30 Items at 20% Chance to RE Each and Not a Single Hit? Yeah, RIIIIIIIIIGHT....

[publicenemies]

new in the cartel market crafting crit chance crits paks! only 1000 cartel coins!

 

The sad part is this isnt far off from bieng true

[ivanhedgehog]

You seem to not understand probability. Truly random sequences will have streaks in them.

 

If vegas had "random" number generators goughing up streaks like swtor they would be investigating it. Today on illum doing the kill beasts part of 1 of the missions. I got all fifteen of my drops in the first 5 or 6 groups, my wife got 1 drop off of each group. that has been the exact pattern for every day of the event. funny how that works that way.

 

Seriously, there needs to be streak breaker code added, like other more successfull games have.

[AshlaBoga]

The only trick I've noticed is that if you RE when you're on your speeder it seems lower your chances...

 

But that could just be coincidence.

[ivanhedgehog]

Yup. Unfortunately I don't have the same luck in lotteries.

 

In any case, I have seen players that I believe are reputable post about similar and longer streaks in the crafting section of the forum. The threads pop up, generate the same discussion about probability and the number of players that attempt to RE and how anything can happen in a small sample and then die away until the next frustrated person starts the same old thread again.

 

One reason that there are a lot of complaints about the RE process is that the unlikely is likely to happen and does so more often than one would expect.

 

RNG is RNG except that in this game it seems more RNGy that one would expect.

 

the rng is only as good as the programmers that put it into use. Seeing as how bug free this game is, there could very likely be problems.

[Nitblade]

That thing's been horrible since day 1.

And it's not been fixed either despite numerous threads about it in the first few months.

People just gave up on it.

[Socialist]

Probability of 30 items giving 0 schematic REs at 20% chance,

 

(excel)

=BINOM.DIST(0, 30, 0.2, FALSE)

= 0.00123 = 0.123%

 

That horribly sucks Never been that unlucky yet.

 

I am an Economist and I approve this post.

[mothear]

I am sure overall I will probably average out at the correct % but tthe thing that bugs me is that you never seem to get the correct pre-fix which makes it seem worse than it actually is.

[annabethchase]

Thanks for that detailed response

 

Totally see where your coming from and how you came to these conclusions

 

So refreshing as well cause so often we just see 2 word answers that are confrontational by design with out adding anything constructive or meaningful to the discussion at hand.

 

Really hope EA is watching closely to see your articulately presented counter

 

Take a statistics course. No explanation is needed or justified.

[mastirkal]

While individuality 1/5^30 is 1.0737418e-21 technically speaking it should be impossible for you to miss that many times on a single item in your lifetime you have to remember that successes in this game are accumulative. If you want the highest chance of getting a crit in that 1/5 ratio RE 5 of them at a very unpopulated time or a peak populated time. You'll have the most success then. Edited February 18, 2013 by mastirkal

[Kaerwolf]

Based on the probability system many of you defend the current "random-system" which means you either end up in streaks of endless misses or streaks of luck noone should continue RE after 5 misses and instead wait a while and try again until they succeed within the first couple of RE attempts and continue for multiple frequent successes, thus exploiting the system. Successes do come frequent once the pop comes and a costly streak of misses can easily go on for 20-30 failed RE attempts in a row. I think this is the problem the thread opener was adressing.

[ConradLionhart]

There's always a possibility that the RNG is really broken and if the devs find out about this, they will secretly deny it.

 

But the most likely scenario is that you were EXTREMELY unlucky, and for that I offer you my condolences.

[TheShootist]

early days I remember REing 100+ times on level 49 items before getting a schematic. One implant I did finally hit after just over 90 attempts, worst part is it gave the command version for presents. After that I went another 40 tries before giving up.

[Aries_cz]

As many have said, it is 20% chance on each attempt, it does not stack.

Although I would welcome a system that even failure to RE an item would increase a chance of next RE attempt. Although it would be a programmer's nightmare to keep these numbers stored for each item.

[Kalfear]

Take a statistics course. No explanation is needed or justified.

 

and I could respond with

 

Take a English language course to learn how to put your thoughts into a written response

 

Your response doesn't mean a thing and frankly makes me laugh (pretty sure my post secondary accounting beats your grade 9 statistics and neither have anything at all to do with game).

 

Seriously, I made a proposal

if you don't like it

EXPLAIN WHY and be constructive addition to thread

If you cant explain why

see above suggestion for English language course on how to put your thoughts into a written post

 

No need to troll the forums as you currently doing

Your adding nothing to the thread other then crappy attitude and unearned ego

[Kalfear]

As many have said, it is 20% chance on each attempt, it does not stack.

Although I would welcome a system that even failure to RE an item would increase a chance of next RE attempt. Although it would be a programmer's nightmare to keep these numbers stored for each item.

 

naaaa no need to store long term

just in that session

 

So you RE a item at 20% and fail

wait 20 minutes and RE 2nd item you restart at 20%

 

BUT

 

If you make 5 items (or get 5 Black Hole items) and RE them 1 after another in 1 sitting till you get success

Then you should gain chances at greater success

 

As I posted before

 

first try 20%

2nd try 40%

3rd try 60%

4th try 80%

5th try 100%

 

If you RE in one sitting a 20% chance item

 

And for high level items (would have to be stuff added after this as changing percentages on items that others RE'd at already completely unfair) they could start coming out with lesser % chances of like 2%, 1% so people understand going to 100 tries could be a issue and possible.

 

Im sorry but I just dont buy 57, 63, 93 tries at 20% as unlucky.

Unlucky doesnt even scratch the surface of naming the chances of doing that multiple times.

Do people realize the odds of missing a 20% chance that many times in a row?

And then multiple people doing so, not just 1,

meaning the odds get even lower and more outlandish?

Completely undermines the laws of average.

As someone said in this thread, if Vegas or any casino worked the way this is suppose to work, they would be under investigation for corruption and machine fixing on their slots

 

Its funny, some "poster" told me to take a statistics course to under stand this but funny thing is any statistics course would show you how insanely small and unlikely the odds of missing a 20% margin is, the more times you roll. Her attempted insult in fact completely blows up in her face and completely negates the attempted point/insult she was trying to make.

 

But yeah, it doesnt have to be complicated

 

store the numbers for a set period of time (5 minutes) and then it starts over so you only improve your chances by covering the odds number wise

[Malastare]

For those of you who don't know, computers are unable to generate truly random numbers; their choice-selections are based off a time-clock, multiplied, sometimes several times within the same algorithm, off of said time-clock. Is it based on the minute? It should be on the hundreths of seconds.

 

You don't seem to understand how pseudo-random number generators work. You're close.... but not really there.

 

PRNGs are based off of semi-complex equations where the previous result is multiplied, divided, raised to a certain power, or any combination of the three followed by some exclusion of the result (eg: Taking only the lowest half of the number). It is that equation that is used repeatedly to generate a number. However, each equation has a certain number of cycles that it can be used before it starts repeating the same series of numbers. So, you customarily provide the initial number, the seed, to ensure it starts someplace unpredictable. And yes, that was commonly supplied as the current system clock value. Which, by the way, is always measured in seconds at the very largest.

 

That is what time is used for. It was generally used once, just to initialize the sequence. Re-seeding every round was a sign of bad coding and a total misunderstanding of how PRNGs work.

 

Of course, that doesn't matter, as that's not how PRNGs are generated now. Much better equations exist, such as the popular Fast Mersenne Twister, which have repeat periods that are so large they effectively don't exist. Even better, most operating systems have their own PRNG mechanisms that use mechanisms so unpredictable that they may not even qualify as pseudo-random number generators. They often use encryption algorithms that repeatedly incorporate entropy from various sources of unpredictable data such as process IDs, the value on a nanosecond clock, cpu/memory statistics, and other odd randomness like the movement of the mouse or milliseconds between keyboard presses.

 

The end result is a series of numbers that are random enough to be used in cryptography. Every major OS has a RNG service like this. Windows has CryptGenRandom. I don't know if SWTOR uses that. Sometimes games will opt not to use it for ever call in fears that it will be too slow. However, most games will at least use it to repeatedly seed a faster local algorithm. And in that case, repeatedly seeding is a good thing.

 

TL;DR: You misunderstand how random numbers are generated. They are far better than you understand and are functionally totally random. Your attempted criticism falls flat because you don't actually understand how things work.

[wanttornow]

I see nothing wrong with there being times when it takes a large number of tries to RE.

 

I could play the lottery every week for 60 years and never win. Then I watch the news and see some guy that bought his first ticket and got the jackpot. Under laws of probability it could be said that shouldn't happen. But it does, and according to the news quite often.

 

Looking at the stats earlier in the thread there was an example of a 0.123% chance. but with, say, 500,000 users there are 615 users that would happen to - and if it hacks them off enough they will probably want to vent and post on the forum.

 

I was RE'ing last night and was quite regularly getting dings off the 2nd or 3rd try. For the laws of probability to work others would have to be taking much longer than that.

 

Unless anything can be shown that they are cooking the books (such as fruit machines do to ensure the correct payout percentages) then I can take this all as random probability. Forcing it so that it increases the hit rate per number of tries would actually mess up probability as the 20% average would be increased due to external measures - if you looked at their averages across the player base I would be surprised if it was much off that 20% figure with the current system. The solution proposed would actually skew that 20% upwards.

[Devlonir]

Simple fact:

 

If you have a 20% chance of something happening, and you perform that chance 10 times, there is still around a 10% chance of it not happening within that entire range.

 

For 30, a quick calculation meant it was a 0.12% chance of it not occuring even after 30 times. Yes, that is a very small chance, but still a chance.

 

This means that 1 in every 1000 sets of 30 items being REed for 20% chance may experience this. It sucks if you are that one, but 1 in 1000 is not really a slim chance when you take larger groups of events/people into account.

[Drudenfusz]

I know it's random, but still after one chain of bad luck and not getting any success with my RE I stopped crafting altogether months ago.

[AsheraII]

Just RE'd 2 different items, 15 each, both with 20% chance to RE. And?

 

NOTHING.

 

20%? Yeah, right. You're going to tell me that 30 items, each with a 20% chance to RE didn't ONCE hit within that 20% range?

 

What unseen element is it, exactly, that stops the random rolls in RE'ing from differing? For those of you who don't know, computers are unable to generate truly random numbers; their choice-selections are based off a time-clock, multiplied, sometimes several times within the same algorithm, off of said time-clock. Is it based on the minute? It should be on the hundreths of seconds.

 

This is far from the first time that I've wasted hundred of thousands of credits, either, in pursuit of the same thing. It wasn't level 47 stuff, either; the Might Mods were level 29, and the Blue Vehemence Crystals were level 31.

 

We need either a re-write of the "random"-number code in SWTOR, or a queue to guarantee that you get under-50 items to RE after so many tries.

 

/EndNerdRage

 

Statistically, this should happen about once for every 800 items you attempt to aquire an upgraded blueprint from. Statistically. Ofcourse, some people have more luck, others even less.

[MajikMyst]

I think the point that isn't being made is that one attempt has no bearing on the other.. Your chance doesn't increase the more your try.. If you have a 20% chance with one RE attempt, then you have a 20% chance with every attempt.. The 10 you RE'd before has no impact on your chances with any future attempt.. Your 20% remains the same..

 

Statistically speaking, it is possible to RE a 100 items and never get a proc.. Again, each attempt is it's own attempt..

 

Another way to describe this, you have a 20% chance of rolling an 80 or better out of 100.. Of course you also have the same odds of rolling a 20 or less.. If you are like me, I spend most my time in the 20 or less area.. Some people will hit it on the first try or couple of tries.. Other people will need a few rolls to hit 80.. It is all random chance.. The 20% is just your odds of winning with each RE attempt.. 1 in 5 chance to win.. 4 in 5 chance to lose.. Simple as that..

Edited February 18, 2013 by MajikMyst

[Natan]

Id like to see the chance % go up with every failed reverse, so it starts at 20% then adds 5% chance every time you RE something that doesn't succeed. Would at least give you the feeling that at some point the item will upgrade for you.

 

BUT, the downside would be this could be exploited, people could RE a few low level parts until they get a high % and then RE some black hole mods.. etc

[sanchito]

Probability of 30 items giving 0 schematic REs at 20% chance,

 

(excel)

=BINOM.DIST(0, 30, 0.2, FALSE)

= 0.00123 = 0.123%

 

That horribly sucks Never been that unlucky yet.

 

that's not even that low of a probability, there are 1000s of players, it's bound to happen to some

[Elfa]

The worse time I ever had was RE'ing a green level 32 Resolve stim to a blue. 103 times it took before I got the blue.

 

103, enough said.

Edited February 18, 2013 by Elfa

[SOULCASTER]

well keep in mind, while you have a 20% chance to successfully RE, you have an 80% chance to fail