20% chance for research scematic? ...LOL?!

[HexHammer]

More realisticly it's 10% in average, it just kills me to make least 15 for success.

[Ironcleaver]

15 would be insanly awesome, sadly for most people it's 50~200 per schematic. This has imporved since 1.2 but people are still getting large unlucky streaks.

[IsfetTheGreat]

Seems like no matter how big my stack is, I tend to sit around 15-20% success rate for new research. This rate does not change regardless of whether I RE 10 items at a time, or one at a time.

 

The only way my mind can justify that is that the programmers are either bad at math, or forgot to take into account the fact that if your chance for new research is 20%, and you RE a stack of 10, there's no way that you shouldn't be getting new schematics.

[Dredredre]

Seems like no matter how big my stack is, I tend to sit around 15-20% success rate for new research. This rate does not change regardless of whether I RE 10 items at a time, or one at a time.

 

The only way my mind can justify that is that the programmers are either bad at math, or forgot to take into account the fact that if your chance for new research is 20%, and you RE a stack of 10, there's no way that you shouldn't be getting new schematics.

 

 

Food for thought.....http://forum.dohforum.com/viewtopic.php?f=6&t=626

[Cerion]

Seems like no matter how big my stack is, I tend to sit around 15-20% success rate for new research. This rate does not change regardless of whether I RE 10 items at a time, or one at a time.

 

The only way my mind can justify that is that the programmers are either bad at math, or forgot to take into account the fact that if your chance for new research is 20%, and you RE a stack of 10, there's no way that you shouldn't be getting new schematics.

 

I think you should be careful about who you're accusing of being bad at math, lol.

[Zajen]

Putting 20 percent on the icon is just plain audacious and a flat out lie.

 

It almost always takes over 10 RE's and has been as many as 26 in my experience.

 

Things like this just make people angry.

[Dabrixmgp]

20% is a tooltip error it mean to say .20%

[Cerion]

Putting 20 percent on the icon is just plain audacious and a flat out lie.

 

It almost always takes over 10 RE's and has been as many as 26 in my experience.

 

Things like this just make people angry.

 

That is why experiential evidence is poor evidence. My experience has been quite the opposite. I landed a few blues today on 1st or second RE. None of my REs on the way to 400 went over 10.

 

What I find ironic is that the devs put the percentage chance for RE on each item, thinking it would inform the player base, and pacify them...when in fact the devs over-estimated the playerbase's ability to grasp statistics.

 

However, the devs aren't entirely guiltless here...they designed a system best suited for a small population of users, thus minimizing the outlying players who'd have streaks of bad luck. With a large population of crafters, the number of players with long streaks of bad luck is more frequent, and thus more problematic.

[siegeshot]

Yeah it routinely takes 15-20 of an item for that 20% to proc. Other times I'll get 3 procs in a row of something I didn't even really care about that much.

 

I wish it remembered that I had been trying to reverse engineer an item and maybe just put in a cold hard "reverse engineer 5 of these to get the next one" or something. I'd rather that then get 5 greens > blues on first try and then a blue > purp at about 20 tries.

 

Which is what I'm at right now with the lvl 47 yellow crit crystal. I got the orange in 3 tries, the yellow I'm at 15'ish and still haven't gotten it. I would have settled for a static number, even if the static number was high. It is painful to blow through mats like this wondering when the 20% will proc.

Edited May 13, 2012 by siegeshot

[pokota]

This is strange. Perhaps I'm doing something differently, because my average since 2.1 has been very close to 20%, and I've RE'd hundreds of times. That it's a dramatic improvement from before 2.1 is without doubt.

 

So far, I'm pretty content with it. For every steak over 10, I've had a win under 5. I dunno, but it's working for me.

[Cerion]

This is strange. Perhaps I'm doing something differently, because my average since 2.1 has been very close to 20%, and I've RE'd hundreds of times. That it's a dramatic improvement from before 2.1 is without doubt.

 

So far, I'm pretty content with it. For every steak over 10, I've had a win under 5. I dunno, but it's working for me.

 

You aren't doing anything differently...except that your luck is matching your expectations, so you don't start a thread about it The people who are having bad streaks post about it most...which is understandable.

[Sithhelmet]

237/1263 schematics to tries since 1.2. 18-19%. That's high enough of a count for me to make it seem right and still low enough to account for it not being 20% even.

 

Related music:

[SteveGarbage]

It seems like I post this or a variation everyday. I really need to just have it somewhere to C/P.

 

The 20 percent chance does not mean that you should be able to RE once for every five times you try. It doesn't even mean that you'll get it in 10 tries or 20 or 50. The reality is every time you RE you have a 20 percent chance to succeed vs. an 80 percent chance to fail. Since your chance for success doesn't cumulatively increase with each failure, you have a 80 percent failure rate every time.

 

While as a whole, statistically the chance of failing 25 times in a row is small, but not nearly impossible since each RE try is independent of the others.

 

Consider it this way, flipping a coin, the chances of flipping 20 heads in a row is 1 in 1.05 million. But if you did happen to flip 19 heads, the chance of flipping a tail on that 20th flip is 50 percent, the same as flipping that 20th head.

 

RE works the way it's designed. Sometimes you have good luck, sometimes bad. I was able to RE three purple-grade Lvl 48 shield schematics today in six tries. However I've RE'd like 15 green-grade generators for Rusk and still haven't gotten my blue schem. Happens.

[okwathne]

Like the one above me, I do have the knowledge to do the math......

20% chance is not the same as 20% of RE will give recepie.....

20% each time, 80% chance each time that it will not..........

Every time you try to get a recepie, it is 4 times more likely you will fail......

They should do something to de strikes not getting it, but it is still 20% no mather how much ppl with no math skill complain......

[RickRedOne]

Seems like no matter how big my stack is, I tend to sit around 15-20% success rate for new research. This rate does not change regardless of whether I RE 10 items at a time, or one at a time.

 

The only way my mind can justify that is that the programmers are either bad at math, or forgot to take into account the fact that if your chance for new research is 20%, and you RE a stack of 10, there's no way that you shouldn't be getting new schematics.

 

The Vegas casino's would like to fly you first class to come and play at their tables.

 

What happened the last 19 times you RE'd has no effect on the 20th RE.

Edited May 13, 2012 by RickRedOne

[VeovisX]

I haven't been able to RE anything successfully for days!

But I'll post this again since some people don't understand real math:

 

Given the 20% chance to RE something: the probability of actually succeeding at least 1 RE in a streak is 1-(4/5)^x. (if you are doing a 10% chance use 1-(9/10)^x, but we'll be using the more common 20%)

 

Given 5 tries (x=5) the the chance of succeeding one RE in a streak is 1-(1024/3125)=0.67232 or 67.232%

lets continue with some math now for longer streaks:

x=6;1 - ((4 / 5)^6) = 0.737856

x=7;1 - ((4 / 5)^7) = 0.7902848

x=8;1 - ((4 / 5)^8) = 0.83222784

x=9;1 - ((4 / 5)^9) = 0.865782272

x=10;1 - ((4 / 5)^10) = 0.892625818

x=12;1 - ((4 / 5)^12) = 0.931280523

x=15;1 - ((4 / 5)^15) = 0.964815628

x=20;1 - ((4 / 5)^20) = 0.988470785

x=30;1 - ((4 / 5)^30) = 0.99876206

 

But consider everyone that plays SWTOR as a population(number of created characters is more then likely high enough to warrant this), these chances are the same for everyone so still if everyone is crafting 6.89% of players will in theory be getting a streak of 12 before they they see results, and 6.89% out of all the players is still a large number. Also statistics states something is not an uncommon occurrence until it falls into the -/+5 percentile. So missing 13 RE's shouldn't strike you are odd, but once you are hitting 14+ there is probably something wrong.

[UltimaST]

Well I'm bad at math as hell, so I wouldn't know about these things but I too em unhappy with the currant system. Well to tell the truth I don't like anything that is a lottery. Its not too bad that you get a random upgrade on the schematic but to only have it once every 15 tires is very bad. That's like lottery times 2. And the mats I have burned and credits I'm almost constantly with out money. Well in the end its my choice to waste mats money and time on this but I would very much like to see something more reliable. I agree with a system where u research several to get one schematic or something similar. I most definitely don't have time to farm mats all day every day to get a full set of gear that i want I need to go to work take care of my dog and do chores and errands I sit with a game only 2 to 3 hours a day.

[psandak]

I haven't been able to RE anything successfully for days!

But I'll post this again since some people don't understand real math:

 

Given the 20% chance to RE something: the probability of actually succeeding at least 1 RE in a streak is 1-(4/5)^x. (if you are doing a 10% chance use 1-(9/10)^x, but we'll be using the more common 20%)

 

Given 5 tries (x=5) the the chance of succeeding one RE in a streak is 1-(1024/3125)=0.67232 or 67.232%

lets continue with some math now for longer streaks:

x=6;1 - ((4 / 5)^6) = 0.737856

x=7;1 - ((4 / 5)^7) = 0.7902848

x=8;1 - ((4 / 5)^8) = 0.83222784

x=9;1 - ((4 / 5)^9) = 0.865782272

x=10;1 - ((4 / 5)^10) = 0.892625818

x=12;1 - ((4 / 5)^12) = 0.931280523

x=15;1 - ((4 / 5)^15) = 0.964815628

x=20;1 - ((4 / 5)^20) = 0.988470785

x=30;1 - ((4 / 5)^30) = 0.99876206

 

But consider everyone that plays SWTOR as a population(number of created characters is more then likely high enough to warrant this), these chances are the same for everyone so still if everyone is crafting 6.89% of players will in theory be getting a streak of 12 before they they see results, and 6.89% out of all the players is still a large number. Also statistics states something is not an uncommon occurrence until it falls into the -/+5 percentile. So missing 13 RE's shouldn't strike you are odd, but once you are hitting 14+ there is probably something wrong.

 

Something else to take into account though is the possibility that you are getting a new schematic you are just getting one you already know and therefore the result is null. It may not announce it in chat anymore, but was this functionality actually removed/fixed?

[MarkusBiggus]

I'm not convinced it is a true 20% chance on every RE.

My experience has been that a new schematic has never resulted from crafting an item while its difficulty colour is red or orange. These are the ones that do 10 in a row unsuccessfully but by the time it works they've gone to green difficulty. It also seems to take more crafts of an items of higher crafting skill to change from orange to yellow to green.

Green level difficulty seems to be where the 20% is working and for grey (the easiest) I frequently get a new schematic on first or second try. I've tested this by going back to craft lower levels item that I had never RE'd at the time they were my highest level items.

The pattern I have observed having RE'd a lot does not appear to be random 20% chance on each occasion but heavily influenced by the difficulty level to craft the items being RE'd.

The other anecdotal evidence I have is that crystals I craft are harder to RE than other items but crystals that dropped from missions RE on 1st (and only possible) attempt on several occasions. (small sample size, maybe i was lucky)

I'd be interested if anyone else has noticed a correlation between ease of crafting and chance of RE success that I've observed.

[MaverickXIV]

A random variable with p (probability) = .2 (20%) has a mean of:

 

1/p = 1/.2 = 5

 

The average number of attempts it should take is 5. The average number of attempts. That means sometimes it will take you a few attempts and sometimes it will take you many more.

 

The probability that it takes you 10 attempts is:

 

(1 - p)^10 = .8^10 = .107374182 = 10.7%

 

The probability that it takes you 20 attempts is 1.15%

Edited June 29, 2012 by MaverickXIV

[Astarica]

Isn't there some schematics that really only have 10% chance to RE? Or are these bugged and actually meant to be the 20%?

 

The ones tha say 20% I've no reason to believe they're not 20%, but the ones that say 10% sure doesn't feel like it's 20%.

[Teykos]

Isn't there some schematics that really only have 10% chance to RE? Or are these bugged and actually meant to be the 20%?

 

The ones tha say 20% I've no reason to believe they're not 20%, but the ones that say 10% sure doesn't feel like it's 20%.

 

The higher level second tier schematics that aren't linear (i.e. not mods, armorings, barrels, etc) are only 10%.

Edited June 30, 2012 by Teykos
A little more clarity

[Grayseven]

I get 20%. Long streaks of not getting a hit are made up for by streaks of getting 3 schematics in 3 RE's. You can't look at a small sample size and say it's broken.

 

Every time you RE, you have 1 chance in 5 of getting a schematic. Each one is independent. If you have any d20's laying around (c'mon, you know you do) or if you have a knack with Excel and can make a RNG, roll and track 1000 die rolls and see how many come out a 1 through 4. It should be pretty close to 20%, give or take a few percentile.

 

I've not seen any evidence to the contrary of the 20% chance not really being 20%.

[Duart]

It could be. Coded so that for at 20% chance after 10 failures the item could be REed with a 100% chance of it working

 

adjust the point at which a given item gets assured RE result. As needed for balance... If this was done the complaints on this topic would go away.

 

Given that the folks who are creating this game are smart programmer s they must have considered this option and rejected it for good reasons.... I would love to know what those reasons are.

[Grayseven]

It could be. Coded so that for at 20% chance after 10 failures the item could be REed with a 100% chance of it working

 

adjust the point at which a given item gets assured RE result. As needed for balance... If this was done the complaints on this topic would go away.

 

Given that the folks who are creating this game are smart programmer s they must have considered this option and rejected it for good reasons.... I would love to know what those reasons are.

 

Easily grinding out every possible schematic is the most plausible reason against "auto" RE success after a set number of iterations. The system works fine as it is set up as long as you don't feel the need to get every single possible schematic. If you do, be prepared to work for it.