Silence the Voices theoretical statistical values

Borderlands 2 Krieg the Psycho
#1

I have seen a lot of controversy over Silence the Voices now im not about to say for a fact its not
12% chance however if it is i can say i have extremely terrible luck with it as any test i have done has been around 25%, but what i wanted to share is repeated hitting theoretical values and why you should almost never hit yourself more then twice in a row

1 Like
#2

Most numbers are rounded off after 4 decimal places, all final odds are rounded to a whole number

Silence the Voices statistical theory

12% chance

2 times in a row = 12(0.12)= 1.44
Odds = 100á(12(0.12))= 69 meaning in 69 2 swing combos there is one chance of self hit 2 times in a row

3 times in a row = 12(0.12^2)=0.1728
Odds = 100á(12(0.12^2))= 579

4 times in a row = 12(0.12^3)=0.0207
Odds = 100á(12(0.12^3))= 4 823

5 times in a row =12(0.12^4)=0.0025
Odds = 100á(12(0.12^4))= 40 188

Silence the Voices melee theory

88% chance to melee attack

17 times in a row= 88(0.88^16)= 11.3817
Odds = 100á(88(88^32)) = 9
meaning odds are 1 in 9, 17 swing combos should be all melee attacks

149 swings, based off formula not rounded numbers 149á69 gives the number of 2 self combos you should see between 1 melee combo, if my math is right
1, 2 self hit combo

31 times in a row = 88(0.88^30)=1.9009
Odd = 100á(88(0.88^30))= 53

1631 swings
1, 3 self hit combo

47 times in a row= 88(0.88^46)= 0.2459
Odds = 100á(88(0.88^48))= 407

19 117 swings
1, 4 self hit combos

64 times in a row= 88(0.88^63)=0.0246
Odds = 100á(88(0.88^63))= 3574

228 712 swings
1, 5 self hit combo

1 Like
#3

Now that all being said it does take a lot more swings to get a huge not hitting yourself combo however odds are odds and in theory you are more likely to see an 82 hit combo then a 5 self hit combo, i know personally i have seen a few more then 20 hit some stretching to even 33-34 but i have never seen anything near 82 and i have seen a good dozen 5-6 self hits and hundreds of 3 hits, my Kreig is only level 60

Edit more likely to see a 63 hit combo then a 5 self hit combo

#4

I’m not a statistician, so can’t opine on your math. I’m just going to gripe that the self-hitting aspect of StV is the most annoying part of BL2 and wish they’d just get rid of it.

#5

It isnt accurate im tired right now going to edit it a bit cause it should include how many swings it would take

#6

Oh if only, this would be so beautiful! Sadly, I’ve downed myself more often with self-hit combos than I can recall.

I disagree, I think it fits perfectly with Krieg’s high-risk-high-reward and overall lunacy playstyle. I wont deny that it’s irritating almost beyond comprehension, but I think it adds to the fun of playing Krieg. One of the most common Krieg discussions I’ve seen/heard is, to StV or not to StV, RNGesus did not favour Krieg but it does make him a hell of a lot of fun (and frustration) to play. I couldn’t imagine how absurdly powerful a melee-based Krieg would be with the full StV benefits but without the self-hit chance.

3 Likes
#7

Statistic should now be accurate

#8

Now what i kind of think is the RNG for it is broken (which seems obvious), i think its a 12% chance after a melee attack but more of a 35% after a self attack, as i know for a fact i have never hit more then the mid 30s in melee attacks and only got there a handfull of times, however i have hit probably close to a dozen 6 self hits in a row… which should only happen once every 2 000 000 swings or so.
Any thoughts?

#9

Without actually keeping a count, you wouldn’t know if it was selection and / or confirmation bias. Your theoretical numbers are more likely accurate, but there’s a lot of broken stuff in the game so it’s possible that could apply with an STV self hit chain, but no way to know without real (boring) data to back it up.

#10

I strongly believe it is more likely to hit yourself if you just hit yourself, the biggest point everyone brings up for “its more like 30%” or “12%? More like 75” is self hit chaining, and the most common argument for “its 12% you just remember the bad times” is that people have had tons of 30 some chains of no self hitting. Now i agree that sometimes it does feel like its more like 20-30% very frequently, as i said before my krieg is only 60 ao i dont have a huge pool of ‘testing’. But other times it feels like the 12% people are right because it you can get a great 35 hit chain going, then get double hit then a few more melee attacks so on.

#11

It’s entirely possible, there’s a lot of weird, bugged stuff in the game. It’s just that it would be very hard to test. You’d basically have to set some key bindings so you could just hold down a button to melee repeatedly and then you’d have to have a counter just for hits after the first self hit, instead of the usual self hit. I haven’t noticed any weird patterns in double hitting or chaining self hits, but again, without any data we’re both just speculating.

#12

It’s 12%. It just seems more than that when you are pissed after a chain or have downed yourself witb it.

(Not my vid btw)

4 Likes
#13

All of your math is correct, but one thing that you’re forgetting in your analysis is the memory-less property of this distribution. Obviously, all melees are independent, so P(Hitting Yourself of nth melee | Hitting Yourself on (n-1) the melee) = P(Hitting Yourself of nth melee | Not Hitting Yourself on (n-1) the melee).

The best estimate of the true mean of the actual probability is to just take total times you hit yourself divided by the total times that you press melee.

#14

I wasnt looking calculating the percent of times you hit yourself, i was calculating the odds of multiple hits in a row and 12(.12^n) gives you the odds, incorperating the ‘fact’ that each melee has a 12% chance to hit yourself.

#15

Now is 300 swings really an adequate number to claim withought a doubt it must be 12%?
If you posted a video of 10 000 swings and he hit himself 1150- 1250 times that would be an adequatetest size to say with confidence its about 12%.
One time i ran around sanctuary meleing for around 45 minutes, now i dont have exact numbers i was not keeping a tally or recording it, but i was trying to get 7 melee for every self hit which works out to 12.5% and i got in around 1 for every 5 give or take a little, so 16-17%, however what made it so high was chains, now i know people are going to dismiss this as there is no proof. However as i have said i do believe it is 12% chance, so long as its not after hitting yourself already with my expirience on krieg, he has a higher chance to hit himself if he has just hit himself rather then if he just melee attacked, it appears to be more of a 20-30% chance of hitting himself if he has just hit himself.
Then again maybe thats from mashing melee while the self hit animation is active?

#16

I’d still trust his results over yours since you weren’t taking a tally. Without hard numbers results are extremely easy to be skewed in your head, especially if the sample size is a thousand and you are just making an approximation. What you felt was ‘1 in 5’ could easily be ‘1 in 7’ or ‘1 in 8’. Garnering 12% from 300 isn’t ideal, but actual video evidence is, for me, superior to any gut feelings I have seen from people who claim otherwise.

Mashing the melee button faster than Krieg’s actual melee “fire rate” is obviously going to lead to more STV procs than normal. If you melee 12 times but hit the button 18, you now have an 18% chance to STV, not 12%.

1 Like
#17

I knew you would before you said that, but i never once said i felt like it was like 1 in 5 i said it was 1 in 5 give or take a little

#18

We have a winner. This is indeed the case. A lot of research was done on this back in the day, including what is commonly perceived as ‘chains’.

It’s 12%, every time. It just feels like more when it happens.

2 Likes
#19

So you’re saying it counts every time you press the button, not every time a melee/self attack animation happens?

#20

Yeah, pretty much that.

Don’t hammer it, instead, try to get into a rhythm.

1 Like