156 réponses à ce sujet

[dbushik]

Lexa_D wrote...

I test the hypothesis of random distribution and see evidence against it. My explanation might not be the right one, but the clustering observations persist.

hudder92, it's not a question of understanding how an RNG or PRNG works, it's a question of putting it to the test of actually doing that, as Lexa-D is saying, so no need to start assuming people don't know what they are talking about just because you are not hearing what they are actually saying.

Modifié par dbushik, 29 novembre 2012 - 11:09 .

[Lexa_D]

Heggy wrote...

Lexa_D wrote...

Bought 6 PSPs while shopping for Huntress (durr broken cloak), zero black. That's 12 gold cards more. In between this and previous posts bought 2 separate PSPs - no blacks too. So the current streak goes at least for 26 non-URs. If it was completely random with 14% probability of success, the odds of such outcome would be 2%. Stay tuned.

For 26 continuous non-URs, assuming 13 premium spectre packs, with a 14% chance of a UR inside any one pack, the probability of your streak is (0.86)^13 =0.14, not 0.02.

Umm, nope. Each PSP has 2 gold cards and you can get UR instead of EACH of them - happened to me (too) few times. So it's (0.86)^26

[Tarelgeth]

Ramsutin wrote...

Ah so OP won't accept anything but his own word. I'm out.

He's posting a 'conspiracy theory' thread.  Did you expect logic and facts to work?

[Lexa_D]

hudder92 wrote...

Some people just do not understand what a Random Number Generator is, or how they work. There is no way to cheat the RNG. What you speak of is just total luck. It has nothing to do with a pattern. An RNG resets after every run of the algorithm.

Have you ever seen a slot machine do two jackpot payouts back to back? I doubt it.

I have spent real money on packs and the drops rates aren't any better or worse spending real money. I have received 5 UR's in a row, and I have spent $50 and not received a single UR.

Yes, but what constitutes a run? Each purchase? Empirical evidence does not support it. Each store visit? In my experience - more like it.

Spending real money on PSPs? Ouch, that must've hurt.

[Lexa_D]

Kommandant Schäfer wrote...

Lexa_D wrote...
Civilization IV

is that some new kind of ammo?

I see what you did there 

[Lexa_D]

Tarelgeth wrote...

Ramsutin wrote...

Ah so OP won't accept anything but his own word. I'm out.

He's posting a 'conspiracy theory' thread.  Did you expect logic and facts to work?

Logic? Please show me some. Probabilty theory is though a more applicable branch of mathematics here - I'm ready to be proven wrong, don't hold out on me.

[dbushik]

born2beagator wrote...

Alien Number Six wrote...

......I hate the store. The store is evil. The store breeds baseless speculation. The store is the worst thing about Mass Effect 3 multiplayer. The store has been the topic of endless threads. The store gives you M-8 Avenger upgrades in your Spectre Pack. The store is the reason new players get kicked. The store is in league with Satan. The store voted for Mitt Romney. The store supports the zombie apocalypse. The store was a close personal friend of Hitler. The store is responsible for the downfall of Mayan civilization. The store made the Roanoke colonists disappear. The store is the reason some teenage males are not getting layed. The store has led to divorce in some cases. The store killed JFK. The store keeps your soul in a Premium Spectre Pack......

No reason to add this.  Not everyone here supported Mr. Hussein Obama and its only asking for trouble.

Yeah, and not everyone thinks Hitler or killing JFK was that bad either!

But seriously, you might want to take your own advice, Six, with the biogeted "Hussien" emphasis, as it's only asking for the same kind of trouble you are warning against.

[Sethvir Fireborn]

It all comes down to space magic

[Alien Number Six]

dbushik wrote...

born2beagator wrote...

Alien Number Six wrote...

......I hate the store. The store is evil. The store breeds baseless speculation. The store is the worst thing about Mass Effect 3 multiplayer. The store has been the topic of endless threads. The store gives you M-8 Avenger upgrades in your Spectre Pack. The store is the reason new players get kicked. The store is in league with Satan. The store voted for Mitt Romney. The store supports the zombie apocalypse. The store was a close personal friend of Hitler. The store is responsible for the downfall of Mayan civilization. The store made the Roanoke colonists disappear. The store is the reason some teenage males are not getting layed. The store has led to divorce in some cases. The store killed JFK. The store keeps your soul in a Premium Spectre Pack......

No reason to add this.  Not everyone here supported Mr. Hussein Obama and its only asking for trouble.

Yeah, and not everyone thinks Hitler or killing JFK was that bad either!

But seriously, you might want to take your own advice, Six, with the biogeted "Hussien" emphasis, as it's only asking for the same kind of trouble you are warning against.

I didn't write the whole Obama thing. That was born2beagator. It wasn't a serious post. I can't believe I have to write this. Please lighten up people. Six smiles for you!
:happy::P:D

Modifié par Alien Number Six, 29 novembre 2012 - 11:36 .

[Heggy]

Lexa_D wrote...

Heggy wrote...

Lexa_D wrote...

Bought 6 PSPs while shopping for Huntress (durr broken cloak), zero black. That's 12 gold cards more. In between this and previous posts bought 2 separate PSPs - no blacks too. So the current streak goes at least for 26 non-URs. If it was completely random with 14% probability of success, the odds of such outcome would be 2%. Stay tuned.

For 26 continuous non-URs, assuming 13 premium spectre packs, with a 14% chance of a UR inside any one pack, the probability of your streak is (0.86)^13 =0.14, not 0.02.

Umm, nope. Each PSP has 2 gold cards and you can get UR instead of EACH of them - happened to me (too) few times. So it's (0.86)^26

The drop rate is about 14% per PSP, and is about 5-6% per spectre pack. In a PSP, each card then has roughly 7% chance of UR.

[dbushik]

Sorry Six. My mistake in reading the names from the quotes.

[jordie3000]

Sethvir Fireborn wrote...

It all comes down to space magic

[HolyAvenger]

I bought 4 PSPs in a row. The first three gave me just gold cards. The fourth one upgraded my CSMG.



What now, OP?

[madmaxmovies]

Adding my two cents, I just want to say, that I have experienced this. When I buy PSPs, or even SPs, and they're dropping the cards I want (not just URs, I might be fishing for particular other card types for other upgrades), I keep buying. I've experienced strings of URs in PSPs doing this 3 or 4 times. I've been in disbelief of some of the runs of UR cards I've dropped since I started following this strategy.

I used to just spend my credits all the time and take what I got, and for a long time I'd buy large volumes of packs in one hit, with no URs at all - I'd say my drop rate was truly "random". Once I noticed this pattern once, I've followed it, and my UR drop rate has been much better.

I understand all the RNG logic, and realise that it is just as likely that things are balancing out for me than there is actually any particular pattern. However, I've been following this strategy the last couple of months since I noticed it, without ever seeing this mentioned before, and it has been working for me. Seems to me that there is some bias or pattern to their RNG generator, I do agree with what one poster wrote that it is possibly grabbing a handful of numbers in one hit (to save server resources), and you buy packs until those dry up, and then it grabs another set. This is pretty much what I observe happening - when you're getting good packs, there will be 6 to 8 in a row, and then it just switches.

Not expecting anyone to convince anyone that hasn't observed this, but like I said, it appears to have been working for me, and that's how I buy my packs these days.

[dbushik]

HolyAvenger wrote...

I bought 4 PSPs in a row. The first three gave me just gold cards. The fourth one upgraded my CSMG.



What now, OP?

I'm not the OP, but nothing you experienced contradicts what he is talking about in the least.  Had you bought 4 PSPs in a row and the first one had a UR but the next three didn't, that might impact what the OP is getting at, but that's not what you did, right?

[dbushik]

So, it would be a somewhat longish term endeavor (a couples weeks say), but do we want to organize test groups? Three or four people running the "cluster" strategy and the same number buying as they earn credits? Lexa_D is saying they'll record their results and post them, I would be willing to do that also in this or a similar thread.

One group hordes credits. Once they reach $1M or so they go to purchase and they take 1 PSP. If they get no UR, they stop for the day (and then re-run the procedure again the next day). If they do get a UR they purchase 2 more. Again, if no URs they stop for the day. If they get another UR, they continue buying PSPs until their money runs out. Etc.

The other group (the control group) buys PSPs as they earn the credits. Play a match, earn credits, check to see if you have 99K for a PSP and buy one if so, repeat.

Record all the data (log PSP purchases and URs pulled), compare.

Or something like that. Any interest? If you want to define a protocol Lexa_D, I'll be more than happy to follow it and report findings.

[Shampoohorn]

dbushik wrote...

HolyAvenger wrote...

I bought 4 PSPs in a row. The first three gave me just gold cards. The fourth one upgraded my CSMG.



What now, OP?

I'm not the OP, but nothing you experienced contradicts what he is talking about in the least.  Had you bought 4 PSPs in a row and the first one had a UR but the next three didn't, that might impact what the OP is getting at, but that's not what you did, right?

This thread is really ridiculous, not in the least your statement above.

At the very least go take artificial ignorance's pack buying results and see if there is even a shred of consistency between the OP's foil hat hypothesis and a reasonably large set of data.

Modifié par Shampoohorn, 30 novembre 2012 - 03:24 .

[mrcanada]

HolyAvenger wrote...

I bought 4 PSPs in a row. The first three gave me just gold cards. The fourth one upgraded my CSMG.



What now, OP?

Keep buying.

[SweetWilly013]

This thread's still alive?

Yes they're pseudo random numbers. No they're not perfect. But a computer processor moves way too fast for you to be able to exploit it and the numbers might as well be completely random to any human, especially at the snails pace one is able to buy packs. (processors are in the GHz range, you can buy packs at fastest maybe around the 0.1 Hz,,,,,that's a 10^10 magnitude difference, good luck).

This "evidence" all sounds very placebo.

[madmaxmovies]

SweetWilly013 wrote...

This thread's still alive?

Yes they're pseudo random numbers. No they're not perfect. But a computer processor moves way too fast for you to be able to exploit it and the numbers might as well be completely random to any human, especially at the snails pace one is able to buy packs. (processors are in the GHz range, you can buy packs at fastest maybe around the 0.1 Hz,,,,,that's a 10^10 magnitude difference, good luck).

This "evidence" all sounds very placebo.

Assuming the random numbers are generated each time you buy a pack, and not cached in groups - which for various technical reasons, they might be.
 
It really all depends on exactly how BW implemented this, and to what extent they have (or don't have) controls in place to guarantee the percentages and so on. i.e. the RNG for slots at a casino will be tightly controlled and monitored to ensure the house wins 55% of the time (or whatever % the operator wants). If the slot hands you a win, chances are it isn't going to throw you another one real quick. Or it might, to keep you hooked. Or it might throw you some patterns to make you think it might, to keep you spending. These systems are very well crafted.

I doubt the ME3 store has the same level of detail or control -but it could have, if BW really cared about controlling the payouts to that level of detail. To me it seems the store system here would be much simpler, including the possibility of patterns / flaws in the system. A Casino never wants to hand out more money than is coming in, but do BW really care if the numbers get generated not in their favour all the time, and a few too many URs get handed out in some circumstances? And not enough other times? I doubt they care about controlling the payouts that closely.

Without seeing their code, it is all guesswork. I have certainly seen some behaviour that indicates the store isn't truly random, but without seeing the underlying code to the behaviour I saw on screen, it is just a guess. It could just as easily be what was meant to happen statistically - or random blind luck.

[SweetWilly013]

madmaxmovies wrote...

SweetWilly013 wrote...

This thread's still alive?

Yes they're pseudo random numbers. No they're not perfect. But a computer processor moves way too fast for you to be able to exploit it and the numbers might as well be completely random to any human, especially at the snails pace one is able to buy packs. (processors are in the GHz range, you can buy packs at fastest maybe around the 0.1 Hz,,,,,that's a 10^10 magnitude difference, good luck).

This "evidence" all sounds very placebo.

Assuming the random numbers are generated each time you buy a pack

I don't see many reasons why they wouldn't be, but then again when thinking about EA's past schemes to maximize profits, a little tin-foil hat just naturally grows on my head.

Modifié par SweetWilly013, 30 novembre 2012 - 05:25 .

[cowwy]

Lexa_D wrote...

hudder92 wrote...

Some people just do not understand what a Random Number Generator is, or how they work. There is no way to cheat the RNG. What you speak of is just total luck. It has nothing to do with a pattern. An RNG resets after every run of the algorithm.

Have you ever seen a slot machine do two jackpot payouts back to back? I doubt it.

I have spent real money on packs and the drops rates aren't any better or worse spending real money. I have received 5 UR's in a row, and I have spent $50 and not received a single UR.

Yes, but what constitutes a run? Each purchase? Empirical evidence does not support it. Each store visit? In my experience - more like it.

Spending real money on PSPs? Ouch, that must've hurt.

But anecdotal evidence does?

10/10 would read trolling again.

[Lexa_D]

Heggy wrote...

Lexa_D wrote...

Heggy wrote...

Lexa_D wrote...

Bought 6 PSPs while shopping for Huntress (durr broken cloak), zero black. That's 12 gold cards more. In between this and previous posts bought 2 separate PSPs - no blacks too. So the current streak goes at least for 26 non-URs. If it was completely random with 14% probability of success, the odds of such outcome would be 2%. Stay tuned.

For 26 continuous non-URs, assuming 13 premium spectre packs, with a 14% chance of a UR inside any one pack, the probability of your streak is (0.86)^13 =0.14, not 0.02.

Umm, nope. Each PSP has 2 gold cards and you can get UR instead of EACH of them - happened to me (too) few times. So it's (0.86)^26

The drop rate is about 14% per PSP, and is about 5-6% per spectre pack. In a PSP, each card then has roughly 7% chance of UR.

Was sure UR probabilty was told before PSPs were out and was for single card, not 2. Might need to unearth some ancient scrolls from forum graveyard. However, I'd expect this to be said for 2, as you supposedly have two independent realizations, and I doubt BW was calculating total probability.
Even if you're right, though, it does not undermine my case as a whole. The "dry" spell becomes more probable, but the "wet" seasons become increasingly more unlikely (8 URs in 13 PSPs, 9 URs in 27 PSPs), providing evidence against pure random independent outcomes.

[Lexa_D]

dbushik wrote...

HolyAvenger wrote...

I bought 4 PSPs in a row. The first three gave me just gold cards. The fourth one upgraded my CSMG.



What now, OP?

I'm not the OP, but nothing you experienced contradicts what he is talking about in the least.  Had you bought 4 PSPs in a row and the first one had a UR but the next three didn't, that might impact what the OP is getting at, but that's not what you did, right?

@HolyAvenger: nothing. I do not propose a panacea or magic rule. Notice I  never said that you can't get Black if you start from Gold - that'd be statistically moronic. What I say is that due to the tendency of URs to come in clusters seeing a UR first makes seeing more URs in the next packs more likely. It could as well be a fluke (see the OP), when the probability of UR in the quasi-random sequence used to generate outcomes for this batch of purchases is low but you get one in the first pack, and the odds of it being a fluke increase if you get no Blacks later - so see the OP, cut the spending if you don't see 2nd UR soon enough. So no, it won't impact what I'm getting at.
As for the advice, I agree with mrcanada: keep buying 

[Lexa_D]

dbushik wrote...

So, it would be a somewhat longish term endeavor (a couples weeks say), but do we want to organize test groups? Three or four people running the "cluster" strategy and the same number buying as they earn credits? Lexa_D is saying they'll record their results and post them, I would be willing to do that also in this or a similar thread.

One group hordes credits. Once they reach $1M or so they go to purchase and they take 1 PSP. If they get no UR, they stop for the day (and then re-run the procedure again the next day). If they do get a UR they purchase 2 more. Again, if no URs they stop for the day. If they get another UR, they continue buying PSPs until their money runs out. Etc.

The other group (the control group) buys PSPs as they earn the credits. Play a match, earn credits, check to see if you have 99K for a PSP and buy one if so, repeat.

Record all the data (log PSP purchases and URs pulled), compare.

Or something like that. Any interest? If you want to define a protocol Lexa_D, I'll be more than happy to follow it and report findings.

Doing that on my own, although with larger sums usually. Anyone is welcome to post their results here if they wish.