Skill ups are more likely from success? True or False!!

[Silnyil]

I have see many time proples say skill ups are more likely from success.

Answer> False

Test> Combine Tumpy's tonic form Brewing skill next to zero no success skill pumping up. When close to trivial alot of success.

*note*
( When close to trivial alot of success)
When you are doing power skill up success is not the one that will bring your skill up. Not too much plat to put in but 0-135 around 2 hour.

[Kronepsis]

http://mboards.eqtraders.com/phpBB2/viewtopic.php?t=5388

That proves absolutely nothing. You're starting a recipe from when it has a 5% chance of success and skilling up on that all the way to a >100 trivial--of COURSE it's going to look like it's false.

[Hibashira]

I agree with Kron.

The only true test would be to test a skill up point across many many many toons under identical conditions. Say:

1) Level 1 Halfling
2) Log skill up for, say, brewing (its cheap) to whatever (122 sounds good, fetid essence is easy)
3) Repeat 50 times (as in 50 halflings, same stat, same class, gearless)

Then you can compare skill up from 0-1 across the 50 toons, then 1-2 across 50 toons, 2-3 and so on.

That would be a valid test.

[kiztent]

I agree with Kron.

The only true test would be to test a skill up point across many many many toons under identical conditions. Say:

1) Level 1 Halfling
2) Log skill up for, say, brewing (its cheap) to whatever (122 sounds good, fetid essence is easy)
3) Repeat 50 times (as in 50 halflings, same stat, same class, gearless)

Then you can compare skill up from 0-1 across the 50 toons, then 1-2 across 50 toons, 2-3 and so on.

That would be a valid test.

I've only done 3. Results posted on page 1 of this thread:

http://mboards.eqtraders.com/phpBB2/viewtopic.php?t=5388&start=0

A skill up was 50 percent more likely to occur on a success over a 360 data point sample. Methodology for testing is enclosed in the thread.

[Wandor]

There are way too many variables to prove this one way or the other. When I decided to do baking way back when, I started making fishrolls from skill 0. I distinctly remember thinking I would have fishrolls for months with the inevitable successes I would get even though trivial was so much higher than my skill. When I got my first success my skill was over 50 (I forget exactly what skill, but I remember thinking 50% o master and not 1 fish roll yet).

Using that example, 100% of my skill increases came on failures.

Along the same lines, when I took JC from 200-240 I used the lowest possible combo that was still trivial with a geerlok equiped. When it was all said and done I had done a little over 600 combines in 2 days. Of those 40 skill increases, not a one came on a failure.

Using this set of data, 100% of my skill increases came on successful combines.

Basically what I am saying is that the chance to skill up is completely independent from the chance to successfully complete a combine. It is highyl unlikely that VI coded tradeskills in such a way for the skill up check to be somehow variable on the success check. It would be unnecessarily complicated and would serve no purpose other than to increase the amount of code require on an already complex trade skill system.

[Delfontes]

Skill ups are more likely from success? True or False!!

True.

Next topic?

[Calabar]

Whether or not its true doesn’t really mater to me. When appropriate I try to make items close to my skill so that I’ll have a better chance to have something to sell back to the vendor, or possibly even to a player. If I’m making Hero Brew, which is very cheap, it’s not an issue, but on high plat items this just makes sense to me. I’ve never had much in the way of money, so the more success I get the less money it costs me to skill up. If I happen to get more skill ups that way all the better. If it has no effect on skill ups I’m still getting a better financial return so that’s fine too.

[bearcaller]

/em Kicking a dead horse....


It has been drilled into our heads so many times:

The ONLY thing that affects success is SKILL LEVEL

The ONLY thing that affects a chance to skill up is the higher of your WIS/INT (or DEX for fletching and STR for smithing)

PERIOD, END OF STORY, THAT'S IT!

The only reason that skill ups seem to come more often on successes is because tradeskillers are, for the most part, are a frugal sort and do combines in a progression such that you already have a 50% chance or better of success in the first place.

The skill up doesn't come because you magically "fashioned the items together to create something new". It happens because the RNG decided to be nice to you for one fleeting moment

[muana]

skillups on success

The answer is TRUE. It is true that you will get many skillups on failures when the trivial is far away, but if instead of going for something so far from trivial that you can't possibly expect to have successes then you are creating an anomaly in your basis for the skillups.

The best way to test this is to take 50 toons all the same, and start 25 on a brewing recipe that is trivial at say 35 and check the skillup rate (not success) and then do the other on say Mino brew and see if your skillup rate was faster or slower.

Now you might say well we should be only looking at the successes, the thing here is that it is only as you get closer to the trivial does the skillups get faster.

Higher end items may vary more widely due to the fact that many are not really trivial at 250 so don't do anything that says it has a 250 trivial.

my 2 cents

[Iustus]

This was asked in at a fanfaire and Dan said quite clearly that the chance to skill up was higher on a success.

Someone did some tests which showed twice the number of skill ups on success than on failure.

-I

[PinyonTreedotter]

It has been drilled into our heads so many times:

The ONLY thing that affects success is SKILL LEVEL

The ONLY thing that affects a chance to skill up is the higher of your WIS/INT (or DEX for fletching and STR for smithing)

PERIOD, END OF STORY, THAT'S IT!

There have been many things that have been "drilled into our heads" that have been, quite simply, wrong. The original game designers created the skillup/success system. However, the programmers actually implemented it. There are ways to have a relationship between success and skill increases without consciously creating a relationship. The easiest one to imagine is the missuse of the pseudo random number generater.

The PR people at Verant, and now SOE, don't know the system. I doubt if the current game designers really know, and the original programmers are likely long gone. The game designers may have created the relationship, but not told the PR folks, & etc.

To Wandor, and others concerning the chance of success: There really are not that many variables here. The problem is the limited dataset, and the dynamic nature of skill level and success rate. Using a high skill recipe equalizes the success (nominally 5%), but even so, the chance of skillup changes with skill level (very high when low skill, very low when high skill level).

However, I am NOT saying you have more skillups with success than failure. What I am saying is you have a higher chance of skillup with sucess than failure. Those two statements are not the same, and understanding the difference will help you to understand the rest of the posts on this subject.

[Taushar]

It has been drilled into our heads so many times:

The ONLY thing that affects success is SKILL LEVEL

The ONLY thing that affects a chance to skill up is the higher of your WIS/INT (or DEX for fletching and STR for smithing)

PERIOD, END OF STORY, THAT'S IT!

There have been many things that have been "drilled into our heads" that have been, quite simply, wrong. The original game designers created the skillup/success system. However, the programmers actually implemented it. There are ways to have a relationship between success and skill increases without consciously creating a relationship. The easiest one to imagine is the missuse of the pseudo random number generater.

The PR people at Verant, and now SOE, don't know the system. I doubt if the current game designers really know, and the original programmers are likely long gone. The game designers may have created the relationship, but not told the PR folks, & etc.

To Wandor, and others concerning the chance of success: There really are not that many variables here. The problem is the limited dataset, and the dynamic nature of skill level and success rate. Using a high skill recipe equalizes the success (nominally 5%), but even so, the chance of skillup changes with skill level (very high when low skill, very low when high skill level).

However, I am NOT saying you have more skillups with success than failure. What I am saying is you have a higher chance of skillup with sucess than failure. Those two statements are not the same, and understanding the difference will help you to understand the rest of the posts on this subject.

When I asked the development team at a fan faire, these guys are actually the guys who writes the code, they actually had to look at the skill up code since the new PoP AAs allowed stats above 255. Based on that, they said without doubt that only the stats effect the skill up. I believe that they know what they are saying. Also, they have confirmed that the success depends solely on the skill level. Based on these comments, I don't see reason to doubt them.

Taushar

[Delfontes]

Guys, find the thread with stats a few pages back.

I got 249 and 250 on 2 consecutive Failed combines in Brewing. So naturally coupled with a majority of my skill-ups on failures I thought it was just the RNG and it didn't really matter.

The previous thread had some very scientific data. Multiple toons created and PLed in brewing (or baking).

One ignored trivials, started at zero on a 135 (or so) trivial item, 2 toons did this and showed chance for skill-up on success to be twice as high mathmatically. I think somewhere along the lines of they failed on 95% of their combines, but 10% of their skill ups were on successes.

Another toon even was created and made things close to trivial, and took about 30% fewer total combines to get to 135.

It was so proven in that thread that I changed my mind. That isn't easy to do, I am stubborn as heck.

So to beat a dead horse again and again....

You have a higher chance of skill-up on success.

[Canto Wolfheart]

I'm surprised with how many people think this is false.
Read the thread that was given in earlier posts. It contains mathematical evidence that is very strong that this is TRUE.
Anyone that doubts it, show me mathematical evidence to support your claim.

[KyrosKrane]

Here's an interesting thought.

All the trials and experiments that I know of have been performed at low trivial levels, starting at zero or one, and going to the mid-100's at most. Those results are then extrapolated to the high end, or simply assumed to be correct for all trivial levels.

We know this is not correct.

We know that the rate of skillups can vary. At low levels, it's normal to get multiple skillups per stack of combines. At high levels, getting one skillup per stack is considered a good rate; averages well below that are common. Frequent discussions of the "hell levels" between 190 and 210 are also common.

That implies that something in the skillup check formulae changes (or perhaps takes into account the trivial of the item) when doing high-trivial combines. If the rate of skillups does indeed vary, why could this not apply to whether skillups are more likely on success or failure also?

I'm not disputing the evidence in the previous thread; I agree it's convincing -- as far as it goes. And until proven otherwise, I will also believe that the chance of a skillup is greater on a success.

My point is merely that the evidence we have doesn't take into account the possibility that the skillup check differs based on the trivial of the item. Hence, we have no reason to assume that what holds true at low trivials also holds true at high trivials.

Does anyone know of any work or experiments that disprove this? It would make life so much simpler if all skillup checks are based on one formula that is not dependent on trivials.