(Skip down to edit)Restore LnT back to how it was but have it divide the chance of it activating by the pellet multiplier of the weapon firing. This would fix what was broken about it. Apparently this was done for the Bee shield in BL2. So why not here?
For example, if LnT has 3 points in it, its chance is 36%. If someone is using Hellwalker which has a x10 multiplier, that’ll give each pellet a 3.6% chance to proc instead of each having a 36% chance. This would make pretty much every gun have the same chance to proc LnT. At the very least no gun sould be able to break the skill this way.
EDIT: It seems my suggestion would have actually made the chance to proc LnT decrease as the pellet multiplier increased.
So I came up with a simple solid solution. All they need is this equation:
N = 1- C^(1/P)
N is the chance per pellet
C is 100% minus the present LnT chance
P is the pellet multiplier
Say LnT is at 36% making C 64% and Hellwalker is being used (10x).
1 - 0.64^(1/10) = 1 - 0.956 = 0.044 = 4.4%
So for Hellwalker, LnT would grant each pellet roughly a 4.4% giving the weapon a 36% to succeed.
10 Likes
Piemanlee
(Goodbye friends and not friends, it's been a fun ride)
#2
This is something most could agree on, so support.
Yeah it’ll be better now but sucks that’s you still won’t have the possibility of procing the skill 2+ times in a row. My suggestion wouldn’t have allowed it to happen too often, but at least it would’ve still been a possibility.
If ricochets count, then yeah they’ll have an increased chance but this will only be true when there are multiple targets getting hit (essentially negligible in most boss fights). Even still, more bullets against more targets sounds fine to me. It may be the next best thing they could do for his mobbing capabilities with his crit build outside of restoring its original duration.
It’s a fine suggestion, but ultimately would nerf multi pellet guns substantially.
#Pellets
3LNT
6LNT
1
0.36
0.72
2
0.3276
0.5904
3
0.318528
0.561024
5
0.31176
0.540412
8
0.308126
0.529747
10
0.306941
0.526326
18
0.304865
0.520397
These would be the effective percent chances to get ammo added assuming all pellets crit for various LNT if you do this.
Meaning if, like in a normal situation, only some of the projectiles hit, then it won’t work as well.
EDIT: The ideal thing to divide by would be =18*25^(1/P-1)/(25^(1/P)-7^(1/P)) where P is the number of pellets. It equalizes the percentage chances for 6 LNT which would yield the following assuming all pellets crit:
P
f∘P
0.12
0.24
0.36
0.48
0.6
0.72
2
1.529150262
0.150791594
0.28926655
0.415425
0.529267
0.630792
0.72
3
2.082208254
0.163120743
0.307461885
0.434172
0.544399
0.639292
0.72
5
3.203300801
0.173789201
0.322529981
0.449033
0.555891
0.645486
0.72
8
4.894409919
0.18011128
0.331172199
0.457295
0.562096
0.648743
0.72
10
6.02371911
0.182270359
0.334075784
0.46003
0.564122
0.649793
0.72
18
10.54519295
0.186172654
0.339262816
0.464863
0.567669
0.651618
0.72
where here f∘P is the function given which you divide the LNT chance by.
I don’t understand how it would decrease the overall chance for shotguns. Can you show how you got your first set of values?
Piemanlee
(Goodbye friends and not friends, it's been a fun ride)
#10
1.2×10=12, let’s assume the hellwalker here, so if every pellet crits you have a 12% chance, compared to
12×10=120, which means with enough sample size you should almost always get ammo back.
That’s literally a 1,000% nerf if you divide it up like that
It’s not really a nerf but a fix. With LnT in it’s original state, shotguns had a far higher chance than what was described. You showed it yourself in your example. So the only choices GB has is to either limit how often it can proc (which is what they’re choosing), or make every gun have pretty much the same chance to proc it.
1 Like
Piemanlee
(Goodbye friends and not friends, it's been a fun ride)
#12
Making shotguns 10 times less likely to proc it wouldn’t exactly give them an equal chance.
Well not exactly 10 times less likely, but still less likely in the long run
Yup the divide thing sounds good in theory, but doesn’t quite work in practice. It results in multipellet weapons having a lower chance than single pellet - AND only if you land every single pellet.
Say something like a Hellwalker. Dividing by 10 gives each pellet a 3.6% to add 1 ammo. So each pellet effectively has a 96.4% not to add ammo. But the Hellwalker has 10 pellets so:
(0.964)^10 = 0.693
This means that if all 10 pellets connect, you would have a 69.3% chance not to add ammo, and a 30.7% chance to add ammo.
So this idea wouldnt work. Multipellet weapons need an inherently higher chance not to consume ammo anyway, because the chance of landing every pellet on crit is slim.
Ok now I see where y’all got your numbers from. Honestly I didn’t know how to calculate group probabilities until now. So yeah admittedly my option was flawed. However, wouldn’t missing pellets help out? Yes it would but the chance would still be less than a single bullet.
So I came up with a simple solid solution. All they need is this equation:
N = 1- C^(1/P)
N is the chance per pellet
C is 100% minus the present LnT chance
P is the pellet multiplier
Say LnT is at 36% making C 64% and Hellwalker is being used (10x).
1 - 0.64^(1/10) = 1 - 0.956 = 0.044 = 4.4%
So for Hellwalker, LnT would grant each pellet roughly a 4.4% giving the weapon a 36% to succeed.
Thanks for the feedback btw. I learned a bit today lol.
So the probability to get ammo added for a gun is 36%. If a gun has 1 pellet, then the probability it DOES NOT get ammo added is 1-0.36 = 64% chance.
Suppose a gun has 8 pellets. Then if one or more pellets crit, you receive exactly one bullet in your magazine. This means that the chance you do NOT get a bullet added is the probability that all crits fail to proc an ammo add, which is (1-0.36)^8 = 0.028, so you have a 97.2% chance to receive ammo back.
Now let’s suppose the probability for bullets to be added scales down by a factor of 8, then each pellet has a 0.36/8 = 4.5% chance to add ammo, so the probability that a full shot gets no ammo added is of course (1-0.045)^8 = 69.2%, which means only a 30.8% chance of getting a bullet added.
Now a single pellet gun (which is easier to get every pellet a crit since there’s only one pellet) would be at a 36% chance, so multi pellet guns get nerfed.
And while this solution is also great, where you solve for the correct pellet multiplier to make it even in the event that all pellets hit, my only concern is that the solution for N depends on both C and P. Ideally it should depend only on P and work for all C, depending on how it is coded, hence my silly and complex example I gave.
It’s rough, which is why they should ultimately try different solutions like those I detail in this thread to avoid the potential computational nightmare, depending on their idea of what LNT should be. It’s obviously OP with multi - pellet guns rn, but there are still better fixes than delays or even multipliers like this one.
Not sure what you were trying to get at in the first part of your reply. I’m assuming you wrote it before seeing my new formula? The new formula I made covers it correctly as you said “in the event that all pellets hit”. I thought a bit about pellets missing and you know what? It’s fine. The same kind of logic can essentially be applied to single pellet guns when missing shots.
Meaning, if you miss a critical hit with a single pellet gun, that shot’s chance automatically reduces to 0%. Whereas if you miss a pellet or so on a shotgun shot, that shot’s chance reduces sure, but not automatically to 0%. So the trade off is a single pellet shot is either going to be (say) 36% or 0% and you have to be precise every shot. While a multi pellet shot has more of a forgiving accuracy and can total anywhere from 0% to 36%. In other words, the multi pellet shot is more flexible, and the single pellet is more accurate. So as I said, it’s fine.
As for the last part I think you’re misunderstanding something. In my formula, it does only depend on P and work with all of C since C is a constant and P is the variable. C is the failure rate of LnT. I only represented C as a variable in the equation because there are 3 values C can be set at constantly (technically 4 if you include 100% lol). But no matter, after that value is determined by the player, it becomes a constant.
I’ll look at your solutions when I get the chance but my new one is perfectly sound, though honestly I do like the present change they made to LnT. It’s not as strong as it was originally, but it’s still strong and I’d be fine if it stayed this way.
My first reply was indeed before your new formula.
Agreed about multi pellet guns being forgiving, so it should be more fine though there are better approaches such as only allowing one check for ammo per trigger pull rather than one ammo added per trigger pull regardless of number of crits.
C is a constant which depends on points put into LnT, of which there are 6 values; it is variable with your points put into LnT. P is also a constant if you only use one gun, or a variable depending on which gun you use. “Constant” and “Variable” are two sides of the same coin. The question is when would they calculate the new solution N? Every time you pull the trigger? When you equip a new gun? I guess a simple computation like this wouldn’t have a large effect on performance, but still.
Your new solution is reasonably sound, fails on unlisted projectiles and unlisted crits such as those by megavore Hex / tediore reloads / torgue stickies, while there exist solutions that account for all of those “exploits”. It’s definitely good, but not perfectly sound. Agreed that present LnT is good.
I think you know when I said that LnT only has 3 values, I was referring to the skill points the player can manually put in it. Mods are outliers that don’t really need to be stated here.
Concerning C, it’s more of a constant than a variable since its value is more static than dynamic whereas P is the opposite making it more of a variable. Sure you can go out of your way to make C a variable (by constantly changing mods every second or so lol) and P a constant by choosing to run all the same pellet sized guns. But in their natural state one is more constant, C,and the other more variable, P. You can continue to argue this if you want but either way the formula is fine regardless.