skillup formula

[Kaysha Soulsinger]

Using the formula as detailed above, we can reverse-engineer it to find out two issues regarding Tradeskills. This requires a fair sample size of tradeskill runs, but combined it should yield fairly reliable information on the following:

- Which Tradeskill has what difficulty modifier
- Which Tradeskill has a tertiary stat (outside WIS / INT)

Of course, this also requires info on whether a skillup was achieved on a failure or a success.

[scarbrowed]

Originally posted by Tarelna Prexus
People have always talked of hell levels for tradeskills from about 180 to 220 but with this published formula, its looks like its diminishing returns till it bottoms out at 190 and never gets any better.

It does get better in many tradeskills because the chance of success gets higher. For example, tailoring, you're only choices after 190 trivial at 242 or 335. If you are making acrylia reinforced at a trivial of 242 with current skill of 190, you should expect about 30 combines per skillup. Due to increased successes, you should expect 25 combines per skillup at skill 220 on the same recipe.

Baking never had a hell level for me, because at 190, you could make several recipes that trivial at 202, which makes skillups more likely. Staying close to trivial is also why I've not heard much about hell levels in Jewelry Craft.

Finally, don't forget that sometimes the RNG just decides you are going to have a hell level. This is completely random and can happen at any level.

[kiztent]

edit: covered by skillup calculators

[Ursawraith]

This formula shows that in the endgame of tradeskilling you can expect about a 5% change of skill up per combine which works out to about 1 in 20 of course the RNG does still have a mean streak.

[Vivamort-Rodcet Nife]

Which would put the amount of attempts needed to get from 190 to 250 skill, assuming using near max stats and staying near the trivial range, to be about 1200 combines. Everyone say "thank you new trader interface". A 5% chance of increase was indeed what I calculated for

Vivamort

[andyhre]

Okay just to make sure I understand.

I have max wis these days of 260, being capped at 61st level.

I am working on my last two tailoring points before the 200 cap, 198->199 and 199->200.

I am doing a variety of things including Solstice Robes.


Y=3
F=2 when I fail, which is usually (by design).

N=260-15=245

So (N*10)/(Y*F) = 2450/6 = ~409

So when compared to a rand(1000), I have about a 41% chance of passing the test, after which I then would compare a rand(200) with 190 (since my skill exceeds 190), giving me a 5% chance.

So on every failure of a Solstice Robe, at 198 skill and 260 wisdom, I have just about a 2% chance (41% of 5%) of a skill up -- so I can expect to do, on average, 50 failures per skill up.

On the few that I succeed, the formula is then changed b/c F=1:


So (N*10)/(Y*F) = 2450/3 = ~817, and as a result, my chance just about doubles to 4% (actually it's 81.7% of 5%).


And, since I succeed at about a 5% rate, given the disparity between my skill and the trivial, my overall skill-up rate will be darn close to 2%.

Correct?


if so this would explain why it seems to make no difference at this point, my samples of 60-100 tries at a time are now big enough to make any skill up (on success or failure) anything more than noise in a stream of non-skillups.

[andyhre]

Another question.

Nowhere in this formula is there room, I think for the oft-observed fact that skill ups come more frequently on Bows than Arrows in fletching.

Is this a correct observation?
Can someone help find how that can be true if this is how skill ups are determined?

Is there room for the number to vary by the kind of combine within a skill (i.e. be a "2" for bows and a "4" for arrows)?


EDIT: I think I just answered my own question: the arrows are 335 trivial and the bows are 255. Doh!

third edit: deleting my second edit. d'oh

[Yakk]

Correct?

Yes. =)

if so this would explain why it seems to make no difference at this point, my samples of 60-100 tries at a time are now big enough to make any skill up (on success or failure) anything more than noise in a stream of non-skillups.

Yes.

As an aside, lets say you are an int caster, and get to L 65 with max int/wis AAs.

Your skillup chance would go up by 39%. Ie, from 1 skillup in ~ 50 to 1 skillup in ~ 36.

[Curufea]

The Y given here are just guess ? Or are they true facts ?

Otherwise here is a simple test to determine them :

For smithing/fletching (ie : with a secondary stats)
Make a lvl 1 mule with 100 as the max of int/wis/str/dex.

For other skills (ie : without secondary stats)
Make a lvl 1 mule with 115 as max of int/wis

Proceed to the test :

Make it combine an item until the mule skill up to 1.
Note the following :
number of success with skill up
number of success without skill up

Delete mule, remake a mule and do the same thing.

The %age of skill up on a success over all your mules is expected to be 1/Y.

[Kaysha Soulsinger]

Curufea, this is something I also suggested a few posts up (I guess it was missed, though). In addition to determining Y, we should be able to determine if the skill in question uses 2 or 3 stats, for certain. There are still debates on some skills if they use an alternate stat or not.

[Bregalad]

Does it say anywhere weather 2nd check uses geerlok-adjusted skill, or raw skill?
Past 190 raw skill, geerlok has no bearing on 2nd check, but before that...

If 2nd check uses effective skill, it would mean that before 190, geerloks hurt skillup rates, peaking out at 180, where they cut 2nd check chance in half.

[GraemeFaelban]

Originally posted by scarbrowed
It does get better in many tradeskills because the chance of success gets higher. For example, tailoring, you're only choices after 190 trivial at 242 or 335. If you are making acrylia reinforced at a trivial of 242 with current skill of 190, you should expect about 30 combines per skillup. Due to increased successes, you should expect 25 combines per skillup at skill 220 on the same recipe.

Well, for those that only have a geerlock, they can do the Ethereal swatches from 210 to 212. For those that are lucky enough to get Shei's Shears, they can do 200 to 212 on the swatches. (I think the shears are +10% anyway, don't have them yet). Sure, 2 points is not huge, but, it's a few less combines on average. For 12 points it becomes quite worthwhile to do.

[KyrosKrane]

Originally posted by Kaysha Soulsinger
There are still debates on some skills if they use an alternate stat or not.

Tanker told us that only smithing and fletching have alternate stats -- STR and DEX, respectively. No others do.

[Calliopeia]

Tanker told us that only smithing and fletching have alternate stats -- STR and DEX, respectively. No others do.

Just to confirm, that means that DEX does not affect tailoring as we previously thought?

[KyrosKrane]

To my knowledge, it was never confirmed, even from when they first added alternate stats. DEX for tailoring was only ever mentioned as a possibility.

To directly answer the question, no. To the best of my knowledge, tailoring does not have an alternate stat.