Jump to content

[Guide] Reverse Engineering


Slaign

Recommended Posts

Great post!

 

Just wanted to educate you people in some probability theory:

 

If the chance of getting an improved schematic is 10%, Reverse engineering 1 items gives you a 10% chance of getting it of course, but Reverse engineering 2 items does not give you a 20% chance of getting it here is why.

 

You have a 90% chance of not getting the schematic per attempt (the complement of the chance of getting the schematic), so in two attempts to not get the schematic you have to not get it on the first one (90% chance) AND not get it on the second chance (90% chance). A rule of thump in probability theory is that AND is multiplication, OR is addition. Following that rule you have a 0.9*0.9 = 0.81 chance of not getting it in two attempts and the complement of that is of course getting it which is 1-0.81 = 0.19 = 19%.

 

A generalized formula is p = 1 - (1-p_0)^n where:

p - Chance to get the improved schematic with n attempts.

p_0 - Chance to get the item on a single attempt.

n - Number of attempts.

 

An example of the 10% progression:

 

1 Attempt 10 %

2 Attempts 19 %

3 Attempts 27.1 %

4 Attempts 34.39 %

5 Attempts 40.951 %

6 Attempts 46.856 %

7 Attempts 52.17 %

8 Attempts 56.953 %

9 Attempts 61.258 %

10 Attempts 65.132 %

 

As you can see you only have about a 65% chance of getting the schematic after 10 Reverse Engineering attempts even tough the individual chance is 10% i.e. 1 in 10, if you're really unlucky you can do it a thousand times and still don't get it but some people may get it on their first try so it evens out to a 10% chance (to with a 99% certainty get the schematic you would in fact have to reverse engineer 44 items and for a 100% chance you would have to reverse engineer an infinite amount of items).

 

I'd be willing to bet that P_0 is calculated from several variables, including crafting skill, rating of the item to be crafted, and potentially the number of other items in the same family that you have previously RE'd and/or crafted. This would hold in line with a lot of what people have reported (for example, those that crafted an item 50 times and didn't get it are possibly running up against an item whose rating outweighs their skill). Don't know how that fits with the endgame crafting RE issues, though.

 

I do know that item Difficulty is strangely calculated, I'm seeing weird patterns when I sort item tress by difficulty that don't align with rating, I used to think was proportional to rating, but now I think maybe has something to do with how many of the item you've crafted.

 

Edit: Assuming P_0 is a function of crafting skill(x) and item rating (y) and potential unknown variable (z), then we can calculate the probability based on brute force. So assuming P_0 = f[x,y,z], then essentially P=1-(f[x,y,z])^n

 

If enough people pick one item and report these following variables:

 

How many times did you have to attempt before you got schematic?(n)

What was your current crafting level?(x)

What was the crafted item rating?(y)

 

Then we could potentially calculate the distribution (P) from the number of answers, and use that to back calculate the dependent function f[x,y,z], because then the equation would only have a single unknown variable.

 

My math is rusty, this right?

Edited by Arch-Stanton
Wrong variable
Link to comment
Share on other sites

  • Replies 1.1k
  • Created
  • Last Reply

Top Posters In This Topic

oh wait another thing real quick....so i have the blue critical nano-optic enhancer....are you also saying that by upgrading the same green, or RE'ing, i might get not just the critical version but the overkill or redoubt version as well? so basically each of the greens that we can train have the possibility to make 3 different types of blue, which in turn have the chance to make 5 types of purples?

 

That's correct. It's the principle point of the guide, actually. :)

Link to comment
Share on other sites

If enough people pick one item and report these following variables:

 

...

 

My math is rusty, this right?

 

the math is right, but that enough ppl in there is... a large number. you'd also need them to be evenly spread over the skill level they have, and most ppl that would participate I guess got 400 already. you're also talking about only 1 skill which limits that even further down...

Link to comment
Share on other sites

I'd be willing to bet that P_0 is calculated from several variables, including crafting skill, rating of the item to be crafted, and potentially the number of other items in the same family that you have previously RE'd and/or crafted. This would hold in line with a lot of what people have reported (for example, those that crafted an item 50 times and didn't get it are possibly running up against an item whose rating outweighs their skill). Don't know how that fits with the endgame crafting RE issues, though.

 

I do know that item Difficulty is strangely calculated, I'm seeing weird patterns when I sort item tress by difficulty that don't align with rating, I used to think was proportional to rating, but now I think maybe has something to do with how many of the item you've crafted.

 

Edit: Assuming P_0 is a function of crafting skill(x) and item rating (y) and potential unknown variable (z), then we can calculate the probability based on brute force. So assuming P_0 = f[x,y,z], then essentially P=1-(f[x,y,z])^n

 

If enough people pick one item and report these following variables:

 

How many times did you have to attempt before you got schematic?(n)

What was your current crafting level?(x)

What was the crafted item rating?(y)

 

Then we could potentially calculate the distribution (P) from the number of answers, and use that to back calculate the dependent function f[x,y,z], because then the equation would only have a single unknown variable.

 

My math is rusty, this right?

 

My point was merely that a 1 in 10 chance doesn't necessarily mean you will get it in 10 attempts. But yeah that's an excellent idea! If you had enough data that would most likely be calculable tough. I'd use a multiple regression analysis tough to find an empirical formula for P_0.

 

Start by trying:

a_0 + a_1 * x + a_2 * y = 1/n

 

(where a_i are constants)

 

and then try adding x^2, y^2 and 1/x, 1/y and some cross products and see if you get a viable model with a low deviation from the statistical data.

 

I'll try this on my alt when I'm leveling up his Cybertech.

Link to comment
Share on other sites

So to put together an experiment someone tell me if this sounds viable:

 

I pick a low level green item, and RE it all the way through all 17 of the potential family tree iterations, recording my armormech level and each of the item ratings. This should give me 3 data points for the level 2 RE, and 14 data points for the level 3 RE.

 

Since a sample set of 30 is statistically significant, if I finish out two green item family trees of the same rating, that will give me 28 data points for the level 3 RE (should be able to figure regression per your equations, gnar, but think need a z variable as catchall for any other variables that might be in there?). Will only have 6 data points for the level 2 RE, could see what that looks like, or maybe have to run a few more items.

 

This is not too much work, I know several people that have finished out item families for some low level items, just finishing out two of them should give us something to go on. May only be valid for a particular crew skill, but I'd be willing to bet they're similar.

 

Edit: number error

Edited by Arch-Stanton
math error
Link to comment
Share on other sites

My apologies if this has already been answered, but I went through about 20 pages without finding my question.

 

If I read the opening post correctly, it is possible to start with a Blue base item and RE into a T2 Purple item?

 

In my particular case, I am trying to go from Enigma Handwraps/Bracers to Veracity Enigma Handwraps/Bracers.

 

I've successfully unlocked the T1's Redoubt/Critical/Overkill, but even after RE'ing about 25 Redoubt Bracers and 10 Redoubt Handwraps, I still haven't learned a single T2 schematic.

 

So before I burn through any more mats, I wanted to confirm if it its even possible to get to a T2 from a Blue base item.

 

Thank you for any help!

 

Went through a nother 30 Redoubt Enigma Bracers this evening with no T2's being unlocked. I'm going to bow out of the RE game until I see some picture proof of a Blue Turning into a T2 Purple.

Link to comment
Share on other sites

Went through a nother 30 Redoubt Enigma Bracers this evening with no T2's being unlocked. I'm going to bow out of the RE game until I see some picture proof of a Blue Turning into a T2 Purple.

 

What is your crew skill level? Every time you RE an item you increase its rating. If you are very low level, you might want to work your armormech level up a little by manufacturing some items and try again later. Not sure what effect crew skill has on RE success, but it couldn't hurt.

Link to comment
Share on other sites

Went through a nother 30 Redoubt Enigma Bracers this evening with no T2's being unlocked. I'm going to bow out of the RE game until I see some picture proof of a Blue Turning into a T2 Purple.

 

http://mrflippy.net/images/screenshots/swtor/T2Purple.jpg

 

I started with Sensory Reaction D-Module (blue), RE'd to Redoubt Sensory Reaction D-Module (purple T1), then RE'd to General's Sensory Reaction D-Module [superior] (Purple T2)

Link to comment
Share on other sites

question:

do the T1/T2 RE rules apply to all schematics across all crafting skills?

 

like, could you reverse engineer a purple quality lightsaber crystal if it didn't have [superior] attached to it?

 

I believe Item Modifications follow the Linear path which is tied up in the Quality of the item and NOT any prefix, so Artifact is the most powerful you can get.

 

(Could be wrong, hope I'm not)

Link to comment
Share on other sites

I believe Item Modifications follow the Linear path which is tied up in the Quality of the item and NOT any prefix, so Artifact is the most powerful you can get.

 

(Could be wrong, hope I'm not)

 

Pretty sure you are right. I think the biggest hurdle for some people is to differentiate the two RE models, when they get excited seeing the huge explanation for the Complex model.

 

Hmm... Started a new alt to have a change of pace, maybe I should keep a detailed crafting and RE'ing log of Biochem Although it'd mean not queuing crafting on the companion(s) so that I can get accurate Affection and Skill for every craft and RE.

Link to comment
Share on other sites

Great update Slaign, much easier to read. Suddenly after reading through all these posts it has become a lot more complicated than I previously thought!!

 

You have done an awesome job of answering peoples questions too ^^. Fantastic work :)

Link to comment
Share on other sites

My point was merely that a 1 in 10 chance doesn't necessarily mean you will get it in 10 attempts. But yeah that's an excellent idea! If you had enough data that would most likely be calculable tough. I'd use a multiple regression analysis tough to find an empirical formula for P_0.

 

Start by trying:

a_0 + a_1 * x + a_2 * y = 1/n

 

(where a_i are constants)

 

and then try adding x^2, y^2 and 1/x, 1/y and some cross products and see if you get a viable model with a low deviation from the statistical data.

 

I'll try this on my alt when I'm leveling up his Cybertech.

 

The problem is affection of compagnon does affect crafting, so it is another variable. Then rules might change depending on item levels or level range (ie level 50 items might have lower probabilities due to difficulty but might also have lower probabilities due to being level 50).

 

The problem is to get enough data we would need many people registering their data with all th epossible variable noted. There could also be bizarre bonus (crafting or RE in cantina compared to while questing and so on).

 

 

 

Then there is the comment of the compagnon crafter (mentinonned in a post not so long ago), I tried it yesterday, and got a schematic from the commented on item, but also got a schematic from the previously crafted item so I am not sure if his comment is spot on or just lucky coincidence.

Link to comment
Share on other sites

The problem is affection of compagnon does affect crafting, so it is another variable. Then rules might change depending on item levels or level range (ie level 50 items might have lower probabilities due to difficulty but might also have lower probabilities due to being level 50).

 

The problem is to get enough data we would need many people registering their data with all th epossible variable noted. There could also be bizarre bonus (crafting or RE in cantina compared to while questing and so on).

 

 

 

Then there is the comment of the compagnon crafter (mentinonned in a post not so long ago), I tried it yesterday, and got a schematic from the commented on item, but also got a schematic from the previously crafted item so I am not sure if his comment is spot on or just lucky coincidence.

 

The nice thing about RE is the companions don't do it, your main has to do it him/herself. AFAIK companions don't have anything to do with RE, just with crafting. Right now we're just looking at the RE aspect.

 

You're right on multiple unseen variables. However, mathematically I think we can reduce them to a single unknown variable to vastly simplify. I THINK if you use a 'z' variable for a catch-all for all the many other unseen/unknown variables that may affect RE success, then we should still be able to see a trend, unless 'z' is much more influential than 'x' (your crafting level) and y (crafted item rating).

 

Recap:

 

X - Your crafting level

Y - Crafted item rating

Z - All other unknowns

Link to comment
Share on other sites

I can confirm that lvl 49 Blues from armormech Can in fact be Re'd into T2 Artifacts. I got mine just this morning.

 

Starting Item Dreadnaught Gauntlets (118)

T1 Critical Dreadnaught Gauntlets (126) + armor + crit

T2 Leadership Dreadnaught Gauntlets (superior) (126) + Presence

 

My T1 to T2 was off of a piece that was modable. So it was an (advanced) modslot item that Re'd to get this new pattern... I had previously RE'd about 20 or so of this exact same item. I had also RE'd many other purples with no success. I'm thinking the Mod slot might have somoething to do with success chance? I will have to test this further when i have money for mats :)

 

Also, This was confirmed by someone with a lower lvl blue item in this forum as well. I am late for work but felt everyone should see this post ill try to edit tonight with a screenshot..

 

SUCCESS!!!!!

Link to comment
Share on other sites

Should I be REing purple items that didn't have a green tier?

 

For example, there is no [Yellow Critical Crystal] — you learn [Yellow Critical Crystal] from the crew skill trainer. REing those learns [Advanced Yellow Critical Crystal]. Should I be REing those?

Link to comment
Share on other sites

I can confirm that lvl 49 Blues from armormech Can in fact be Re'd into T2 Artifacts. I got mine just this morning.

 

Starting Item Dreadnaught Gauntlets (118)

T1 Critical Dreadnaught Gauntlets (126) + armor + crit

T2 Leadership Dreadnaught Gauntlets (superior) (126) + Presence

 

My T1 to T2 was off of a piece that was modable. So it was an (advanced) modslot item that Re'd to get this new pattern... I had previously RE'd about 20 or so of this exact same item. I had also RE'd many other purples with no success. I'm thinking the Mod slot might have somoething to do with success chance? I will have to test this further when i have money for mats :)

 

Also, This was confirmed by someone with a lower lvl blue item in this forum as well. I am late for work but felt everyone should see this post ill try to edit tonight with a screenshot..

 

SUCCESS!!!!!

 

Thats good to hear. Will remove my theory so people dont get confused reading it.

 

Would be good to see some screenies, so can see the value increase amounts... now all you have to do is get the Expert version, crit craft them and you have better Tier 3 pve gloves!

 

And I seriously hope the RE coming from a crit craft had nothing to do with it.... otherwise it makes RE patterns so much harder.

Link to comment
Share on other sites

Should I be REing purple items that didn't have a green tier?

 

For example, there is no [Yellow Critical Crystal] — you learn [Yellow Critical Crystal] from the crew skill trainer. REing those learns [Advanced Yellow Critical Crystal]. Should I be REing those?

 

I had a theory that you couldn't for the high end ones, but EndersRage discover disproves that.

 

So yes you should be able to RE the purple, though incredibly costly!

Link to comment
Share on other sites

Great guide Slagin,

 

I have a theory I would like to discuss, as I believe the rules governing reverse engineering may be slightly different to current thought. (Note this comes from armortech experience)

 

The current theory with reverse engineering states that

  • All base premium(green) / prototype(blue) schematics can be Reverse Engineered twice
  • Base Artifact(purple) schematics cannot be reverse engineered.

 

Looking through the forums, I have seen evidence of mid game prototype(blue) schematics being reverse engineered to T2. However I have seen no evidence of an end game base Protoype(blue) schematic (rating 118) being reverse engineered into T2. In fact I have seen the opposite with people complaining they have unsuccessfully been able to create such T2 items. If anyone has any evidence of this, please share. This could simply be a time/cost issue, however I believe that it is in fact not possible to create such item (which would break the current theory rules), and in fact the rule governing whether an item can be reverse engineered is not based on the quality of the base schematic but instead depends on the item rating, and is simply:-

 

  • All schematics can be reverse engineered twice up to a rating of 126

 

This would explain why artifact schematics cannot be reverse engineered at all - these items are already >=126 rating

It would also explain why mid game prototype schematics can be reverse engineered into T2, but there is no evidence of an end game prototype schematics being reverse engineered into T2, as the T1 version is already rating 126

 

I also think it makes common sense. I have looked through a handful of dps end game prototype(blue) schematics (118 rating) and calculated what their T2 stats would be if they existed. Whilst they lack endurance, their actual dps stats are roughly equivalent to drops from nightmare operations. Crit Craft versions of them would surpass nightmare ops drops (but do lack set bonuses). At the moment it is fine as they are ridiculously costly to make, but imagine the effect in a month's time when they were in production. A person could hit endgame, farm credits from dailies for a couple of weeks and buy a set of gear equal/better to those obtainable in nightmare ops. It would completely nullify the operations end game. Which is why I think these don't exist.

 

Appreciate your thoughts.

 

Oh, and if anyone has evidence of a base prototype(blue) schematic of rating 118 being reverse engineered into a T2 please share as it will contradict this theory.

 

 

 

 

I would like to know this as well. The level 49s can RE to superior purple but what about the prototype blue 50? I have not seen any purple crafted lvl 50. Only the purple prototype that drop from artifact 50 lock boxes. Another thing I noticed: the prototype artifact lvl 50 items have lower armor ratings than the superior artifact 49 RE recipes you get from bought 49 recipes. I havent seen any of the Raid schematics yet. Maybe the only way to get lvl 50 artifact items (besides the BoP rataka from vendor) is from raid drops?

Link to comment
Share on other sites

I would like to know this as well. The level 49s can RE to superior purple but what about the prototype blue 50? I have not seen any purple crafted lvl 50. Only the purple prototype that drop from artifact 50 lock boxes. Another thing I noticed: the prototype artifact lvl 50 items have lower armor ratings than the superior artifact 49 RE recipes you get from bought 49 recipes. I havent seen any of the Raid schematics yet. Maybe the only way to get lvl 50 artifact items (besides the BoP rataka from vendor) is from raid drops?

 

I can confirm REing lvl 50 blues (Primeval XXX, rating 116, recipe gotten of GTN) into artifact T1 items (rating 124) and those into T2 artifacts (rating 124). atm I CBA with SSs, but I can provide them if needed.

Link to comment
Share on other sites

I can confirm REing lvl 50 blues (Primeval XXX, rating 116, recipe gotten of GTN) into artifact T1 items (rating 124) and those into T2 artifacts (rating 124). atm I CBA with SSs, but I can provide them if needed.

 

Thanks, SS not needed. Just wanted to be sure I wasnt wasting my time.

Link to comment
Share on other sites

I can confirm that lvl 49 Blues from armormech Can in fact be Re'd into T2 Artifacts. I got mine just this morning.

 

Starting Item Dreadnaught Gauntlets (118)

T1 Critical Dreadnaught Gauntlets (126) + armor + crit

T2 Leadership Dreadnaught Gauntlets (superior) (126) + Presence

 

My T1 to T2 was off of a piece that was modable. So it was an (advanced) modslot item that Re'd to get this new pattern... I had previously RE'd about 20 or so of this exact same item. I had also RE'd many other purples with no success. I'm thinking the Mod slot might have somoething to do with success chance? I will have to test this further when i have money for mats :)

 

Also, This was confirmed by someone with a lower lvl blue item in this forum as well. I am late for work but felt everyone should see this post ill try to edit tonight with a screenshot..

 

SUCCESS!!!!!

 

This is fantastic news. I was getting discouraged RE'ing the level 49 blue Enforcer Boots for tanking after no discoveries despite ~10 RE's, but I'll keep at it then! Thanks for the report EndersRage!

Link to comment
Share on other sites

http://mrflippy.net/images/screenshots/swtor/T2Purple.jpg

 

I started with Sensory Reaction D-Module (blue), RE'd to Redoubt Sensory Reaction D-Module (purple T1), then RE'd to General's Sensory Reaction D-Module [superior] (Purple T2)

 

Thank you Mrflippy, do you recall if yours was off a moddable? Although I dunno if those can be moddable. . .

 

I'll start RE'ing my critical successes as well then.

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...