Fan Faire June 2005 Write-up (Plus Tanker handout)

[Xislaben]

One reason is because that would all be integer math, which is not the way it works. mulitplying by 10 early, then dividing by 10.0 (a floating point number) later gives a more accurate representation of the odds. For S = 313, Y = 3, D = 2, it's the difference between (317 / (3*2)) = 52 and ( 317 * 10 ) / (3*2) / 10.0 = 52.8.
Hmm, doing ( S / (float) (Y*D)) would work better (maybe I'll change it), yielding 52.833333

Another reason was I was trying to mimic the old way it did things as much as possible, and that code contained the *10 and rolled a number 1-1000. To represent it as a percentage, I just divided by 10.0.

Heh, that's mainly a "show your work" kind of thing. ( 10/125.0 ) was a way to show WHY I was multiplying by 0.08 (as a method of spreading 10 points of skillup chance over 125 points of skill).

I code in vb mostly and would have set the pass 1 calc equal to a double so it wouldn't have made any difference for me, but I see how it would in C or the like with floats. Are you saying that basically you only need to take the pass 1 chance to 1 digit past the decimal because you're multiplying it by 10 and comparing it to a /random 1000 to see if it passes?

So while your calculated chance of passing is 52.833333%, your actual chance of skillup due to the calc being compared to a random int between 1-1000 is 52.8%? I knew you guys were shaving little points somewhere :P

[Niami DenMother]

My notes from the CL summit and Fan Faire can be found at the below link.

http://www.eqtraders.com/articles/ar...p?article=g340

Once I have more time to unwind, I'll work on a writeup regarding the expansion as well.

[Maevenniia]

My notes from the CL summit and Fan Faire can be found at the below link.

http://www.eqtraders.com/articles/ar...p?article=g340

Once I have more time to unwind, I'll work on a writeup regarding the expansion as well.

Well, I was hoping that Kyros had just forgotten to post what might have been mentioned about Spell Research, but from your notes, also lacking any mention of Spell Research, it looks like it never came up...

So...maybe Tanker or Absor might comment...

-when will Spell Research get new recipes so we can actually skill up to 300 (cap was raised but highest trivial recipe is 227)?
-when will spell research get new things to do, spells 61-70 or maybe wands or one-shot scrolls?
-when will Spell Research get a TROPHY?

[AbsorEQ]

Well, I was hoping that Kyros had just forgotten to post what might have been mentioned about Spell Research, but from your notes, also lacking any mention of Spell Research, it looks like it never came up...

So...maybe Tanker or Absor might comment...

-when will Spell Research get new recipes so we can actually skill up to 300 (cap was raised but highest trivial recipe is 227)?
-when will spell research get new things to do, spells 61-70 or maybe wands or one-shot scrolls?
-when will Spell Research get a TROPHY?

Considering that research was, until resently, a purchasable skill, I'm not impressed with the idea of a trophy. Perhaps once we get the rest straightened out.
No one-shot scrolls or other new items planned at the moment. I do need to talk to Rytan a bit about researching other spells, tho.
For the most part skill gaps will be filled when then second (or third, I forget which) part of the revamp is done. When that will be I still don't know. Sorry.

A

[Aeght]

you have to pass BOTH... so this means if you do not have 100% from pass 1.... it jsut reduces your overall by that percentage. Even if you pass pass 1 you still ahve to pass pass 2...

Ajusted Skill = 214
Success chance = 74.25%
Chance of skill up on Success: 9.38%, on failure: 7.04%, overall 8.78%
(I guessed on a 255 trivial... since you did not give the trivila of what you were working on...)

looks to me like jsut horrible luck. Is about 1 in 15 (not exact math!!!)

Enchanted velium bits, trivial 222. success rate 95%. Note that the skill up rate I am quoting is not over one skill up but several , so a couple hundred combines, large enough to be statisticly significant.

As for JC,

Ajusted Skill = 283
Success chance = 95%
Chance of skill up on Success: 2.75%, on failure: 1.38%, overall 2.68%
(trivial 300)

so about 1 in every 40 combines. 60 is not way off... but still bad luck.

Actually, since it turns out that JC is a learning difficulty 4 skill, it should be 1 in 52 (approx) so it seems to be working at this level (albiet very annoying and expensive...).

Aeght

[Aeght]

Considering that research was, until resently, a purchasable skill, I'm not impressed with the idea of a trophy. Perhaps once we get the rest straightened out.
No one-shot scrolls or other new items planned at the moment. I do need to talk to Rytan a bit about researching other spells, tho.
For the most part skill gaps will be filled when then second (or third, I forget which) part of the revamp is done. When that will be I still don't know. Sorry.

A

I would still like to make a case for 'useless' routes to skill. That is to say, assuming that any given path offers sufficient raw materials supply so as not to turn all but the independantly wealthy off from tradeskilling, paths that offer useful end product to skill up on will likely merely flood the market with those products, making a mess of regular hunting.

Also, effort to obtain the quantities of raw materials needed to skill up should be taken into account when determining trivials, and weighed against utility of end product. Even utterly useless end products should have higher trivials if they require sufficient subcombines and/or have poor input ratios (ie, needing several each of various parts... ).

Aeght

[Xodar]

Great write up! Thank you Kyros

Based on my results in the field, I think that the formulae need to be rechecked...

In the low 200's in smithing (currently 214) I have been averaging 40+ combines per skill up.

Based on the new formula 2, I should be expecting approximately 10 combines per.

That is at 300 int, and assuming a difficulty factor of 2 (Yielding 142%) on formula 1. Even if the difficulty was a 4, formula 1 would still come to 71%, and approx 14 combines per skill up.

In jewelcraft at 283, I am averaging approx 60 combines per skill up, instead of the expected 26... more than double the number of expected combines, unless of course jewelcraft is a 4? I do not see the jewelcraft difficulty listed (.



Aeght

I did 100 shadowcream combines 3 days ago with 365wis/str and salvage3/mastery 2. I got 10 skill ups (from 227 - 237) which pleased me immensely, lol, and seems in line with the formula. We are still dealing with a random factor.....of those 10 skill ups 3 came in the course of 5 combines, and 1 came after 25 (doing combines in lots of 25, had 1 lot without a skill up out of the 4).

[mithridate]

First and foremost: A hearty thanks to Kyros (I still hate you for having done solstice robes first), Niami (especially for the repeat-link to the straightjacket page), Ngreth (you're still going to post your notes, right? /grin), and the SoE tradeskill peeps - Absor, Maddoc, Tanker et al. All of them provided a treasure trove of information.

Could you perhaps combine all the raw info (impressions, transcriptions, quotes, etc.) into one sticky (e.g., Fan Faire 2005) as well as replace the old "skillup formula" sticky with the most current one? For the sake of posterity and clarity.

[Oona Wynterbreeze]

Actually, since it turns out that JC is a learning difficulty 4 skill, it should be 1 in 52 (approx) so it seems to be working at this level (albiet very annoying and expensive...).

Aeght

Hrrrrmmm.... that still puts me waaaaay off with 1 in 85.1 since 280 and 1 in 157.4 since 290. Apparently the RNG Gods hate me.

[KyrosKrane]

Okay, I'm confused.

At the bottom of the document, it says something to the effect of "if you assume a 100% chance of passing check 1, which is unrealistic". However, if you plug the stats for an average tailoring combine with someone with max wis (with all AA):

Success
( (400 * 10) / (2 * 1) ) / 10.0 = (4000 / 2) / 10 = 2000 / 10 = 200% chance

Failure
( (400 * 10) / (2 * 2) ) / 10.0 = (4000 / 4) / 10 = 1000 / 10 = 100% chance

That doesn't seem unrealistic to me.

Aside from the difficulty factor, which Ngreth already pointed out, it's also very rare to find a person with 400 WIS or INT. Only certain classes with a bucketload of AA and high-end gear can get their primary stat that high. Most tradeskillers will have a primary stat in the 200-300 range, which will have a profound impact on the formula.

[KyrosKrane]

Tanker allowed the players present to select one tradeskill at a time, and he would give us the info for that one tradeskill. The ones selected were done so by players, not by Tanker. I asked Tanker later about the remaining difficulties, and he said he didn't want to reveal them all because there are people who enjoy figuring them out for themselves.

So, don't make any assumptions based on which tradeskill difficulties were announced. It's essentially random.

I would also make a distinction between values Tanker gave us (which are most definitely correct as of this writing) and those Suani derived, which are probably correct but are still not definitive. So, I would amend the list as follows:

4 = Alchemy (Tanker)
3 = Baking (Suani)
3 = Brewing (Suani)
4 = Fletching (Tanker) - effective stat not reduced by 15
4 = Jewelcraft (Suani)
2 = Poison Making (Tanker) - effective stat not reduced by 15
4 = Pottery (Suani)
2 = Smithing (Tanker) - effective stat not reduced by 15
2 = Tailoring (Tanker)
2 = Tinkering (Tanker)

[KyrosKrane]

OK, something screwy here with the time zone setting on the server. My reply to Elfdruidess's post is showing up BEFORE she made it. I'm guessing it's because of the server upgrade that's happening around this time, which Ngreth has posted on the home page.

[Elfdruidess]

Tanker filled a lot of holes where we lacked data. Here's a consolidated list of tradeskill difficulty numbers:

4 = Alchemy (Tanker)
3 = Baking (Suani)
3 = Brewing (Suani)
4 = Fletching (Suani) - effective stat not reduced by 15
4 = Jewelcraft (Suani)
2 = Poison Making (Tanker) - effective stat not reduced by 15
4 = Pottery (Suani)
2 = Smithing (Suani) - effective stat not reduced by 15
2 = Tailoring (Suani)
2 = Tinkering (Tanker)

I put a lot of faith in Suani's numbers because he set smithing at 2 when a lot of people wanted to call it as high as 4.

I have always been a big believer of Suani's results as well because she was confident enough to put a 2 on tailoring when the devs were quoted to say that tailoring is a 3.

However, there is now a new situation. With Tanker releasing 6 of the 10 tradeskill difficulty values, it is as if he is implying the ones he did not mention are 3 (i.e. Baking, Brewing, Pottery, JC). From experience, I'm really more inclined to go with Suani's results (i.e. Pottery and JC are 4). But then, if Pottery and JC are 4, why did Tanker not mention it with the other 4s he mentioned?

[Ngreth Thergn]

Ngreth (you're still going to post your notes, right? /grin),

Eventually.

My main computer is down (Should have it up Wednessday) and I am working all week, then only have ONE day off before work again.... will see.

Most is up. I doubt I have much different notes, but I will try and compare and contrast as soon as I can.

[mewkus]

Getting your strength over 400 is not that difficult. Most melee types in their upper 60 with 2 or 3 levels of Chaotic Potential, Wunshi and Feran Avater will be above 400.