Challenge! (At will damage)

Post/Author/DateTimePost
#42

FrogReaver

Nov 04, 2014 9:33:40

I think for ease of reading the chart can we all format them to 1 decimal place?  Seems more readable and really who cares much about more decimal places than that?  (Besides the math geek in me)...

 

By the way aswesome work achmsed!  I noticed an example calculation section was missing.  Would be helpful so others can check your work.  I also noticed you didn't justify having an ally in melee with an enemy (but that's okay because I think I already did that earlier).  Overall awesome!

 

Now I just need to go calculate my damage numbers... lol

#43

Yunru

Nov 04, 2014 9:31:29 		Also is there any reason for Multiclassing to be arbitarilly banned?
(Reply to #43)

FrogReaver

Yunru wrote:
#45

ORC_Animus

Nov 04, 2014 9:45:39

I’ve removed content from this thread because trolling/baiting is a violation of the Code of Conduct.

You can review the Code of Conduct here: http://company.wizards.com/conduct

Please keep your posts polite, on-topic, and refrain from making personal attacks.You are welcome to disagree with one another but please do so respectfully and constructively.

If you wish to report a post for Code of Conduct violation, click on the “Report Post” button above the post and this will submit your report to the moderators on duty.

#46

Noctaem

Nov 04, 2014 9:51:47

Thank you.

#47

FrogReaver

Nov 04, 2014 10:11:24

achmsed, how did you calculate sneak attack with criticals factored in (since you wont know if the 2nd attack will crit before you decide to apply it on the first attack)?

(Reply to #37)

danyc

FrogReaver wrote:
#49

akschmid

Nov 04, 2014 10:30:02

FrogReaver wrote:
(Reply to #49)

FrogReaver

akschmid wrote:
#51

Yunru

Nov 04, 2014 10:45:10 		Average sneak attack is easy to calculate. SAAvgDmg*((1-chance of missing with every attack)+(1-chance of not critting at all))
(Reply to #48)

FrogReaver

danyc wrote:
(Reply to #51)

FrogReaver

Yunru wrote:
#54

Yunru

Nov 04, 2014 10:53:59 		No, he's on about you not getting Sneak Attack if you use the -5/+10 because to get Sneak Attack (which strangely neither requires sneak nor is an attack) you need to hit.
(Reply to #53)

Yunru

FrogReaver wrote:
(Reply to #55)

FrogReaver

Yunru wrote:
#57

Yunru

Nov 04, 2014 11:05:50 		The notion ia disagree with is that any "reasonable player" will apply SA to tje first hit. For one it's the least logical course of action unless said first hit is a critical and for two, as one of those people who do wait until the last moment to SA I'm insulted at your implication that I'm not reasonable.
(Reply to #57)

FrogReaver

Yunru wrote:
#59

Yunru

Nov 04, 2014 11:08:44 		No you don't. You make two attacks, there are four outcomes.

●You crit one both hits

●You crit on the first but not the second

●You crit on the second but not the first.

●You don't crit.

Considering you choose which attack to apply SA to, the only time you don't crit with your SA is when neither attack crits. Hence the formula.

(Reply to #58)

Yunru

FrogReaver wrote:
(Reply to #59)

FrogReaver

Yunru wrote:
(Reply to #14)

Coredump00

Noctaem wrote:
(Reply to #60)

FrogReaver

Yunru wrote:
#64

Yunru

Nov 04, 2014 11:35:34 		Whut? I never once mention Sharpshooter and crits in the same sentence. Also thank you for continuing to belittle all who don't follow your style of playing a rogue. I tried being nice, onto reporting.
(Reply to #61)

Yunru

FrogReaver wrote:
(Reply to #65)

FrogReaver

Yunru wrote:
#67

Yunru

Nov 04, 2014 11:50:49

No, mine is superior, as it considers all options, and gives the highest result. Which when trying to find the upper limit of something, is best.

(Reply to #67)

FrogReaver

Yunru wrote:
(Reply to #67)

FrogReaver

Yunru wrote:
#70

FrogReaver

Nov 04, 2014 11:56:18

Yunru, I don't even know when you think a player can apply sneak attack. 

 

I don't know what you are assuming the player gets to know before he makes his decision to apply the sneak attack damage.

 

There's very little of what you have said that is clear enough to me so that I can be sure of what you are trying to say.

#71

Yunru

Nov 04, 2014 11:56:01

It's simple. If you have X attacks, you have X chances to crit. To arbitrarily decide that you can only consider the first attack is unrealistic, and thus your results are too.

 

Thus you have to consider every attack. Hence:

Yunru wrote:
(Reply to #70)

Yunru

FrogReaver wrote:
#73

FrogReaver

Nov 04, 2014 11:58:47

Yunru wrote:
(Reply to #72)

FrogReaver

Yunru wrote:
#75

Yunru

Nov 04, 2014 12:01:17

FrogReaver wrote:
(Reply to #74)

Yunru

FrogReaver wrote:
#77

Coredump00

Nov 04, 2014 12:02:52

Yunru wrote:
(Reply to #77)

Yunru

Coredump00 wrote:
(Reply to #72)

FrogReaver

Yunru wrote:
(Reply to #67)

Coredump00

Yunru wrote:
(Reply to #78)

Coredump00

Yunru wrote:
#82

Yunru

Nov 04, 2014 12:11:17

Sure it is. The only time the above doesn't result in a critical SA is when there's no critical, hence +(1-chance of not critting).

The first part is the same but for hitting rather than critting: you can sneak attack if you hit once, thus unless you don't hit at all you can SA.

 

Of course, the part ignores the situation when you hit but don't crit, then don't hit again. But that generally gets cancelled out by the law of averages.

 

Also sorry if I'm getting shorty. I blame genetics.

(Reply to #78)

Coredump00

Yunru wrote:
#84

Yunru

Nov 04, 2014 12:35:35

Coredump00 wrote:
#85

Jay_Ibero_911

Nov 04, 2014 12:18:10

Coredump00 wrote:
#86

Yunru

Nov 04, 2014 12:28:09

Phew, glad no-one quoted my maths before I could correct it.

(Reply to #84)

Coredump00

Yunru wrote:
#88

Jay_Ibero_911

Nov 04, 2014 12:35:21

Yunru wrote:
(Reply to #87)

Yunru

Coredump00 wrote:
(Reply to #88)

Yunru

Jay_Ibero_911 wrote:
#91

Jay_Ibero_911

Nov 04, 2014 12:44:56

Yunru wrote:
#92

Coredump00

Nov 04, 2014 12:44:26

So... here is the 'easy' version of the Math

 

 

You hit, so you can automatically get S extra dice of damage, or you can swing again and *try* for a crit.

 

If you need an 11 to hit

1-10 means *no* extra dice 

11-19 means 5 extra dice

20 means 10 extra dice.

 

So there is a 50% chance of getting 5 less dice from waiting, and a 5% chance of getting 5 more dice from waiting. (And a 45% chance it ends up the with the same 5 extra dice)

 

So you are 10x more likely to be worse off than you are to be better off; which leads to doing about 1/2 the SA damage by hoping for a crit on the second attack.

 

The difference between waiting and not gets less the easier it is to hit, but it *never* becomes better to wait..... (If multiple attacks, it becomes worth waiting on the first attack if it is *very* easy to hit....)

(Reply to #90)

FrogReaver

least we agree here

 

Yunru wrote:
(Reply to #92)

FrogReaver

Well done!

 

Coredump00 wrote:
#95

Yunru

Nov 04, 2014 12:49:32

Okay, assuming a rogue with Extra Attack and a bonus action attack:

 Chance that you crit first attack: 1/20 Chance that you crit on the second but not first: 19/400 Chance that you crit on the third but not the second or first: 361/8000 Chance that you crit at least once: 14.2625%

10% chance of hitting: Chance that you didn't crit on the first two and miss on the third: 3249/80000

20% chance of hitting: Chance that you didn't crit on the first two and miss on the third: 2888/80000

30% chance of hitting: Chance that you didn't crit on the first two and miss on the third: 2527/80000

40% chance of hitting: Chance that you didn't crit on the first two and miss on the third: 2166/80000

50% chance of hitting: Chance that you didn't crit on the first two and miss on the third: 1805/80000

60% chance of hitting: Chance that you didn't crit on the first two and miss on the third: 1444/80000

70% chance of hitting: Chance that you didn't crit on the first two and miss on the third: 1083/80000

80% chance of hitting: Chance that you didn't crit on the first two and miss on the third: 722/80000

90% chance of hitting: Chance that you didn't crit on the first two and miss on the third: 361/80000

 

So roughly:

14.2625% chance of getting critical sneak attack damage.

10%: 4.06% chance of not doing any sneak attack damage.

20%: 3.61% chance of not doing any sneak attack damage.

30%: 3.15875% chance of not doing any sneak attack damage.

40%: 2.7075% chance of not doing any sneak attack damage.

50%: 2.25125% chance of not doing any sneak attack damage.

60%: 1.805% chance of not doing any sneak attack damage.

70%: 1.35375% chance of not doing any sneak attack damage.

80%: 0.9025% chance of not doing any sneak attack damage.

90%: 0.45125% chance of not doing any sneak attack damage.

 

So yes, I will say I was wrong that it's the sensible choice for a pure rogue. It is, however, undeniably the right choice if you've Extra Attack.

(Reply to #95)

Coredump00

Yunru wrote:
#97

Yunru

Nov 04, 2014 12:57:12

You're right, the maths looks wrong. The three attack maths though, I know is right, except maybe the final step.

your chance of wasting Sneak Attack with three attacks is: (chance of not critting)*(chance of not critting)*(chance of not hitting), or (361/400)*(chance of not hitting)

with two attacks it's: (chance of not critting)*(chance of not hitting), or (19/20)*(chance of not hitting)

 

Probably shouldn't do it late at night but I will get this right.

#98

Jay_Ibero_911

Nov 04, 2014 12:57:26

FrogReaver wrote:
#122

Raegoul

Nov 04, 2014 14:45:30

Coredump00 wrote:
#123

Yunru

Nov 04, 2014 14:52:02
Coredump00 wrote:
(Reply to #121)

FrogReaver

Drycanth wrote:
#125

Noctaem

Nov 05, 2014 5:17:08

*deleted*

(Reply to #125)

Yunru

Noctaem wrote:
(Reply to #126)

FrogReaver

Yunru wrote:
#128

Yunru

Nov 04, 2014 16:47:46 		Obviously 5% of the time the sneak attack will be a critical. The average damage is worked out independently of the chance to hit, as it always has been. Timesing the accuracy by the average damage gets you the DPR. Thus all that's need is the chance to hit, which my equation replies. To do otherwise is as pointless as reading every thread you've posted in devolve into an argument. Blocked.
#129

Varisthenes

Nov 04, 2014 16:54:04

I was thinking about this the other day.

What's the math on the break even point for a Halfling Rogue with advantage and a 19-20 crit rate?  Since they can reroll all 1s, have advantage, and a 19-20 crit rate is it worth gambling at that point if your first attack isn't a crit?

Also I do appreciate the irony of you blocking someone Yunra.

#130

Yunru

Nov 04, 2014 16:58:00

With a single reroll on a 1 the numbers get complex. You've a 1 in ten chance of rolling a crit first time and a 1in 200 chance of rolling a 1 then a crit. How that applies to multiattacks I can only guess atm (am on phone).

What really annoys me though is every time you block someone you loose your avatar.

#131

Varisthenes

Nov 04, 2014 17:05:30

Yeah, I wondered why your avatar disappeared.  There seems to be some account issues on this site.  I lost one completely due to log in issues.  When I blocked that poster in the rogue thread though I don't think I lost my avatar.

Was your earlier math on rogue crit fishing based on having advantage?  I can't really see it being viable without extra dice and the 19-20 range.

#132

Yunru

Nov 04, 2014 17:16:16 		No, my maths was based around regular rolls. That said Advantage only makes it more likely to get a crit so the numbers can still provide a bare minimum (say if external factors cancel out your advantage).

I think it only happens on mobiles with regards to losing the avatar.

#133

Yunru

Nov 04, 2014 17:20:16 		Advantage should be the same as adding an extra attack (per attack with advantage): You two rolls, both of which must fail to meet the critera in order to fail.
#134

Jay_Ibero_911

Nov 04, 2014 17:39:17

2 attacks with advantage, lucky, and 19-20 crit range, when hitting on an 11

 

hit but not crit on 11-18 or 1 re-rolled to 11-18

 

miss on 2-10 or 1 re-rolled to 1-10

 

crit on 19-20 or 1 re-rolled to 19-20

 

To find odds of hitting but not critting with advantate, take 1 - odds of missing or critting on both rolls. 1-(.45+.025+.1+.005)^2=.6636

 

So 66.36% of the time you will hit but not crit with the first attack and then have to choose to either apply sneak attack or save it and risk not using it.

 

Your chance of the first attack hitting and not critting and the second attack missing is .6636*(.475)^2=~.1497, ~15% chance of wasting sneak attack

Your chance of the first attack hitting and not critting and the second attack critting is .6636*[1-(.85+.05*.9)^2]=~.1320, ~13% chance of getting a crit by saving

 

So you almost but not quite break even here.

 

(Reply to #129)

Coredump00

Varisthenes wrote:
#136

bid

Nov 04, 2014 23:23:08

Forget the percentages, just look at the 9 cases: mm mh mc hm hh hc cm ch cc where the letters stand for miss hit crit.

 

The difference between the 2 strategies are hit-miss {you lose the SA} and hit-crit {SA gets crit damage}. Since you usually miss more often than crit, you should take SA early when you don't have advantage.

 

If you have advantage on the first attack only, you get :

- lose the SA on mh-m hm-m hh-m

- gain the crit SA on mh-c hm-c hh-c

Just as bad.

 

If you have advantage on the both attacks, you get :

- lose the SA on mh-mm hm-mm hh-mm

- gain the crit SA on mh-mc mh-hc mh-cm mh-ch mh-cc, hm-mc hm-hc hm-cm hm-ch hm-cc, hh-mc hh-hc hh-cm hh-ch hh-cc 

Comparing miss-miss to mc hc cm ch cc = 19/400, slightly better than 4/20 miss = 16/400 miss-miss.

 

If you have 3 attacks, we've just demonstrated that you should take your SA on the second attack. Therefore:

- lose the SA on (h)-mm-mm {got tired of writing mh hm hh, using (h) instead}

- gain the crit SA on (h)-(c)-x (h)-mm-(c) 

(h) is common,  (c) is still 19/400, we're comparing mmmm to (1+mm)*19/400... (m)^2 - (19/400)*(m) - (19/400) to resolve...

(m) = (19/400)/2 +/- sqrt[ (19/400)^2 + 4*(19/400) ]/2 = ~97/400 or 10/20 miss...

If you aren't missing half the time, you should take a chance with the first attack.

@ 5/20 miss, that's 4.1 - 0.3% = 3.8% extra SA damage...

 

Not worth delaying for so little gain.

 

(Reply to #134)

Coredump00

Jay_Ibero_911 wrote:
#138

Jay_Ibero_911

Nov 05, 2014 5:01:18

Coredump00 wrote:
#139

Noctaem

Nov 05, 2014 5:17:39

Yunru wrote:
#140

Varisthenes

Nov 05, 2014 8:20:59

Jay_Ibero_911 wrote:
#141

mrlong78

Nov 05, 2014 10:47:38

Class Rogue (sneak attack feature)
Race: Variant Human (Crossbow Expertise, +1 Dex)
Stats: 16 Dex
Level 1-20 sneak attack
Level 4 feat/stat -> +2 Dex, Assassin
Level 8 feat/stat -> +2 Dex
Level 12 feat/stat -> Sharpshooter

 

#142

ORC_Koph

Nov 05, 2014 17:19:32


I’ve removed content from this thread because trolling/baiting is a violation of the Code of Conduct.

 

You can review the Code of Conduct here: http://company.wizards.com/conduct

 

 

Please keep your posts polite, on-topic, and refrain from making personal attacks.You are welcome to disagree with one another but please do so respectfully and constructively.

 

If you wish to report a post for Code of Conduct violation, click on the “Report Post” button above the post and this will submit your report to the moderators on duty.
 

#143

ORC_Animus

Nov 06, 2014 10:46:43

I’ve removed content from this thread because trolling/baiting/personal attacks are violations of the Code of Conduct.

You can review the Code of Conduct here: http://company.wizards.com/conduct

Please keep your posts polite, on-topic, and refrain from making personal attacks.You are welcome to disagree with one another but please do so respectfully and constructively.

If you wish to report a post for Code of Conduct violation, click on the “Report Post” button above the post and this will submit your report to the moderators on duty.

#144

Coredump00

Nov 06, 2014 10:54:52

Sure.

 

Your approach assumes that the added Crit damage is equally distributed amongst all of the hits.... thus a percentage of hits is the same as the percentage of damage.

 

But the reality is that the damage is concetrated (doubled) on the crit roll only, and this extra damage is not effected by the 'loss' of hits for higher ACs.  Thus the 'average' damage is *different* for each AC.

 

Here is some 'pseudomath' to provide an example.

 

(The numbers used are completely irrelevant, so lets pick easy ones.)  Each hit is 10 damage, and so the Crit is 20.  Thus the average damage is 10.5

 

Since the difference is greater for higher ACs, lets look at a case where you need a 20 to hit.

 

Your method give would give a DPR of (10.5 * .05) or .525 DPR

 

But in actuallity, in 20 attacks, you would hit once for 20 points, giving an actual DPR of 20/20 or 1.0.  Which is about 90% higher than what your formulas would provide.

 

As mentioned, the difference is less as you go lower, but the absolute values are larger... I have not calculated the overal difference in the two methods yet. (just thought of it actually)

 

So if you needed an 11 to hit, your method gives a DPR of (10.5 * .5) or 5.25 DPR

 

In Actuality, in 20 attacks you would get 9 hits and 1 crit (90+20=110   110/20= 5.5 DPR.  Which is about 5% higher than your formulas would provide.

 

 

Hope that makes more sense.

 

 

edit: Of course, just about every post this relates to has just been nuked by the mods... so hopefully it still makes sense.

 

 

 

#145

Yunru

Nov 07, 2014 7:17:28 		I... think I see? Reading from a phone's not the best for maths. But where I think one of the major misunderstandings is: My equation I provided didn't include criticals. When I say criticals occur 5% of the time I mean that whatever my equation spit ou, you then add 5% of SA damage.

Interestingly for a single attack (only concerning SA damage), a critical is the same as +1 accuracy.

(Reply to #145)

bid

Yunru wrote:
(Reply to #160)

FrogReaver

Yunru wrote:
(Reply to #200)

FrogReaver

Yunru wrote:
#203

Yunru

Nov 13, 2014 2:38:55 		Yes, yes it is.

Level 1-3: 2 Attacks, +7 1d6+3

Level 4: 2 Attacks, +8 1d6+4

Level 5: 3 Attacks, +9 1d6+4

Level 6: 3 Attacks +4 1d6+14

Level 9: 3 Attacks, +5 1d6+14

Level 10: 3 Attacks, +5 advantage, 1d6+14

Level 11: 3 Attacks, +6 advantage, 1d6+15

Level 13: 3 Attacks, +7 advantage, 1d6+15

Level 14: 4 Attacks, +7 advantage, 1d6+15

Level 15: 4 Attacks, +7 advantage, 1d6+15, 1d6 SA

Level 17: 4 Attacks, +8 advantage, 1d6+15, 2d6 SA

Level 19: 4 Attacks, +8 advantage, 1d6+15, 3d6 SA