GM smithing/tailoring success rates w/ Mastery 3, possible bug

[Bumkus]

Show me some logs with low skill/no mastery that manages to succeed GM bp/legs more than 75% of the time reliably in significant numbers and I'll believe there's a problem.

Well I think that is part of the problem, and the purpose of this thread. No one person can accumalate a statistically signifigant dataset, because of the rarity of the components that go into the combines.

I can only do a combine here or there, but I do try to post my results. I certainly hope other people continue to post their numbers, as I find these data collection threads to be very informative.

[Bobaten]

Actually having re-read through the JCM thread, the only comments in there are that the combines are believed to be capped. There are no references to anything from the Dev's confirming that.

While we haven't heard specifically which combines have caps. Tanker has told us that some recipies can have a lower success cap.
Incidentally, 70% was the max success rate for cutting radiants at 252 skill with mastery 3. The other parts of augment making have lower trivials (higher success rates), so their success cap might well be at their pre-DoN max success of ~80%.

[Phantron]

Well I think that is part of the problem, and the purpose of this thread. No one person can accumalate a statistically signifigant dataset, because of the rarity of the components that go into the combines.

I can only do a combine here or there, but I do try to post my results. I certainly hope other people continue to post their numbers, as I find these data collection threads to be very informative.

Even with the limited data available there is no indication that having no mastery or low skill ever makes it more likely to make a GM BP/Leg.

There's a difference between that having high skill/mastery has no effect because of high inherent fail rates versus the fact that it actually doesn't work.

[Leana]

I've been looking at the numbers presented in this post, and it struck me that the success rate does not appear to be significantly different in the upper end tradeskills, though it should be. There begins to be enough evidence to believe that there is a bug in the system.

Given that, and the hypothisys made at the start of this thread, I made a hypothisys of my own. Lets assume that a bug is present in the code related to mastery AAs. Let's further assume that the bug was introduced when the cap for tradeskills was expanded from 250 to 300. If you assume 250+5% skill with mastery 3, the tradeskill calculator generates a 62% probablility of success. All results presented in this thread cluster somewhat near this mark. Some results are higher, some are a bit lower, but all appear to be very close to the 62% mark. In fact, if you accumulate the values for the three sets of data, you get 61.59%. Based on this data, the hypothisys seems valid.

See how this flies.

Leana Soulwarden
Blacksmith
Inevitable Storm
The Seventh Hammer

[Xulan]

Though conjecture is great (this isn't towards any one person, btw), more data please and less "feeling." If you do wish to make a statement that you have or have not seen an indication, please post the data which will support the statement. Any data supporting any view is welcome.

[Xulan]

I've been looking at the numbers presented in this post, and it struck me that the success rate does not appear to be significantly different in the upper end tradeskills, though it should be. There begins to be enough evidence to believe that there is a bug in the system.

Given that, and the hypothisys made at the start of this thread, I made a hypothisys of my own. Lets assume that a bug is present in the code related to mastery AAs. Let's further assume that the bug was introduced when the cap for tradeskills was expanded from 250 to 300. If you assume 250+5% skill with mastery 3, the tradeskill calculator generates a 62% probablility of success. All results presented in this thread cluster somewhat near this mark. Some results are higher, some are a bit lower, but all appear to be very close to the 62% mark. In fact, if you accumulate the values for the three sets of data, you get 61.59%. Based on this data, the hypothisys seems valid.

See how this flies.

Leana Soulwarden
Blacksmith
Inevitable Storm
The Seventh Hammer

Cool observation =) Good catch.

So what would be interesting to see is if there exist any runs of success over ~60%, for any skill level and any mastery aa level.

EDIT: Here's a thought. It's a bummer if a success rate is capped in such a way that the cap is evident at 250 skill with no mastery aa's, meaning any higher skill or mastery aa will make little to no difference. I think that caps should be made with mastery 3 and max skill in mind. People have mentioned that JC augments seem to reflect this =/

Maybe what is needed is a rehash on how success caps are distributed over a skill level and mastery aa range, so that one will see a return on skill to 300 and mastery aa's to 3. This would only need to be done on:

OoW augments (JC, Smithing, Alchemy & Poison)
DoN GM armor (Smithing and Tailoring)

I believe (anything that has a success cap on it from OoW expansion forward). This could perhaps become a top 10 item.

Oh! and I am not saying that success needs to be higher. I'm saying to start with the cap at the top (300 skill, mastery 3) and make the success rate go down from there as you lower in skill and/or mastery aa.

[Mannwin Woobie]

Oh! and I am not saying that success needs to be higher. I'm saying to start with the cap at the top (300 skill, mastery 3) and make the success rate go down from there as you lower in skill and/or mastery aa.

Now THAT I can agree with. If you are going to cap it, it should be capped when you are maxxed out on skill and mastery so those that have bothered to raise their skill and/or purchase the AA's, can actually get the benefit that they worked hard to get!

[Bumkus]

Oh! and I am not saying that success needs to be higher. I'm saying to start with the cap at the top (300 skill, mastery 3) and make the success rate go down from there as you lower in skill and/or mastery aa.

lol, Xulan, this is a further derail from your thread, but I just had to comment. Level 70's already control the supply of DoN GM armor combines, since they only drop in high level instances. (Competetive advantage #1.) Not all races have equal skillup paths. (Competetive advantage #2.) High levels get accelerated skillups due to higher stats. (Competetive advantage #3) These provide tradeskilling advantages simply based on what you are, not what you do. On top of that, you are advocating a reduction of success rate, using, say...your current skill level as the starting point. I understand that you want competetive advantages, but man, why not just push for required skill level of 299 on all combines?

Is it conceivable that the success caps are intentional? The flood of DoN armors is devaluing what used to be pretty good looted items. I believe this is the more signifigant rational behind success caps on DoN GM items. (Personally I think DoN armors should have been NO TRADE from the beginning, but that is just one ogre's opinion.)

All that said, I would still like to see more people post numbers for their combines, so we can get a feel where the current caps are.

[Xulan]

lol, Xulan, this is a further derail from your thread, but I just had to comment. Level 70's already control the supply of DoN GM armor combines, since they only drop in high level instances. (Competetive advantage #1.) Not all races have equal skillup paths. (Competetive advantage #2.) High levels get accelerated skillups due to higher stats. (Competetive advantage #3) These provide tradeskilling advantages simply based on what you are, not what you do. On top of that, you are advocating a reduction of success rate, using, say...your current skill level as the starting point. I understand that you want competetive advantages, but man, why not just push for required skill level of 299 on all combines?

Nah, my fault, I derailed my own thread, but it's all good discussion.

In response:

1) I am level 70. That said, I have never looted (won a roll) on a single DoN armor item (MDS, Glossy, Course Silk.). I had to either buy my scales or mostly what I have done is just charge for the combine and let people worry about getting the scales (this has minimized my risk and the method has been successful with over 100 combines now). However, yes, it is still an advantage to have access to the drops from some place I can hunt.

2) I think all of the skill-up paths for smithing aren't all that great at the master level. The only one that could really be used would be the yttrium ore races. I know there is an easy path for a few of the tailoring races, but not the rest. It's how it goes (I'm not one of them).

3) For kicks I put in a 250 triv item with 200 skill in the calculator. The difference between 300 and 200 prime stat was less than 1% to skill up. It does make a difference, but it is small. At 280 skill on a 300 trivial, the difference is less than half a percent to skill up.

It is true, I have 300 smithing, but I only currently have 187 tailoring. I plan to work up to 300 (over a year =) and I would like to make GM tailored armor. I said everything above knowing that this would affect me as well. I say 300 because it is the highest skill level available, not because it is my current skill level in smithing. Otherwise, what is the point of even getting 300 and having mastery 3 if it does nothing over 250 skill? (That said, perhaps future armor combines that are even harder will allow Mastery 3 and 300 skill to have an advantage. If so, I'm ok with the current situation =).

Back to the data collection. Anyone who has over ~60% success on GM armor with a decent bit of attempts, please post =)

[Bobaten]

You should also note that 300 skill with no mod, and no mastery, gives you the same 62% success. Also, my success rate making augs has gone up with skill (never had mastery). At 250 skill pre DoN I got the expected 40% return, and I definitely have better than that at 300 skill.

[Bumkus]

2 tries at GM leg aug, triv 386

1 fail, 1 success
No skillup, but no complaints here.

per calculator, with 256+5%=268 tailoring and mastery 2, i should have a 47.5% chance of succeeding in this combine.

[Xulan]

Good stuff. Things I have learned from this thread:

from Leana:

250 skill + 5% mod + Mastery 3 = 62% success rate

from Bobaten:

300 skill + 0% mod + Mastery 0 = 62% success rate

from Leana:

averaging the 3 data sets at the top = 61.59% (~62%)


So, if anyone has data that is contrary to the 62% success rate at any skill level from 250 or higher (any mastery level or none at all), please post your data.

[Leana]

Actually, you should revise that. You should collect all data you can related to that zone, not just data that might disprove what you think is the case. There is a long thread about combines above 250. Try extrapolating data out of that, providing there is some.

I'll try to update my numbers.


This from Celeste2 on the tradeskill section of the SOE website: Skill 300, mastery 3, gm armor combines 7/26, gm symbol combines 23/25, total 30/50.


Leana Soulwarden
Blacksmith
Inevitable Storm
The Seventh Hammer

[Angelsyn]

Phantron...

The thread in the Smithing Forum gives empirical evidence to suggest that having a skill of 290-300, +5% Mod item and no Mastery gives at least as good (if not better) success rate than 300 skill, +15% Mod item & Mastery 3.

[Phantron]

That doesn't mean having less skill equals to better chance of succeeding, which would be needed for mastery and high skill to be BUGGED.

It is quite possible higher skills do not give additional benefit with the existence of minimum fail rate. This is working intended with how the game works (though perhaps not necessarily with how players think it should work).