105 réponses à ce sujet

[Cryos_Feron]

Two chances never make one chance with bigger odds. It's just two chances.

true

[PatrickBateman]

Two chances never make one chance with bigger odds. It's just two chances.

Exactly, can't remember exactly how to calculate the statistical probability anymore, but it's 2 separate 7,5% probabilities in the same pack, maybe more like 10% for one pack but that's just a random number on my end.

[PatrickBateman]

wraith is one of the very best weapons in game and THE best shotgun if you want to put it on casters imo

Totally right, a lot of people believe the lolReegar is the best but that's only true when used on certain kits, best overall is definately the Wraith.

[ROBOTICSUPERMAN]

Reserves pack is the best. It has a slightly better percentage than the others, (according to the thread about actual pack results) it's not much of a difference, but anything is advantageous.

[Catastrophy]

Totally right, a lot of people believe the lolReegar is the best but that's only true when used on certain kits, best overall is definately the Wraith.

The best is the Kroguard, who is a gun and a bullet by himself, that headbutts and slaps people. He can even equip a Reegar...

[PatrickBateman]

The best is the Kroguard, who is a gun and a bullet by himself, that headbutts and slaps people. He can even equip a Reegar...

Who needs the Kroguard when there is a Batarian Brawler around?

[Catastrophy]

Who needs the Kroguard when there is a Batarian Brawler around?

wooopwooopwooop *splotch*

[niniendowarrior]

What I'm confused with the percentage is that it's referred here that you get a 7.5% chance of getting an ultra rare for each card slot on these psp, but only the last two cards ever drop ultra rares. I'm sure I'm just misinterpreting something.

[Cryos_Feron]

What I'm confused with the percentage is that it's referred here that you get a 7.5% chance of getting an ultra rare for each card slot on these psp, but only the last two cards ever drop ultra rares. I'm sure I'm just misinterpreting something.

 

7,5% per gold card

[Teh_Ocelot]

wraith is one of the very best weapons in game and THE best shotgun if you want to put it on casters imo

Couldn't care less about the Wraith and hardly ever play with it. But the first time I put the Crusader on my Flamerless GT, it was instantly one of my favorite guns. So I'd happily trade a gun I don't use or care about for a gun that's motherfucking amazing

[Cryos_Feron]

Couldn't care less about the Wraith and hardly ever play with it. But the first time I put the Crusader on my Flamerless GT, it was instantly one of my favorite guns. So I'd happily trade a gun I don't use or care about for a gun that's motherfucking amazing

 

well, it required 1200 hours for me to finally fall in love with the Wraith.

 

But everybody knows:  those relationships that develop slowly and that are not love at first sight last the longest

[Teh_Ocelot]

well, it required 1200 hours for me to finally fall in love with the Wraith.

 

But everybody knows:  those relationships that develop slowly and that are not love at first sight last the longest

Nope. Anything I wanted to use the Wraith on I'd just use the Talon. I've no room for the tin can or it's horrible sound in my boomstick rotation.

 

Having said that, I'm not knocking it, it's power, weight, or utility, I understand based on it's numbers it's certainly a very good weapon. I'm just saying, I don't use it often and don't plan to start.

[Ashevajak]

I love the Wraith, but even I will concede it sounds like a tin can, looks like something you'd expect to see Uday Hussein, Colonel Gaddafi or an especially tasteless pimp using, and calling it's fire rate "glacial" is perhaps being over generous.

[Teh_Ocelot]

I love the Wraith, but even I will concede it sounds like a tin can, looks like something you'd expect to see Uday Hussein, Colonel Gaddafi or an especially tasteless pimp using, and calling it's fire rate "glacial" is perhaps being over generous.

+1, at least you can be honest with yourself about it 

[Mgamerz]

7,5% per gold card

Per slot that can drop an ultra rare. Not all gold slots drop ultra rares.

PSP Arsenal Last 2
SP Last 1
Other IDK.

Jumbo Spectre Pack last 4.

There is so much wrong information in this thread.

[PatrickBateman]

SP have one slot, PSP/AP/RP have 2 slots where each individual slot have a 7,5% probability to drop an UR card.

[Cryos_Feron]

Per slot that can drop an ultra rare. Not all gold slots drop ultra rares.

PSP Arsenal Last 2
SP Last 1
Other IDK.

Jumbo Spectre Pack last 4.

There is so much wrong information in this thread.

I was assuming that SP, PSP, AP, RP
had 1,2,2,2 gold slots.

you say that this is not true??

my bad then, will check tonight again.

was never really watching out for colors, only knew the mentioned numbers

[Mgamerz]

I was assuming that SP, PSP, AP, RP
had 1,2,2,2 gold slots.

you say that this is not true??

my bad then, will check tonight again.

was never really watching out for colors, only knew the mentioned numbers

Idk about RP(never bought one even) but your number on num slots is correct.

[FasterThanFTL]

Commendation Packs.

[Cryos_Feron]

Nope. Anything I wanted to use the Wraith on I'd just use the Talon. I've no room for the tin can or it's horrible sound in my boomstick rotation.

Having said that, I'm not knocking it, it's power, weight, or utility, I understand based on it's numbers it's certainly a very good weapon. I'm just saying, I don't use it often and don't plan to start.

the Talon is indeed a good alternative.
however, I really hate that you always have to scope-in if you want the (strange) accuracy bonus on middle distances. that's the main reason for me to prefer the wraith

[FOZ289]

I don't really understand the Crusader.  Too much of the game going "eh, that wasn't really a headshot" and barely doing any damage, missing Husks at close range because of the way the crosshair works when firing him hip, the fact that it's a precision weapon yet isn't hitscan, and it just doesn't fire nearly as fast or reliably as the Black Widow.  I don't understand the Saber either.  When would it be more effective than using the Valkyrie or a fully automatic assault rifle, which will already be incredibly accurate on any class that can effectively use the Saber?  Feels like a mediocre sniper masquerading as an assault rifle.  Maybe if the rate of fire wasn't so low.  I play a lot of Infiltrator so these two guns feel pretty underwhelming when I already have a class that can kill any mook in two headshots from across the map while invisible.  Feels like there's no point to either gun.

[justinman114]

There are specialized pools like gold and goldconsumable goldmp4gun etc. If you max out goldmp4gun why would it ever choose that pool?

The background color of the card has literally nothing to do with its drop rate. See the N7 characters.

You may be right, but why don't you ever get IV equipment before maxing rare guns? They seem to basically replace it.

[Fuenf789]

Since Fermat was brought up it's time to look at Comradeshepard's BSN proof and see for yourself whether it's correct.

Mmm. Ok

scanned through it again (only the AP raw data):

Some minor irregularities :
1) data samples - a number of samples doesn't meet the n*p sampling criteria rule of thumb. Considering a theoretical drop rate as low as 7.5% we should have "at least" 2* (1/7,5%) >=26 consecutive samples per member before it counts as a representative submission. (To be sure I would make N=54)

In common english it means that small samples (<26) are noise and should be thrown out .

2) outliers - at least 2 submissions deviates significantly from 80% of the samples ( far outside the 9% margin of error =2*St.dev). Simply put - a minority of the submissions contains jokers/ are substance abusing / plain confused . Those 2 should also be thrown out.

3) looking at the drop rate probability of the large samples, e.g from Firerider100, it's significantly larger , in total, for 2 slots in AP, than one would expect from our current 7.5% 0% correlance assumption. It means that the two slots combined have higher probability and possibly also correlated ,...if so, much higher than the 10,5 guestimate that patrick bateman (and also my gut feel) suggests.

4) looking again at the stats, I realized that the buying "instance" behavior was not considered: For a number of some large samples the consecutive drops are far beyond what our hacked variable, suggests. Ie it was never tracked whether a bsn'er spent 2m credits over 4 weeks or in 2 minutes. The slot independence calculation was from a small population set, and drop rate probabilities do not match(correlate) with the larger samples.

5) it would really make sense if some aspiring doctorate verified the calculations again, just 2 b sure.


Ps. Hence I removed my table earlier in this thread, which were based on the 7.5% asumption, to recalculate based on larger actual samples. Since the larger sets have only a 9% error margin between themselves - I belief them to carry significant weight , together. From the raw data , ito very large batches, some values are significantly higher than anticipated, across the samples- we have to ask some of the submitters to tell us whether short timespan indeed played a role...

[Catastrophy]

Mmm. Ok

scanned it through it again (only the AP raw data):

Some minor irregularities :
1) data samples - a number of samples doesn't meet the n*p sampling criteria rule of thumb. Considering a theoretical drop rate as low as 7.5% we should have "at least" 2* (1/7,5%) >=26 consecutive samples per member before it counts as a representative submission. (To be sure I would make N=54)

In common english it means that small samples (<26) are noise and should be thrown out .

2) outliers - at least 2 submissions deviates significantly from 80% of the samples ( far outside the 9% margin of error =2*St.dev). Simply put - a minority of the submissions contains jokers/ are substance abusing / plain confused . Those 2 should also be thrown out.

3) looking at the drop rate probability of the large samples, e.g from fireraider100, it's significantly larger , in total, for 2 slots in AP, than one would expect from our current 7.5% 0% correlance assumption. It means that the two slots combined have higher probability and possibly also correlated ,...if much higher than the 10,5 guestimate that patrick bateman (and also my gut feel) suggests.

4) looking again at the stats, I realized that the buying "instance" behavior was not considered: For a number of some large samples there consecutive drops far beyond our hacked variable, suggests. Ie it was never tracked whether a bsn'er spent 2m credits over 4 weeks or in 2 minutes. The slot independence calculation was from a small population set, and drop rate probabilities do not match(correlate) with the larger samples.

5) it would really make sense if some aspiring doctorate verified the calculations again, just 2 b sure.


Ps. Hence I removed my table earlier in this thread, which were based on the 7.5% asumption, to recalculate based on larger actual samples. Since the larger sets have only a error 9% margin between themselves - I belief them to carry significant weight , together. From the raw data , ito very large batches, some values are significantly higher than anticipated, across the samples- we have to ask some of the submitters to tell us whether short timespan indeed played a role...

It is First Contact. You are approached by a blue spectacled alien. It's an Asari mathematician. The Asari expresses interest to discuss number theory with you.

[Mgamerz]

You may be right, but why don't you ever get IV equipment before maxing rare guns? They seem to basically replace it.

You can get those before you max your rare guns. Once your rare guns are maxed the weight of those consumables vs the total pool weight is significantly shifted to it.