PDA

View Full Version : 5-20 crit range with Disciple of Dispater (3.5)



Frog Dragon
2010-07-21, 01:14 PM
This is an upgrade of the old aptitude kukris and insane crit range trick. I haven't seen this in builds yet so am I missing something, did I actually discover something new (very unlikely) or is there something else about this, like a better trick?
The Disciple of Dispater's (BoVD) Iron Power ability doubles your crit range at level 4 with iron and steel weapons and triples it at level 8. This specifically stacks with Improved Critical. So, using Aptitude Kukris with the Improved Critical Feat you get..
Normal Crit Range 17-20. 4 numbers crit. Imp Crit 13-20, 8 numbers crit.
With Iron Power +1 9-20, 12 numbers crit. Iron Power 5-20, 16 numbers crit.
You could accomplish this at level 14 (as you need BAB +6 to enter). Add Blood in the Water for extra fun.
So yeah.

Pechvarry
2010-07-21, 01:20 PM
I don't own BoVD so don't know the class, but your math appears wrong. It looks like you're applying Dispater's multiplication AFTER improved crit which isn't what "stacks with" means, though 17-20 wouldn't be the result anyway.

Ok, you're simply using 17-20 as base stats for kukri. I don't know how you're getting that instead of 18-20.

Kukri: 18-20
Improved crit: 15-20
+1 class feature: 12-20
+2 class feature: 9-20.

Still flippin' potent, but not 5-20.

Private-Prinny
2010-07-21, 01:20 PM
Screw BitW, slap Lightning Maces on those puppies.

Keld Denar
2010-07-21, 01:20 PM
Do note that the Aptitude Kukiri trick with Lightning Maces is pretty questionable. The intent was obviously to allow an aptitude weapon to benefit from a feat that allows you to CHOOSE a weapon, such as Weapon Focus, Weapon Spec, etc, rather than feats that require a specific weapon, such as Lightning Maces, Three Mountains, or god forbid Boomerang Daze.

That said, nice crit range!

Dusk Eclipse
2010-07-21, 01:30 PM
Shame you can't use DoD abilities with kaorti weapons, 1d4 9-20 x4 weapon sounds extremely nasty even before rider effects such as prismatic burst

9mm
2010-07-21, 01:32 PM
Screw BitW, slap Lightning Maces on those puppies.

your thinking to small... go the full Olo (who came up with this trick and demonstrated it hilliariously in ToS) and get completely metal handcrossbows and add the splitting enhancement: enjoy the exponential damage curve of "how many bolts do you have x2".

Eldariel
2010-07-21, 01:36 PM
your thinking to small... go the full Olo (who came up with this trick and demonstrated it hilliariously in ToS) and get completely metal handcrossbows and add the splitting enhancement: enjoy the exponential damage curve of "how many bolts do you have x2".

With Roundabout Kick.

Keld Denar
2010-07-21, 01:37 PM
I wrote some MatLab code using the RandomInt function to run simulations of how silly a normal Lightning Maces Aptitude (if you allow it to work) would get. I had a feedback loop that increased damage per hit by 1 every time you confirmed a crit and in theory should be able to calculate the average damage you'd deal before a nasty string of 1s end the iteration.

Unfortunately, I don't have MatLab to actually run the code and check it for bugs, but it looks flawless. Then again, I'm a TERRIBLE debugger, and MatLab is a bassackwards language to code in anyway. Plus, I don't trust the RandomInt function to be completely unbiased...no computer function is ever truely "random".

Dusk Eclipse
2010-07-21, 01:38 PM
9mm: wasn't that the build that critted itself to death? I also think it was a night infinite loop.

Eldariel
2010-07-21, 01:40 PM
9mm: wasn't that the build that critted itself to death? I also think it was a night infinite loop.

Yes. It's a rather rudimentary loop; basically every attack is a crit and every crit generates more hits (Maces and Roundabout) and every attack is two hits as per Splitting. It gets rather ridiculous rather fast.

JaronK
2010-07-21, 01:47 PM
Huh, I remember making that build a while back (not for ToS though) but it used Aptitude Great Crossbows. Great Crossbows have a better critical range than hand crossbows, and with Hand Crossbow Mastery + Aptitude you can still reload them at full speed. It required extra hands though.

JaronK

Boci
2010-07-21, 01:51 PM
I wrote some MatLab code using the RandomInt function to run simulations of how silly a normal Lightning Maces Aptitude (if you allow it to work) would get. I had a feedback loop that increased damage per hit by 1 every time you confirmed a crit and in theory should be able to calculate the average damage you'd deal before a nasty string of 1s end the iteration.

Unfortunately, I don't have MatLab to actually run the code and check it for bugs, but it looks flawless. Then again, I'm a TERRIBLE debugger, and MatLab is a bassackwards language to code in anyway. Plus, I don't trust the RandomInt function to be completely unbiased...no computer function is ever truely "random".

Didn't something similar to that spring an argument between between two posters as to whether or not the amount of attacks was theoretically infinite.

Keld Denar
2010-07-21, 01:55 PM
Statistically, its not infinite. On any truely infinite string of numbers that range between 1 and 20, there exists a string of 1s long enough to stall out the recusion. Now, the longer it goes, the more successive ones needed to stall it out, but the chance of such a string of 1s never truely reaches 0% at any point. Thus, since there's no chance it can't fail, it must fail at some point and thus can't be truely infinite.

Go go gadget limit calculus!

EDIT for clarity: If you had, say, 5 attacks per round when you started attacking, if you rolled 5 consecutive 1s, the iteration would stall out. If you rolled one hit but no crits, it would take 6 1s to stall out, and if you rolled one crit, it would take at least 7 1st to stall out. If, for example, you rolled all 5 crits, you'd then have 10 attacks remaining, and you'd have to roll 10 1s in order to stall out. If you crit enough without stalling, you might have to roll 1000000000000 1s in a row in order to stall out. The odds of this are EXTREMELY low, but on an infinte string of numbers, the probability of this happening is REALLY low, but never 0. Thus, its its possible to stall out even after you get rolling, but when you do finally stall out, you'll probably find out that you need scientific notation to quantify your damage, and its not really relevant at that point.

JaronK
2010-07-21, 01:55 PM
Yup, I was one of them in fact. Never did quite resolve that... I wanted a simulation based on different critical threat values (to see which value went infinite). Basic statistics for this situation doesn't actually work very well, as there's a similar problem that shows a break in how statistics work (the doubling coin flip game).

JaronK

2xMachina
2010-07-21, 02:10 PM
Shame you can't use DoD abilities with kaorti weapons, 1d4 9-20 x4 weapon sounds extremely nasty even before rider effects such as prismatic burst

Depends if you want to kill catgirls...

Steel is ok with DoD. Can you use resins to make steel? You can make composite material with them I guess.

But then, you can mix with Crystal Weapon Master. Steel is a crystalline material, so...

Abuse Material Science!

Keld Denar
2010-07-21, 02:12 PM
Hey now...respect the grain boundaries or I'll Jominy End Quench you, and it won't be comfortable!

Frog Dragon
2010-07-21, 02:49 PM
I don't own BoVD so don't know the class, but your math appears wrong. It looks like you're applying Dispater's multiplication AFTER improved crit which isn't what "stacks with" means, though 17-20 wouldn't be the result anyway.

Ok, you're simply using 17-20 as base stats for kukri. I don't know how you're getting that instead of 18-20.
That's because I war working off memory for the weapons and mistakenly remembered it was an exotic weapon with a crit range of 17-20. I'm applying the multiplication to the base crit and adding together.

true_shinken
2010-07-21, 05:19 PM
Well, there IS a 17-20 weapon. Stump Knife from Sword & Fist, I believe.
Of course, Aptitude abuse is silly.

Tinydwarfman
2010-07-21, 05:40 PM
Uh, actually Disciple of Dispater was exactly what was used before, you just did normal math, not D&D math.

true_shinken
2010-07-21, 05:43 PM
Uh, actually Disciple of Dispater was exactly what was used before, you just did normal math, not D&D math.

Nope, he got the threat range wrong.

Tinydwarfman
2010-07-21, 05:48 PM
Nope, he got the threat range wrong.

Ah, right, I've just seen it done so many times I assumed that was what it was.

SurlySeraph
2010-07-21, 06:32 PM
Throw in 7 levels in Barbarian with the Streetfighter ACF from Cityscape to make it a 4-20 crit.

Tinydwarfman
2010-07-21, 07:23 PM
Throw in 7 levels in Barbarian with the Streetfighter ACF from Cityscape to make it a 4-20 crit.

But only on a charge right?

Dr.Epic
2010-07-21, 07:27 PM
This is an upgrade of the old aptitude kukris and insane crit range trick. I haven't seen this in builds yet so am I missing something, did I actually discover something new (very unlikely) or is there something else about this, like a better trick?
The Disciple of Dispater's (BoVD) Iron Power ability doubles your crit range at level 4 with iron and steel weapons and triples it at level 8. This specifically stacks with Improved Critical. So, using Aptitude Kukris with the Improved Critical Feat you get..
Normal Crit Range 17-20. 4 numbers crit. Imp Crit 13-20, 8 numbers crit.
With Iron Power +1 9-20, 12 numbers crit. Iron Power 5-20, 16 numbers crit.
You could accomplish this at level 14 (as you need BAB +6 to enter). Add Blood in the Water for extra fun.
So yeah.

Then the DM decided you'll only be fighting undead and constructs for the next 100 adventures.

Eurus
2010-07-21, 07:36 PM
Never seen Streetfighter, but can you get both it and Lion Totem?

Flickerdart
2010-07-21, 07:38 PM
Never seen Streetfighter, but can you get both it and Lion Totem?
Streetfighter only replaces Damage Reduction, so yes.

KillianHawkeye
2010-07-21, 10:09 PM
Well they changed the way critical threat stacking works between 3.0 and 3.5, so even though Disciple of Dispater wasn't officially revised your DM should probably change the way it works in order to make it acceptable in a 3.5 game. Sorry.

Optimystik
2010-07-21, 10:29 PM
Then the DM decided you'll only be fighting undead and constructs for the next 100 adventures.

No love for plants/oozes? :smallfrown:

Mongoose87
2010-07-21, 10:48 PM
Well they changed the way critical threat stacking works between 3.0 and 3.5, so even though Disciple of Dispater wasn't officially revised your DM should probably change the way it works in order to make it acceptable in a 3.5 game. Sorry.

I disagree. Disciple of Dispater specifically states that it stacks - this is its shtick, whether used for cheese or not.

Optimystik
2010-07-21, 10:50 PM
Well they changed the way critical threat stacking works between 3.0 and 3.5, so even though Disciple of Dispater wasn't officially revised your DM should probably change the way it works in order to make it acceptable in a 3.5 game. Sorry.

Sorry for what? That you have houserules?

Frog Dragon
2010-07-22, 04:19 AM
I disagree. Disciple of Dispater specifically states that it stacks - this is its shtick, whether used for cheese or not.
Yup, that's the idea. 3.0 or not. It is still the specific source. Therefore, it stacks. :smallbiggrin:
That stump knife would be great for this though. I'll need to check it.
The stump knife has a bunch of limitations, like having had to deal damage already. However what we really want is the bladed gauntlet.
Damage: 1d6, crit range 17-20. The weapon is exotic and small in 3.0 terms. This means light in 3.5 terms. Dual wield a pair of these. Go to town.

KillianHawkeye
2010-07-22, 05:34 AM
I disagree. Disciple of Dispater specifically states that it stacks - this is its shtick, whether used for cheese or not.


Sorry for what? That you have houserules?

It would be like using something from 3.0 psionics when you're playing in a 3.5 game. Psionics was completely changed between 3.0 and 3.5, so anything from 3.0 should probably be altered to match the new psionics paradigm. Same goes for critical stacking. It's not a house rule. Those are the official reccomendations.

Santra
2010-07-22, 06:09 AM
I dont think you understand what they are saying. Disciple of Dispater is made to stack with the new Improved Critical. It is one of the benefits of the class.

http://i8.photobucket.com/albums/a11/blackhelldemon/Saywhut.jpg

Mongoose87
2010-07-22, 06:21 AM
I dont think you understand what they are saying. Disciple of Dispater is made to stack with the new Improved Critical. It is one of the benefits of the class.

http://i8.photobucket.com/albums/a11/blackhelldemon/Saywhut.jpg

Exactly. Houseruling this is basically going out of your way to nerf a class that isn't involved in any more serious cheese than a good charger build, let a lone a spellcaster, especially if you don't allow the Aptitude Lightning Maces trick.

Boci
2010-07-22, 06:25 AM
It would be like using something from 3.0 psionics when you're playing in a 3.5 game. Psionics was completely changed between 3.0 and 3.5, so anything from 3.0 should probably be altered to match the new psionics paradigm. Same goes for critical stacking. It's not a house rule. Those are the official reccomendations.

No, psionics is updated, BoVD is not. Therefor it is still valid. Given that is specifically mentions not stacking with a keen blade, I see no reason to add improved crit to that list.


Then the DM decided you'll only be fighting undead and constructs for the next 100 adventures.

Then the player laughs as they points out that you only need to threatan a critical.

Optimystik
2010-07-22, 08:07 AM
It would be like using something from 3.0 psionics when you're playing in a 3.5 game. Psionics was completely changed between 3.0 and 3.5, so anything from 3.0 should probably be altered to match the new psionics paradigm. Same goes for critical stacking. It's not a house rule. Those are the official reccomendations.

What? It's not like Psionics at all. There is a 3.5 version of the PH. There is no 3.5 version of BoVD.

The only things from BoVD that were explicitly updated are the Devil and Demon lords, in the Fiendish Codices. Disciple of Dispater was not.

Boci
2010-07-22, 08:10 AM
.The only things from BoVD that were explicitly updated are the Devil and Demon lords, in the Fiendish Codices. Disciple of Dispater was not.

Actually, weren't the Devil and Demon lords in Fiendish Codices just their avatars? Even then the ones in BoVD should be updated, but still.

Pechvarry
2010-07-22, 11:45 AM
Maybe Haste would be a better example -- if I find a non-updated 3.0 class that gives Haste as a SLA and says in its text "...gaining an extra standard action each turn..." then we'd have a problem. The specific class states 3.0 rules, which makes it a case of specific > general. But there's a rules update to take into consideration. So if you use the 3.0 part, you can't say "but it getting an extra standard action is its schtick!" because that schtick is now illegal.

Of course, I can't find such a class, but it's a pretty good comparison. And if you nerf it to where it "should" be -- with the ability not stacking with improved crit but stacking with itself at 8th level for a still-impressive threat range, then you have to re-examine entry reqs and determine if the class is even functional anymore... 3.0 sucks.

Tyndmyr
2010-07-22, 11:49 AM
Well yes, if a new Disciple of Dispater class was printed, the new supercedes the old. Im not aware of any such class.

Stacking rules are part of the class though, so yeah, it stacks. It says it does. Specific overrides general.

jseah
2010-07-22, 01:05 PM
I asked a math guy in cambridge about the doubling coin flip problem.
- Rule: Every time I flip a coin and it lands heads, I get two more coins to flip
- I start with 1 coin. All coins are unbiased.
Terminology: A toss is when you flip all the coins you are entitled to from the results of the previous toss.
The first coin is toss 1.

His reply was that the average number of coin flips was infinity but all trials will stop. He also replied (at my confusion) that there was no contradiction in that. =/

His explanation reworded:
If you plot a graph of log(number of coin flips) vs toss number, the average line is a straight line.
The sum of average line across all the tosses is therefore infinity.

To put it another way, every coin I flip gives me, on average, 1 more coin to flip, which gives me 1 more coin... etc


However, if you plot any such trial on that graph, the number of coins available to the tosses wobbles up and down similarly to a random walk (close but not quite). The trial only stops when the number of coins hit zero.
Therefore, the chance of any trial hitting zero, is 1. All trials will eventually stop.


The apparent contradiction is solved by indicating this:
The sum of coins flipped in each trial doubles when the chance to receive that number halves.
Basically, the chance you get 2X coins is half of the chance you get X coins.

Since the average is calculated as the sum of (return x chance of return), that sums to infinity.

Keld Denar
2010-07-22, 03:14 PM
Its never quite infinite. No matter how many coins you have left to flip, there is a chance, however small, that the remaining flips will all be tails. Heck, there is a 50/50 chance that the experiment ends after the first flip, and a 5/8 chance that the experiment doesn't make it past 3 coins. If you have 500 flips remaining, the chance of all 500 being tails to end the experiment is very very VERY small, but its still greater than 0.

Similarly, if every successful attack creates an extra hit (from Roundabout Kick) and every successful crit creates 2 extra attacks (from Roundabout Kick + Lightning Maces), its gonna take a lot of 1s in a row to stall this puppy out once it gets going, but on an infinite string of randomly generated numbers, there WILL always be a string of 1s long enough to stall it out, no matter how many attacks are remaining at any given point in the simulation.

There is no way to mathematically calculate out how many swings you'll truely make before stalling out each time because it depends too much on the order of numbers that are inherantly unordered. The only way to quantify it is through experimental data aquisition, which may require a more powerful computer than I have, simply because there is the potential to perform SOOOOOOOO many iterations before stalling. Unlike the 1d2 crusader who has a 100% chance of going to the next step and repeating itself, a 5-20 crit range character has a 24% chance (20% + 5%*80%) of NOT generating 2 attacks, and a 5% chance of not generating 1 attack at each iteration. Since it doesn't have a 100% chance of repeating the iteration, it has to have a 100% chance of stopping...eventually.

JaronK
2010-07-22, 03:25 PM
That problem is very similar to the one I was thinking of when we first worked out that build... standard statistics gives you the wrong answer (it always terminates, but basic math says it's infinite). The version I'd heard was a bet... you pay me some money, and I flip a coin. If it's heads, I give you a dollar and flip again. If it's heads again, I give you two dollars and flip again. If it's heads again, I give you four dollars and flip again, and we keep doing this, where I keep giving you double the last value. But if I ever flip tails, it ends. How much should you be willing to bet?

Standard statistics says bet as much as you like, you'll come out ahead in the long run if you do it enough times. That's because your chance of returns keeps halving but your prize keeps doubling, so your expected returns are 1/2 * 1 + 1/4 * 2 + 1/8 * 4... = 1/2+1/2+1/2+1/2... = infinite. But your actual expected returns if you just try it are about $11, give or take.

Fun stuff, huh? But we got into a debate because the other poster didn't understand this concept.

JaronK

Baron Corm
2010-07-22, 03:35 PM
I dont think you understand what they are saying. Disciple of Dispater is made to stack with the new Improved Critical. It is one of the benefits of the class.

http://i8.photobucket.com/albums/a11/blackhelldemon/Saywhut.jpg

Based on this text, it is very clear to me that the people who wrote it intended the class feature to essentially function as having all of your weapons become Keen. They did not want it to be an overpowered ability - none of the Disciple classes are very good. So translate that into 3.5, and Keen no longer stacks with Improved Critical, therefore it doesn't work. It works perfectly well in TO since there was never any errata, but it should only be brought to the same tables as Pun-Pun.

senrath
2010-07-22, 03:39 PM
That problem is very similar to the one I was thinking of when we first worked out that build... standard statistics gives you the wrong answer (it always terminates, but basic math says it's infinite). The version I'd heard was a bet... you pay me some money, and I flip a coin. If it's heads, I give you a dollar and flip again. If it's heads again, I give you two dollars and flip again. If it's heads again, I give you four dollars and flip again, and we keep doing this, where I keep giving you double the last value. But if I ever flip tails, it ends. How much should you be willing to bet?

Standard statistics says bet as much as you like, you'll come out ahead in the long run if you do it enough times. That's because your chance of returns keeps halving but your prize keeps doubling, so your expected returns are 1/2 * 1 + 1/4 * 2 + 1/8 * 4... = 1/2+1/2+1/2+1/2... = infinite. But your actual expected returns if you just try it are about $11, give or take.

Fun stuff, huh? But we got into a debate because the other poster didn't understand this concept.

JaronK

That would be the Martingale betting system. It'll generally bankrupt you if you ever actually try it, although it does have an infinite return rate (provided an infinite input).

Tinydwarfman
2010-07-22, 03:40 PM
Based on this text, it is very clear to me that the people who wrote it intended the class feature to essentially function as having all of your weapons become Keen. They did not want it to be an overpowered ability - none of the Disciple classes are very good. So translate that into 3.5, and Keen no longer stacks with Improved Critical, therefore it doesn't work. It works perfectly well in TO since there was never any errata, but it should only be brought to the same tables as Pun-Pun.

Only brought to the same tables as PUN-PUN? That, sir, is ridiculous, especially since it's not even remotely as strong as many 3.5 PrCs that were intended to work they way they were written. What exactly is so terrible if you disallow Lightning Maces? I mean really, Lightning Maces + this is TO, but on it's own it's not that bad.

Frog Dragon
2010-07-22, 03:46 PM
Based on this text, it is very clear to me that the people who wrote it intended the class feature to essentially function as having all of your weapons become Keen. They did not want it to be an overpowered ability - none of the Disciple classes are very good. So translate that into 3.5, and Keen no longer stacks with Improved Critical, therefore it doesn't work. It works perfectly well in TO since there was never any errata, but it should only be brought to the same tables as Pun-Pun.
Exactly. This trick is TO. It's munchkinry.
Edit: If you use lightning maces and BitW.

Keld Denar
2010-07-22, 03:50 PM
Fun stuff, huh? But we got into a debate because the other poster didn't understand this concept.


Who, me???