemy but not the second and wind up in the appropriate square)


Finally, I like the fact that Mobility would kick in and help with these AoO that anyone incurs on a failed Tumble check. I know it always has applied, but with these rules I think it looks a *LOT* more attractive than it has in the past.

DrSpunj

 

[Silver Griffon]

Other mechanics rule that if a failed action forces you into a specific space but that space is occupied, you end up prone in that space. I'd have you end up prone in B's space in this case.

Daniel

IDHMBWM, but the new PHB combat section has an entry for what to do when you "accidentally" end up in an illegal square. I believe you wind up in the nearest possible open space. I kind of like the prone idea though as opposed to winding up 10 or 15 feet away from where you were headed.

 

[DrSpunj]

IDHMBWM, but the new PHB combat section has an entry for what to do when you "accidentally" end up in an illegal square. I believe you wind up in the nearest possible open space. I kind of like the prone idea though as opposed to winding up 10 or 15 feet away from where you were headed.

Hmm, can anyone confirm this?

I'll probably house rule it back to prone in the square you're coming from (though admittedly that requires the DM to adjudicate the weirdness that is an opponent standing directly over a PC).

Does it make sense to say that if you land prone in opponent B's square (since you failed to tumble through C's square), that if you're still lying prone under B on your turn you have to first move to a free square before trying to stand up?

DrSpunj