Please upgrade your browser for the best possible experience.

Chrome Firefox Internet Explorer
×

Reverse Engineering is not 20%

STAR WARS: The Old Republic > English > Crew Skills
Reverse Engineering is not 20%

SuperGrunt's Avatar


SuperGrunt
03.01.2013 , 11:07 AM | #91
Quote: Originally Posted by DarthTHC View Post
If it's really a 20% chance, and you do it enough times, the actual results should show something very close to 20%.

The question is, what is "enough times" to prove the 20% out? It's not going to be measured in 100's or even 1,000's. Any statisticians care to answer that?
In order to have any chance to prove/disprove it would take an exhaustive effort, to either get the entire server to co-operate and provide their results for an entire day worth or RE results, or take a server where you have unlimited access to things to RE and are the only person on the server. Good luck finding that amount of co-operation in this community, b/c you are not going to get the second option.
My propsed changes to Loot Rules.
My Twitch channel, both a link and the address so you can choose how you visit it. twitch.tv/otawo

SuperGrunt's Avatar


SuperGrunt
03.01.2013 , 11:11 AM | #92
Quote: Originally Posted by Khevar View Post
*Sigh*

I may disagree with the OP about sample size, but there is NO DOUBT that he has a thorough and advanced understanding of statistics. You are completely misunderstanding the point he is trying to make. I would suggest looking up "Confidence Interval" and "Standard Deviation" and re-reading his post.

I also recommend re-reading post #17 and #18 and review the graphs both posters supplied supporting their points.

This thread is, at it's heart, a mathematical analysis of RE.
How can you possibly provide a mathematical analysis of RE without having access to the code? TBQH without knowing what the RE code is, what the RNG code is, and how it is implemented you can't possibly give any real accounting of the results that you see in the game.
My propsed changes to Loot Rules.
My Twitch channel, both a link and the address so you can choose how you visit it. twitch.tv/otawo

Zoekhan's Avatar


Zoekhan
03.01.2013 , 11:17 AM | #93
Quote: Originally Posted by Darth_Sweets View Post
after going through too many materials i decided to check the reverse engineering rate of going greens to blues which is stated to be at 20% according to the tool tip in game. I have collected the number of reverse engineering attempt and success that i have had going from a green to a blue, I have not counted the times where I RE'ed and there was no chance of a plan to me gained.

<Wall of numbers>

This gives a 13.4 percent mean. Using a standard confidence calculation with 99.7 percent boundary if 20 percent is the true mean as defined in the tool tip the average for the sample above should be between 14.3 to 25.7 percent. With the mean sample is out of the 99.7 boundary that mean that it is almost impossible that the 20 percent is the true rate of getting a new plan.
You're doing it wrong.

SuperGrunt's Avatar


SuperGrunt
03.01.2013 , 11:23 AM | #94
Quote: Originally Posted by Blackardin View Post
The current crafting system is equivalent to wearing level 50 tank gear that does not increase any stats, improve one's chances to mitigate damage, dodge, absorb. A crafting system without a learning curve, without the ability to improve one's skill, chances, percentages is a failed system. Its really that simple.
The crating system is fine, it does indeed have a learning curve to it. The more items you RE the more recipes you have and less you have to learn, which equates to a CURVE if you graph it. But seriously, why are you even complaining? most of the time when I RE a loot drooped MOD, Armoring, or Enhancement lately (even lowbie ones which used to have 0% chance to gain a schem) I have gotten a schematic. Maybe you are REing the wrong items? Instead of trying to bilk all the people trying to level characters out of credits, why don't you try REing those sexy end game mods that loot drop in OPS?
My propsed changes to Loot Rules.
My Twitch channel, both a link and the address so you can choose how you visit it. twitch.tv/otawo

psandak's Avatar


psandak
03.01.2013 , 11:46 AM | #95
Quote: Originally Posted by SuperGrunt View Post
In order to have any chance to prove/disprove it would take an exhaustive effort, to either get the entire server to co-operate and provide their results for an entire day worth or RE results, or take a server where you have unlimited access to things to RE and are the only person on the server. Good luck finding that amount of co-operation in this community, b/c you are not going to get the second option.
Not just this community.

Fact is that everything we as players do to try to prove or disprove a feature is working as intended is irrelevant; anything anyone professes is anecdotal at best. I've said before and I'll say it again, we as human beings cannot help but see clusters as purposeful and even distribution as random when in fact the opposite is overwhelmingly more often true (clusters are random and even distribution is purposeful). SuperGrunt is on the right track that the only way to get reliable data from players' observations would be to get a massive number of players doing a massive number of REs. And he is right, good luck getting that to happen.

Khevar's Avatar


Khevar
03.01.2013 , 12:28 PM | #96
Quote: Originally Posted by SuperGrunt View Post
How can you possibly provide a mathematical analysis of RE without having access to the code? TBQH without knowing what the RE code is, what the RNG code is, and how it is implemented you can't possibly give any real accounting of the results that you see in the game.
Actually, that was sort of the point I was trying to make to the OP. His sample size was too small.

However, the nature of his analysis was, in fact, a mathematical one. It was refreshing to see a post about RE that wasn't grounded in an emotional perspective.

Headpunch's post that "the OP doesn't understand statistics" threw me, as anyone who has studied the type of math being described in this thread can tell that Darth_Sweets isn't likely to be lying when he presents his field of expertise (post #11).

SuperGrunt's Avatar


SuperGrunt
03.01.2013 , 01:49 PM | #97
Quote: Originally Posted by psandak View Post
Not just this community.

Fact is that everything we as players do to try to prove or disprove a feature is working as intended is irrelevant; anything anyone professes is anecdotal at best. I've said before and I'll say it again, we as human beings cannot help but see clusters as purposeful and even distribution as random when in fact the opposite is overwhelmingly more often true (clusters are random and even distribution is purposeful). SuperGrunt is on the right track that the only way to get reliable data from players' observations would be to get a massive number of players doing a massive number of REs. And he is right, good luck getting that to happen.
Like I said though it would have to be from one server, unless you did multiple servers and compiled the results based on server. Time is the only variable that would change on a computer that would allow you to change the number generated with a random function, so you would want the server time as a constant, as 2 different computers using the same time based random function would generate 2 different random numbers.
My propsed changes to Loot Rules.
My Twitch channel, both a link and the address so you can choose how you visit it. twitch.tv/otawo

Arvig's Avatar


Arvig
03.01.2013 , 09:09 PM | #98
Quote: Originally Posted by Darth_Sweets View Post
after going through too many materials i decided to check the reverse engineering rate of going greens to blues which is stated to be at 20% according to the tool tip in game. I have collected the number of reverse engineering attempt and success that i have had going from a green to a blue, I have not counted the times where I RE'ed and there was no chance of a plan to me gained.

<Snips RE "statistics">

This gives a 13.4 percent mean. Using a standard confidence calculation with 99.7 percent boundary if 20 percent is the true mean as defined in the tool tip the average for the sample above should be between 14.3 to 25.7 percent. With the mean sample is out of the 99.7 boundary that mean that it is almost impossible that the 20 percent is the true rate of getting a new plan.
As others pointed out, all you did is prove that random is random.

I've literally looted an item that stated a 10 percent chance to RE, went ahead and did it on a lark, and got the next quality tier schematic.

And yes, I've hit it where it takes at least 10 items to get a schematic from RE.

Lodril's Avatar


Lodril
03.03.2013 , 06:38 PM | #99
Quote: Originally Posted by psandak View Post
Fact is that everything we as players do to try to prove or disprove a feature is working as intended is irrelevant; anything anyone professes is anecdotal at best.
Hehehe

This is exactly the sort of 'knowledge is impossible' reasoning I was talking about. There is no margin for error, there is only absolute knowledge, and as absolute knowledge cannot be obtained, we cannot learn anything from observation.

'Anecdotal' information is still information. People love that word because it lets them sound science-y, yet still be dismissive without having to engage in any actual analysis. The original poster took a number of 'anecdotal' data points and compiled them into a collection of data. Your argument that no one could ever take a number of instances and draw conclusions from that is not reasoning, but instead just a convoluted call to faith. You can argue with the original poster's margin for error based on his sample size, but instead you claim nothing is provable because observation cannot result in understanding, merely 'anecdote'.

Close your eyes, foolish mortals, and cover your ears, for nothing that passes can you ever comprehend.

JouerTue's Avatar


JouerTue
03.04.2013 , 07:19 AM | #100
more RE you do more you get close to 20%..the numbers you are working on aren't sufficiently high