Reverse Engineering Broken - or am I crazy???

[Saaz]

I know that some people have bad luck.... but

 

I was making a Might Hilt 9 for an alt.

 

The tooltip states 20% chance to discover.

 

I have made approximately 22 hilts trying to get the discovery.......

 

Using my basic math I think the probability of that happening is less than 1 percent.

 

Am I just unlucky or is this broken?

[Kruxxor]

Just bad luck.

 

You have a 20% chance of it re-ing to the next one.

 

Because you make 5, doesn't mean 1 will definitely re into the other.

 

You have 20% chance per one you re....

 

So 80% of the time you re you will not successfully get the new schematic.

 

Take dice for example, you have a 1 in 6 chance of rolling a 6, you roll it 6 times and never get a 6. Or you roll is 6 times and you always get a 6. The chance of getting a 6 is still 1 in 6.

[Saaz]

But my specific scenario is (0.8)^22 which equals 0.00738 probability of occuring....

 

or 0.738%

 

Just seems crazy.

 

I'll make another 5 and see how I go

[Saaz]

another 5

 

makes it 0.241%

 

---------------------

 

 

sorry finally got it around 33ish odd hilts

Edited April 15, 2012 by Saaz

[Valkyriesuki]

your mistake is thinking past failures have any effect on future attempts.

 

and as i reread your post i realize you are calculating odds on having 33 bad rolls.. these are two different statistics. while you may have a <1% chance of losing 33 times in a row, you still have a %20 chance of winning the next time.

Edited April 15, 2012 by Valkyriesuki

[Thebigkabooa]

Your math is off. You don't dip under 1% probability until you make 101 and fail.

[Arzoo]

Your math is off. You don't dip under 1% probability until you make 101 and fail.

 

 

His math is fine. He calculated the odds of not getting the recipe after however many tries, which is (1 - odds_of_failure) ^ number_of_tries.

[Darknscub]

I know that some people have bad luck.... but

 

I was making a Might Hilt 9 for an alt.

 

The tooltip states 20% chance to discover.

 

I have made approximately 22 hilts trying to get the discovery.......

 

Using my basic math I think the probability of that happening is less than 1 percent.

 

Am I just unlucky or is this broken?

 

Why do lame ppl keep complaining about the RE rate it use to be 4%. So then a bunch of lame noobs complain, so they up it to 10 and 20% and you still complain. You RE'd 22 do you want a medal 22 is nothing, if it was made as easy as the whining minority want crafting would be lame.

 

But then if it was made even easier so someone with half a brain cell could do it you would complain that it cost too much to send your companions out and the drop rates on rares were to low.

 

So shut up stop ************ and keep trying you will get the craft if you want it bad enough

[kaboro]

Why do lame ppl keep complaining about the RE rate it use to be 4%. So then a bunch of lame noobs complain, so they up it to 10 and 20% and you still complain. You RE'd 22 do you want a medal 22 is nothing, if it was made as easy as the whining minority want crafting would be lame.

 

But then if it was made even easier so someone with half a brain cell could do it you would complain that it cost too much to send your companions out and the drop rates on rares were to low.

 

So shut up stop ************ and keep trying you will get the craft if you want it bad enough

 

wow pvp forums are that way ---->

I guess you can take a player out of the barrens but cant take the barrens out of the player.

Thats what happens when bioware makes harvesting and crafting attractive for the kind of people that were usually staying out of it.

[kaboro]

I know that some people have bad luck.... but

 

I was making a Might Hilt 9 for an alt.

 

The tooltip states 20% chance to discover.

 

I have made approximately 22 hilts trying to get the discovery.......

 

Using my basic math I think the probability of that happening is less than 1 percent.

 

Am I just unlucky or is this broken?

 

I had the same thing happening repeatedly always on one category of items, while the other categories work too often.

On some categories the 20% is actually 50% while on others the 20% is actually 3-5% and its pretty consistent, always works like that on the same categories.

I assume on a whole the rate is correct, but not for the individual item categories.

I will give it more time to see if the trend persists and also try it on my alts.

If it does persist i will post a bug ticket or whatever its called in this game.

[venomlash]

Noticed this also. The 20% posted has no basis in reality.

[rjkishida]

It's broken. I wrote a similar post.

[Terrorble]

Maybe we should ask, can anyone else get the particular schematic in question? And, did it take many attempts to suggest its rate is not 20%?

 

RE has been working for me so far, but I've not been collecting data to say if the rates are as stated.

[MutatedSpleen]

I'm having this problem as well trying to craft a Level 37 kinetic shield (artifice).

 

I've RE'd around 20, with no luck. I definitely understand the statistics (i.e., it's a 20% each time, and past failures do not change the chance of future successes [side note: that's called an illusory correlation, in case you were curious]), but a 20% chance relates to 1 out of every 5. Which, of course, being a chance and not a guarantee, means it's not going to *actually* be 1/5, but rather, as a whole, 1/5. The problem is there seems to be a significant amount of reports of not getting a new schematic after 20-30 tries, which really makes the 20% look ridiculous. I mean, if you /dice 20 times, you're damn near guaranteed to get either 1-20 or 80-100 depending on how you look at it.

 

Anyway, I just want someone from Bioware to look and make sure their math is right. Because it doesn't seem right.

[GnatB]

Well, to point out... if there's around a 1% chance of going on a streak that bad...

 

that means, on average, around 1 out of every 100 people are going to be going on a streak that bad on the getgo.

 

How many people post on these boards? Surely SOME are going to be going on a streak that bad right at the start.

 

 

And, of course, that's only the *first* time somebody tries to RE an item. if there's 10 of us trying to RE 10 items, on average 1 of those items will go on a streak that bad...

 

so really 1% chance of a streak "that bad" means it's actually fairly likely, considering the number of times being tried.

[Kaskali]

I cannot believe someone thought it would be a good idea to display tooltip percentages for reverse engineering.

[GnatB]

I think it's a good idea.

 

It lets us know that there are serious mental problems with the design team, as they apparently gave linear RE items twice the chance (20%) as tiered items (10%), even though when you successfully RE a tiered item, you only have a 1 in 4 or 1 in 5 chance of getting the *good* one, whereas you always get the linear item you're going for.

 

Common sense would have seemed to indicate that the one with the sucker prizes should have had a higher chance of success.

[MutatedSpleen]

After doing a little bit of fun internet research, I found the following link: http://www.stat.tamu.edu/~west/applets/binomialdemo.html

 

This is a binomial probability calculator. "n" = number of attempts at REing, "p" = probability (20% of greens, 10% for blues, which equates to .2 and .1 on a 0-1 scale).

 

You can select "Prob X is exactly:" and put in 0, and it will tell you the odds that you won't get a new schematic after x tries.

 

Or you can select "Prob X is at least" and put in 1 (or any other number) and it will tell you what the odds of having gotten at least one of said item is.

 

The odds of someone REing 20 things and not getting a single schematic are, according to this calculator: 0.0115, or 1.15%. Tiny, but considering how many people RE, not impossible.

 

The odds of someone REing 20 things and only getting one are: 5.76%.

 

And for the "1 out of 5" crowd, the odds of getting 1 when you RE 5 is: 40.96%

 

The moral of the story is...if Bioware's equations and what not are right, it sucks, but it's still possible - and not even THAT unlikely, to RE 20+ and not get a single proc. My whole original question, however, was if they are even right to begin with.

[TheEdgeOfMadness]

Well, to point out... if there's around a 1% chance of going on a streak that bad...

 

that means, on average, around 1 out of every 100 people are going to be going on a streak that bad on the getgo.

 

How many people post on these boards? Surely SOME are going to be going on a streak that bad right at the start.

 

 

And, of course, that's only the *first* time somebody tries to RE an item. if there's 10 of us trying to RE 10 items, on average 1 of those items will go on a streak that bad...

 

so really 1% chance of a streak "that bad" means it's actually fairly likely, considering the number of times being tried.

 

Wow, great explanation. Really puts it in perspective! Btw the devs have said a couple of times that they are planning on implementing some sort of 'safety net' to deal with sucky strings of bad luck like this.

[Evgen]

Well, to point out... if there's around a 1% chance of going on a streak that bad...

 

that means, on average, around 1 out of every 100 people are going to be going on a streak that bad on the getgo.

 

How many people post on these boards? Surely SOME are going to be going on a streak that bad right at the start.

 

 

And, of course, that's only the *first* time somebody tries to RE an item. if there's 10 of us trying to RE 10 items, on average 1 of those items will go on a streak that bad...

 

so really 1% chance of a streak "that bad" means it's actually fairly likely, considering the number of times being tried.

How about this.

I was playing with a friend of mine last night, and we both had the same problem.

Him on Armormech, and me on Cybertech.

And while we were playing about 70% of the items dropped were for Jedi knights, some light armour, and 3 near identical trooper helmets. I think the problem here may be a not so random number generator.

It would explain why after finally getting the blue after 30 attempts, the purple comes after 3 or 4. For my mate last night it was only one attempt to get the purple.

[jolleebindu]

Well, to point out... if there's around a 1% chance of going on a streak that bad...

 

that means, on average, around 1 out of every 100 people are going to be going on a streak that bad on the getgo.

 

How many people post on these boards? Surely SOME are going to be going on a streak that bad right at the start.

 

 

And, of course, that's only the *first* time somebody tries to RE an item. if there's 10 of us trying to RE 10 items, on average 1 of those items will go on a streak that bad...

 

so really 1% chance of a streak "that bad" means it's actually fairly likely, considering the number of times being tried.

 

would it not make sense for them to utilize a pseudo random - so if you do go thru 15 of them on a 1/5 chance you are automatically granted the crit on the 16th - if it only applies to that 1/50ish case would it really be that big of a deal? (or if anything after the first 5 increase the odds to 40%, then after next five increase them to 60%)

[IAlkon]

Im up to 47 REs and 8+ hours trying to get the Artifact Might Hilt 22.

 

Working as intended I guess...

[Endymionsama]

Wow, great explanation. Really puts it in perspective! Btw the devs have said a couple of times that they are planning on implementing some sort of 'safety net' to deal with sucky strings of bad luck like this.

 

Aion had something similar. Their crafting system had a supposed 1/5 proc rate. However most players did not know there was a streak breaker that was easily exploitable. To beat the system in that game you could craft a ton of mats before hand. Once you got above a certain number you would attempt your actual craft. Because the game thought you did not proc x amount of times in a row the next 2 rolls were near guaranteed procs. For obvious reason it would not work for this game because you do not have to process mats to the same extent as Aion, but this gives you an idea of what a streak breaker could do.

[GnatB]

Oh, I'm perfectly able to ignore other players using them, just like I can ignore the siths running around with blue or green. (Jedi running around with red doesn't really bother me all that much)

 

Just from reading the wikia page (which is where I got the info from too) and looking at the pictures there I got the feeling that darksaber was some sort of *very* primitive precursor, possibly just some regular sword encased in plasma or something, (tech equivelent of a flaming sword?) and not a "real" lightsaber that just happened to be a different color. (Basically, still want to know if it actually "extended" in the show)

[dinchydoo]

I think it's down to the compion that makes the hilt, armor etc. and would like to know how can do what better ! but don't ask me !