yeah, while I applaud his efforts, there is still issues with that kind of thinking.
When you have a lottery, there is literally one prize, only one number that will win, sure multiple people can get that number, but there's only one drawing, so you can calculate the odds somewhat reliably.
In this case however, there is no set number of prizes, Cryptic hasn't decided that they will only award X JHAS. Every time you open a box, the game does a new drawing, in which the bug is only one of a dozen or more possible prizes. That means that the odds are really only one in however many prizes are possible. But that is meaningless, because it does a new drawing every time you open one.
yeah, while I applaud his efforts, there is still issues with that kind of thinking.
When you have a lottery, there is literally one prize, only one number that will win, sure multiple people can get that number, but there's only one drawing, so you can calculate the odds somewhat reliably.
In this case however, there is no set number of prizes, Cryptic hasn't decided that they will only award X JHAS. Every time you open a box, the game does a new drawing, in which the bug is only one of a dozen or more possible prizes. That means that the odds are really only one in however many prizes are possible. But that is meaningless, because it does a new drawing every time you open one.
But like a slot machine it is set at a payout level. So it is fairly accurate (within 1-2%). Over a massive sample size you will see it even out. But people opening a 100-200 is not near enough. A thousand isn't either.
Today I received the Bug ship on Doff pack 90. (I opened 43 packs today and had Excel records from previous attempts during previous promotions amounting to another 47.)
Just because you received the ship upon opening the 90th pack, that doesn't mean the probability of winning one is 1/90. The probability may be 1/200, and you were just lucky; or the probability may be 1/50, and you were unlucky. It is true that if the probability were 1/90, then the average number of boxes you would have to open before seeing one ship is 90. Please note the emphasis on the word "average". See the geometric distribution for details.
Looking at your premise that the chance is actually 1.1%, where that is from is no matter lets agree that it is solid.
Wouldn't opening 412 packs at that rate give you only a 4.8% chance of getting A bug ship?
And a 94 point blah blah blah % chance of getting MORE than one bug ship?
Just asking.
I think he meant at least one ship, not exactly one ship. Suppose that the probability of winning a ship from opening a lockbox is 0.011. Then, the probability of not winning is 0.989. The probability of winning no ship from opening 412 boxes is 0.989^412 = 0.0105. Thus, the probability of winning at least one ship from opening 412 boxes is 0.9895. The probability of winning exactly one ship from opening 412 boxes is 412 * 0.011 * (1 - 0.011)^411 = 0.04808. More generally, the probability of winning exactly k ships from opening n boxes is (n choose k) * 0.011^k * (1 - 0.011)^(n-k). See the binomial distribution for details.
And this is the thing ... There is no such thing as a "truly" random number in computing ... The RNG can 'simulate' a random number, but it still requires fixed parameters to work with, so you have to ask (and Cryptic will NEVER tell) when those parameters are set?
No, the pseudorandom number generators used by computers aren't truly random, but they may be good enough for your purposes, depending on what those purposes are. Different applications have different needs. For example, cryptography has stronger requirements than most other applications, and most cryptographic libraries come with their own pseudorandom number generators, because the ones provided by most standard libraries are too weak; the PRNGs that come with standard libraries are meant for general applications, not for cryptography. Another thing to bear in mind is that many mistakes with PRNGs have more to do with improper usage than with weaknesses in the PRNGs themselves; for example, mistakes in how they're seeded, how they're called, and what's done with the values they return.
You fall into the same false trap that many other "number crunchers" fall into ... You seem to think that Cryptic has some Whiteboard somehwere, ticking off how many boxes you've opened, and when you get enough "ticks" you'll get one ... WRONG!!!
The simple fact is, that the RNG is coded to set parameters on (I hope anyway) each and every Lockbox openeing, that means that each and every time you open a box, the ONLY odds or ratio is 50% or 50/50 period!
However if the parameters are set differently, who knows WHAT might happen ... As I said, I've had 3 ships drop from 5 packs consecutively... and I've had 0 drop from 10 consecutively ...
This is an oft repeated fallacy that many seem to believe ... You cannot put meaningful odds or use statistics from "past" events to determine the likely outcome of future events in a random (or close approximation of random) system ... It just doesn't work that way!
You can however change the parameters of the system, which is what Casino's and Lotteries do. (and I hope Cryptic doesn't do)
Instant Scratch prizes for instance are a good example ...
Say 1 in every 10000 tickets has a major prize, this can be set before any tickets get printed, so the odds of anyone specificallyfinding the ticket is 1 in 10000, but the odds of "someone" winning the prize is 100% and every time you scratch one you have precisely a 50% or either winning or losing (unless of course the prize was won yesterday, in which case you now have 0% chance ...
You can change the original parameters by printing more tickets i.e. 20000, 30000, 100000 and the odds of 'anyone' will change, but note that this has absolutely NO effect on your odds of winning.
I'm not entirely sure what you are trying to say. You sound like someone without training in computer science or probability and statistics. Maybe, someone told you about pseudorandom number generators and the gambler's fallacy. Or maybe, you took one programming class and one class in probability and statistics, but failed to completely grasp what they were teaching you.
If I understand you correctly, then you seem to think that each lockbox is hard-coded with a prize, which may or may not be a ship. That is, the box either does or does not contain a ship---you just don't know which. While Cryptic can implement lockboxes this way, they probably don't because it requires much more storage than a simpler solution. It is more likely that every lockbox is exactly the same. The act of opening a lockbox executes a piece of code that uses a pseudorandom number generator to simulate rolling a many-sided die. The result of the die roll determines what prize you win.
Even if each lockbox were hard-coded with a prize, like physical boxes, it would still be meaningful to talk about the probability of winning a ship from opening a box. Yes, either the box contains a ship, or it does not. But you don't know which before opening the box, and it is meaningful to talk about the probability of an event that has not yet occurred. This discussion is starting to venture close to arguments over the interpretation of probability, such as frequentist vs. Bayesian. I'm not well-versed in such arguments, so I am not going to discuss them here. In any case, even if lockboxes were physical boxes and which box you received were determined by some physical process like a raffle, the probability of winning a ship would not be 50%, but whatever probability Cryptic has set.
A method for estimating the probability of winning a ship
To estimate the probability of winning a ship, you can open many boxes, count the number of ships you receive, and then divide the number of ships received by the total number of boxes opened. The problem is figuring out how many boxes you need to open in order to get a good estimate with high confidence. Let p denote the probability of winning a ship from opening one lockbox. Suppose you want to estimate p to within 0.01 with 95% confidence. Let N denote the number of boxes opened and X denote the number of ships received. Then, Hoeffding's inequality gives
Pr[|X/N - p| >= 0.01] <= 2 * exp(-2 * N * 0.01^2)
We wish bound the probability that our estimate X/N differs from p by 0.01 or more. To get an upper bound of 5% on this probability, we solve the following equation for N:
2 * exp(-2 * N * 0.01^2) = 0.05
Solving the above equation gives
N = ln(2/0.05) / (2 * 0.01^2) = 18444.4
Thus, we need to open at least 18445 boxes. The values 0.01 and 95% were chosen arbitrarily. If you are willing to accept an error greater than 0.01 or a confidence less than 95%, then you can open fewer boxes. To figure out how many boxes you need, repeat the above calculation with your desired error and confidence.
98.9%^412 is 1% (well, very slightly higher), which is to say opening 412 packs gives us a 1% chance of not winning the grand prize, which reversed is a 99% chance of winning it.
...
assuming your numbers are correct, reversing it means you get a 99% chance of winning AT LEAST ONE.
The opposite of "winning none" is "winning one OR MORE"..
assuming your numbers are correct, reversing it means you get a 99% chance of winning AT LEAST ONE.
The opposite of "winning none" is "winning one OR MORE"..
I made the same point above. This is really a case of people misunderstanding each other due to not using English with the specificity required when discussing mathematics. Admittedly, some flexibility is allowed in this area, but it takes practice to know how much flexibility to allow. There is also the additional complication that not everyone speaks English as their first language.
I made the same point above. This is really a case of people misunderstanding each other due to not using English with the specificity required when discussing mathematics. Admittedly, some flexibility is allowed in this area, but it takes practice to know how much flexibility to allow. There is also the additional complication that not everyone speaks English as their first language.
yeah your post kind of settled the matter but i feel its less a problem of language and more about ppl misunderstanding how randomness/chances work once more than one throw of the dice is done.
and i seem to have overlooked the upper part of your post so thats why i posted what i did
yeah your post kind of settled the matter but i feel its less a problem of language and more about ppl misunderstanding how randomness/chances work once more than one throw of the dice is done.
Some people in this thread display a poor understanding of probability. But I feel that the specific disagreement between thissler and walshicus is due to imprecise wording. That kind of disagreement can be easily solved by making both parties aware of what the other meant.
A bit late to the table but from my limited knowledge of eastern run games (mainly from Asian friends connected to the industry) when it comes to the gambling mechanic and lock/treasure etc boxes. It is usually set to 0.01% for the top prizes, since they run the business and it falls under their rules it does not have to comply with western gambling laws.
To give them their due I reckon the top prizes here are set at 0.1% !
Hello, this is your fellow math teacher here. Let's learn a little about probability.
No one knows the exact odds of winning of a JHAS ship from a promo pack. For this example we will use 1.1%. Now, ALL STO promo backs, lockboxes, etc are independent events meaning their probabilities do not rely upon a previous outcome. Hence why you can get two ships in a row.
When a probability is independent of each other you multiply the outcomes together to get the combined probability. So, if you opened 400 promo packs your combined probability of getting ONE JHAS is (1.1% * 1.1% * 1.1% * .... nth amount of packs you open = 1.1%^nth, where n is the number of times you open packs.
So, if you opened 400 packs it would be 1.1%^400=3.606401e-784 chance
Hello, this is your fellow math teacher here. Let's learn a little about probability.
No one knows the exact odds of winning of a JHAS ship from a promo pack. For this example we will use 1.1%. Now, ALL STO promo backs, lockboxes, etc are independent events meaning their probabilities do not rely upon a previous outcome. Hence why you can get two ships in a row.
When a probability is independent of each other you multiply the outcomes together to get the combined probability. So, if you opened 400 promo packs your combined probability of getting ONE JHAS is (1.1% * 1.1% * 1.1% * .... nth amount of packs you open = 1.1%^nth, where n is the number of times you open packs.
So, if you opened 400 packs it would be 1.1%^400=3.606401 or 360% chance
When calculating with probabilities, you need to use the decimal version of the value (remember, "%" basically means * 1/100) . 1.1 % = 0.011. 0.011^400 would be a very low number, and represents the chance of winning 400 times in a row. That is not what you are after.
There are two things you can check for plausability:
1) If an outcome has a less than 100 % chance to happen, then no matter how often you try, the chance for it occurring in a sequence can never exceed 100 %, because there is always a possible sequence where you stay "unlucky" the entire time.
2) Try your math with a higher probability and see if that makes sense. (For example, use the standard six-sided die. Does rolling 6 times guarantee a 6?)
What you need to do for this is, as others pointed out before, is to look at the probability of the event not happening in a series of events. You want to find out how likely it is for you still loose after n tries.
If your probability for the outcome you desire is 1.1 % (or 0.011), then a non-desirable outcome happens with a probability of 0.989 (or 98.9 %).
For n = 400, it's 0.989^400 ~= 0.012 or 1.2 %. That means the chance for not getting the deisred outcome is only 1.2 % after 400 tries, or the chance for the desired outcome after 400 tries is 98.8 %.
That means after 400 tries, you have a fairly good chance of winning your bug. But it's still not guaranteed, and it never will be.
Regarding computer-generated probabilities vs "real" probabilities. One of the problems of computer-generated randmization is that it's ultimately predictable, because it's still a deterministic algorithm. But if you do not know the "seed" (often the compter clock time is used initializing a randomizer), you cannot really predic the result, so it's pretty much equal to a pure random distrubtion. Unless the randomizer does favor certain results inherently. (Say, if the randomizer would usually favor numbers around a certain value, that value may be determined by the seed or even inherent to the algorithm). That would be somethnig that would probably be a visible pattern in a statistical analysis of the randomization results.
I have no idea if that is in fact a problem with common randomizers or not.
Star Trek Online Advancement: You start with lowbie gear, you end with Lobi gear.
See, that is why statistics never work in real life. If you have 100 boxes, then statistics would tell you that the first box is a 1% chance, but the 100th box is a guaranteed ship, which is simply not true. In real life, the first box is only a 1%, the 20th box is still 1%, the 50th box is 1%, and the 100th box is still only a 100%. So, you very easily can (and likely will) open hundreds and hundreds of boxes without ever getting a single ship.
See, that is why statistics never work in real life. If you have 100 boxes, then statistics would tell you that the first box is a 1% chance, but the 100th box is a guaranteed ship
I am not sure what stats books you read, but 'no' statistics does not tell you this.
What statistics does show me is that Cryptics RNG is not fully random - it is streaky and can be 'broken'. I have seen people pull 3 and 4 lockbox ships in immediate succession (ie 4 ships in 10 secs). What statistics tells you is that the odds of that happening are not astronomical, they are much, much larger than that.
In fact what Statistics tells you is that based on a null hypothesis H0 = RNG is purely random, the evidence (and from this I am taking logs of players winning lockbox ships, personal loot drops, and crit rolls from the opvp logs) shows that you can be certain to 5 degrees of certainty that H0 is false. RNG is not fully random, and varies from character to character, and from time to time.
It is worth noting that in most areas of science and mathematics, once a theory obtains 4 degrees of certainty (that is 99.999% likelihood based on evidence) that it is assumed as FACT.
But I myself (honestly) have opened more than a hundred Dominion Lock Boxes, and all I got was duty officer packs and mining claims. And then last week, I dumped 400k+ Dilithium into buying Romulan duty officer packs, and never got a Bugship. If statistics were true, then I should have gotten at least 2-3 ships by now.
I work in the financial sector so I use Excel for everything related to PWI both on STO as well as NW. I keep really detailed records of every chance based thing purchased. The $250 or 1:250 per top prize is a very common theme on both of the games across most of the lottery based prizes.
When the Bug ship first came out, it was 125 Zen ($1.25 USD) to get a winter box and reports were that it was 1:250 boxes to get a ship. Now since the Doff packs are 275 Zen ($2.75 USD) and the odds have been adjusted to account for this, it takes 90.9 Doff packs to get 1 Bug ship.
Today I received the Bug ship on Doff pack 90. (I opened 43 packs today and had Excel records from previous attempts during previous promotions amounting to another 47.)
90 Doff packs at $2.75 USD is $247.50.
I think this information is useful for people to know because it helps them make a more informed decision prior to engaging in a lottery mechanic.
I haven't opened any lock boxes on STO, but I have on Neverwinter. Results on that game confirm the general 1:250 premise.
284 Keys
1 Thayan Book of the Dead
0 Giant Beetle
0 Legendary Beetle
109 Keys
0 Apparatus of K
0 Rust Monster
920 Keys
0 Owlbear
4 Puesdodragon
380 Keys
1 Nightmare
TOTAL KEYS: 1,693 ($2,116.25 USD)
TOTAL PRIZES: 6
Odds: 1/282.
Could have just posted enough keys on the exchance to flat out buy a bug ship...
That is just it exactly. The only way most of us will ever get a ship is to simply buy it. But in all honesty, you will get that ship much faster, and for a great deal cheaper, if you sell the keys for EC. Here is the arithmetic.
* The Jem'Hadar Dreadnought costs 800 Lobi.
* With 4 Lobi per box opened, 800 Lobi is 200 keys.
* Keys cost 125 Zen each.
* 200 keys is the equal to 180 or $250, which is the cost of a Lifetime Subscription.
* So your Jem'Hadar Dreadnought just cost you a Lifetime Subscription.
* Now you look on the Exchange, where you find the Jem'Hadar Dreadnought Carrier is 100 million.
* 53 keys sold will get you that 100 million, with some left over for duty officers.
* Your Jem'Hadar Dreadnought just cost you 48 or $66.25.
Moral of the story? Always go the keys to EC route.
All I can say is this is sad really and a shame to be honest.
What should be the case it that for the investment of what you can buy a zen ship for maybe even twice that, you should be able to acquire the rare ship drop from any lock box. So for an inverstment of say $50 in keys you should be damn near at like 99.9% chance of getting one. Hundreds of dollars on average, if you don't get amazingly lucky, is monumentally ridiculous and a bit of a barrier to playing the game in my opinion.
From the perspective of a new player, wanting to get into playing a Elachi Monbosh and S'golth at some point, just knowing that I am stuck either grinding 180+ million EC or spending hundreds of dollars is really disappointing. Honestly I would much rather have a monthly subscription with more open accessibility to things like exotic ships. Free-to-play differently isn't the best MMO model by far.
Hello, this is your fellow math teacher here. Let's learn a little about probability.
So, if you opened 400 packs it would be 1.1%^400=3.606401 or 360% chance
And they claim cannabis has no ill effects on the human brain.
Sanity checks have always been a good tool. If you calculate a probability of 360% then maybe, just maybe, you've done something wrong. Glad you weren't teaching any of my math classes :P
Sanity checks have always been a good tool. If you calculate a probability of 360% then maybe, just maybe, you've done something wrong. Glad you weren't teaching any of my math classes :P
no, the math is right but i realize the website did not upload the full answer. I originally used an editor called Math Type to input it so the full answer would be there. However the short answer is, its 3.60e-784% but that is the odds of getting a ship every time.
Just getting a ship ONCE would be (P)=(# of favorable Outcomes)/(# of total possible Outcomes). i do not know the total list of items you could possibly win from a lockbox but what it will show is your odds will not change significantly just because you open more packs. Though others like to state that over time you "should be guaranteed" a ship due to the improbability of not getting a ship but that is not how this works. What you find is that as you open more packs and add more favorable outcomes you have to add equally the same possible outcomes which in turn will keep the chance the same.
Example, if each pack has 11 chances to get JHAS out of 1000 items thats 11/1000 or 1.1% opening two packs it would be 22/2000 or 11/1000, the same odds. It is incorrect to add 11/1000+11/1000=22/1000 because from two packs you have 22 chances out of 2000 total items NOT 22 chances out of 1000 items. You can not just add the favorable outcomes from pack two but not add the total outcomes.
Comments
When you have a lottery, there is literally one prize, only one number that will win, sure multiple people can get that number, but there's only one drawing, so you can calculate the odds somewhat reliably.
In this case however, there is no set number of prizes, Cryptic hasn't decided that they will only award X JHAS. Every time you open a box, the game does a new drawing, in which the bug is only one of a dozen or more possible prizes. That means that the odds are really only one in however many prizes are possible. But that is meaningless, because it does a new drawing every time you open one.
But like a slot machine it is set at a payout level. So it is fairly accurate (within 1-2%). Over a massive sample size you will see it even out. But people opening a 100-200 is not near enough. A thousand isn't either.
Just because you received the ship upon opening the 90th pack, that doesn't mean the probability of winning one is 1/90. The probability may be 1/200, and you were just lucky; or the probability may be 1/50, and you were unlucky. It is true that if the probability were 1/90, then the average number of boxes you would have to open before seeing one ship is 90. Please note the emphasis on the word "average". See the geometric distribution for details.
I think he meant at least one ship, not exactly one ship. Suppose that the probability of winning a ship from opening a lockbox is 0.011. Then, the probability of not winning is 0.989. The probability of winning no ship from opening 412 boxes is 0.989^412 = 0.0105. Thus, the probability of winning at least one ship from opening 412 boxes is 0.9895. The probability of winning exactly one ship from opening 412 boxes is 412 * 0.011 * (1 - 0.011)^411 = 0.04808. More generally, the probability of winning exactly k ships from opening n boxes is (n choose k) * 0.011^k * (1 - 0.011)^(n-k). See the binomial distribution for details.
No, the pseudorandom number generators used by computers aren't truly random, but they may be good enough for your purposes, depending on what those purposes are. Different applications have different needs. For example, cryptography has stronger requirements than most other applications, and most cryptographic libraries come with their own pseudorandom number generators, because the ones provided by most standard libraries are too weak; the PRNGs that come with standard libraries are meant for general applications, not for cryptography. Another thing to bear in mind is that many mistakes with PRNGs have more to do with improper usage than with weaknesses in the PRNGs themselves; for example, mistakes in how they're seeded, how they're called, and what's done with the values they return.
I'm not entirely sure what you are trying to say. You sound like someone without training in computer science or probability and statistics. Maybe, someone told you about pseudorandom number generators and the gambler's fallacy. Or maybe, you took one programming class and one class in probability and statistics, but failed to completely grasp what they were teaching you.
If I understand you correctly, then you seem to think that each lockbox is hard-coded with a prize, which may or may not be a ship. That is, the box either does or does not contain a ship---you just don't know which. While Cryptic can implement lockboxes this way, they probably don't because it requires much more storage than a simpler solution. It is more likely that every lockbox is exactly the same. The act of opening a lockbox executes a piece of code that uses a pseudorandom number generator to simulate rolling a many-sided die. The result of the die roll determines what prize you win.
Even if each lockbox were hard-coded with a prize, like physical boxes, it would still be meaningful to talk about the probability of winning a ship from opening a box. Yes, either the box contains a ship, or it does not. But you don't know which before opening the box, and it is meaningful to talk about the probability of an event that has not yet occurred. This discussion is starting to venture close to arguments over the interpretation of probability, such as frequentist vs. Bayesian. I'm not well-versed in such arguments, so I am not going to discuss them here. In any case, even if lockboxes were physical boxes and which box you received were determined by some physical process like a raffle, the probability of winning a ship would not be 50%, but whatever probability Cryptic has set.
A method for estimating the probability of winning a ship
To estimate the probability of winning a ship, you can open many boxes, count the number of ships you receive, and then divide the number of ships received by the total number of boxes opened. The problem is figuring out how many boxes you need to open in order to get a good estimate with high confidence. Let p denote the probability of winning a ship from opening one lockbox. Suppose you want to estimate p to within 0.01 with 95% confidence. Let N denote the number of boxes opened and X denote the number of ships received. Then, Hoeffding's inequality gives
Pr[|X/N - p| >= 0.01] <= 2 * exp(-2 * N * 0.01^2)
We wish bound the probability that our estimate X/N differs from p by 0.01 or more. To get an upper bound of 5% on this probability, we solve the following equation for N:
2 * exp(-2 * N * 0.01^2) = 0.05
Solving the above equation gives
N = ln(2/0.05) / (2 * 0.01^2) = 18444.4
Thus, we need to open at least 18445 boxes. The values 0.01 and 95% were chosen arbitrarily. If you are willing to accept an error greater than 0.01 or a confidence less than 95%, then you can open fewer boxes. To figure out how many boxes you need, repeat the above calculation with your desired error and confidence.
assuming your numbers are correct, reversing it means you get a 99% chance of winning AT LEAST ONE.
The opposite of "winning none" is "winning one OR MORE"..
I made the same point above. This is really a case of people misunderstanding each other due to not using English with the specificity required when discussing mathematics. Admittedly, some flexibility is allowed in this area, but it takes practice to know how much flexibility to allow. There is also the additional complication that not everyone speaks English as their first language.
yeah your post kind of settled the matter
and i seem to have overlooked the upper part of your post so thats why i posted what i did
Some people in this thread display a poor understanding of probability. But I feel that the specific disagreement between thissler and walshicus is due to imprecise wording. That kind of disagreement can be easily solved by making both parties aware of what the other meant.
Whatever method you use, the odds are clearly way too expensive for me.
Unless the price is -much- lower, or the odds are -much- higher (say, 1 in 3, 1 in 4, etc) the rare ships are beyond my financial reach.
Maybe with EC someday, but that day is still a long way off.
I am grateful to those who have a lot of disposable income and support this game. I do wish PWE found a way to profit from us "poor folk".
To give them their due I reckon the top prizes here are set at 0.1% !
No one knows the exact odds of winning of a JHAS ship from a promo pack. For this example we will use 1.1%. Now, ALL STO promo backs, lockboxes, etc are independent events meaning their probabilities do not rely upon a previous outcome. Hence why you can get two ships in a row.
When a probability is independent of each other you multiply the outcomes together to get the combined probability. So, if you opened 400 promo packs your combined probability of getting ONE JHAS is (1.1% * 1.1% * 1.1% * .... nth amount of packs you open = 1.1%^nth, where n is the number of times you open packs.
So, if you opened 400 packs it would be 1.1%^400=3.606401e-784 chance
When calculating with probabilities, you need to use the decimal version of the value (remember, "%" basically means * 1/100) . 1.1 % = 0.011. 0.011^400 would be a very low number, and represents the chance of winning 400 times in a row. That is not what you are after.
There are two things you can check for plausability:
1) If an outcome has a less than 100 % chance to happen, then no matter how often you try, the chance for it occurring in a sequence can never exceed 100 %, because there is always a possible sequence where you stay "unlucky" the entire time.
2) Try your math with a higher probability and see if that makes sense. (For example, use the standard six-sided die. Does rolling 6 times guarantee a 6?)
What you need to do for this is, as others pointed out before, is to look at the probability of the event not happening in a series of events. You want to find out how likely it is for you still loose after n tries.
If your probability for the outcome you desire is 1.1 % (or 0.011), then a non-desirable outcome happens with a probability of 0.989 (or 98.9 %).
For n = 400, it's 0.989^400 ~= 0.012 or 1.2 %. That means the chance for not getting the deisred outcome is only 1.2 % after 400 tries, or the chance for the desired outcome after 400 tries is 98.8 %.
That means after 400 tries, you have a fairly good chance of winning your bug. But it's still not guaranteed, and it never will be.
Regarding computer-generated probabilities vs "real" probabilities. One of the problems of computer-generated randmization is that it's ultimately predictable, because it's still a deterministic algorithm. But if you do not know the "seed" (often the compter clock time is used initializing a randomizer), you cannot really predic the result, so it's pretty much equal to a pure random distrubtion. Unless the randomizer does favor certain results inherently. (Say, if the randomizer would usually favor numbers around a certain value, that value may be determined by the seed or even inherent to the algorithm). That would be somethnig that would probably be a visible pattern in a statistical analysis of the randomization results.
I have no idea if that is in fact a problem with common randomizers or not.
My Ship Builds: USS Conqueror, HMS Victorious, HMS Concord, ISS Queen Elizabeth, Black Widow III
Click here to view my DeviantArt.
I am not sure what stats books you read, but 'no' statistics does not tell you this.
What statistics does show me is that Cryptics RNG is not fully random - it is streaky and can be 'broken'. I have seen people pull 3 and 4 lockbox ships in immediate succession (ie 4 ships in 10 secs). What statistics tells you is that the odds of that happening are not astronomical, they are much, much larger than that.
In fact what Statistics tells you is that based on a null hypothesis H0 = RNG is purely random, the evidence (and from this I am taking logs of players winning lockbox ships, personal loot drops, and crit rolls from the opvp logs) shows that you can be certain to 5 degrees of certainty that H0 is false. RNG is not fully random, and varies from character to character, and from time to time.
It is worth noting that in most areas of science and mathematics, once a theory obtains 4 degrees of certainty (that is 99.999% likelihood based on evidence) that it is assumed as FACT.
...#LLAP...
My Ship Builds: USS Conqueror, HMS Victorious, HMS Concord, ISS Queen Elizabeth, Black Widow III
Click here to view my DeviantArt.
Could have just posted enough keys on the exchance to flat out buy a bug ship...
* The Jem'Hadar Dreadnought costs 800 Lobi.
* With 4 Lobi per box opened, 800 Lobi is 200 keys.
* Keys cost 125 Zen each.
* 200 keys is the equal to 180 or $250, which is the cost of a Lifetime Subscription.
* So your Jem'Hadar Dreadnought just cost you a Lifetime Subscription.
* Now you look on the Exchange, where you find the Jem'Hadar Dreadnought Carrier is 100 million.
* 53 keys sold will get you that 100 million, with some left over for duty officers.
* Your Jem'Hadar Dreadnought just cost you 48 or $66.25.
Moral of the story? Always go the keys to EC route.
My Ship Builds: USS Conqueror, HMS Victorious, HMS Concord, ISS Queen Elizabeth, Black Widow III
Click here to view my DeviantArt.
4800 promo packs - 29 JHAS
And they claim cannabis has no ill effects on the human brain.
What should be the case it that for the investment of what you can buy a zen ship for maybe even twice that, you should be able to acquire the rare ship drop from any lock box. So for an inverstment of say $50 in keys you should be damn near at like 99.9% chance of getting one. Hundreds of dollars on average, if you don't get amazingly lucky, is monumentally ridiculous and a bit of a barrier to playing the game in my opinion.
From the perspective of a new player, wanting to get into playing a Elachi Monbosh and S'golth at some point, just knowing that I am stuck either grinding 180+ million EC or spending hundreds of dollars is really disappointing. Honestly I would much rather have a monthly subscription with more open accessibility to things like exotic ships. Free-to-play differently isn't the best MMO model by far.
Sanity checks have always been a good tool. If you calculate a probability of 360% then maybe, just maybe, you've done something wrong. Glad you weren't teaching any of my math classes :P
no, the math is right but i realize the website did not upload the full answer. I originally used an editor called Math Type to input it so the full answer would be there. However the short answer is, its 3.60e-784% but that is the odds of getting a ship every time.
Just getting a ship ONCE would be (P)=(# of favorable Outcomes)/(# of total possible Outcomes). i do not know the total list of items you could possibly win from a lockbox but what it will show is your odds will not change significantly just because you open more packs. Though others like to state that over time you "should be guaranteed" a ship due to the improbability of not getting a ship but that is not how this works. What you find is that as you open more packs and add more favorable outcomes you have to add equally the same possible outcomes which in turn will keep the chance the same.
Example, if each pack has 11 chances to get JHAS out of 1000 items thats 11/1000 or 1.1% opening two packs it would be 22/2000 or 11/1000, the same odds. It is incorrect to add 11/1000+11/1000=22/1000 because from two packs you have 22 chances out of 2000 total items NOT 22 chances out of 1000 items. You can not just add the favorable outcomes from pack two but not add the total outcomes.