crafting success

[keoni]

A few anecdotal observations...

In my experience, how empty or full your bag is has nada to do with your success percentage in crafting.

To the central issue of this thread - I do think something changed with either the crafting algorithm, or the underlying random number generator. It has always seemed unduly "streaky" in how often you end up with a run of failures and then string of perfect successes when overcrafting. It has seemed, if anything, somewhat more streaky since patch 71.

However... overall, at least for my experience in overcrafting jewel sets - the ratio of degrades / total failures / perfect successes seems to be comparable to what it was before that patch, if you average them out over time. Of course - this can still lead to some considerable frustration, as Yoshi & Booster have observed, if you get one of those nasty bad luck streaks when you're working with supreme mats.

[riveit]

even crafting with 95% success - got 3 degrades in row and ~20 "Craft action failed"....

I've gotten the impression that stopping crafting after one failure, closing, doing something else and then reopening the craft session improves my success rate and eliminates the 3 degrade or 3 failures in a row problem. I haven't kept statistics on the tactic so I wouldn't swear by it. But perhaps it is worth trying.

[hans1976]

I've gotten the impression that stopping crafting after one failure, closing, doing something else and then reopening the craft session improves my success rate and eliminates the 3 degrade or 3 failures in a row problem. I haven't kept statistics on the tactic so I wouldn't swear by it. But perhaps it is worth trying.

After some sp mining in jewels for the Matis heavy armor plans, I found that the successrate is actually working rather well at any %.

Roll a d100 a 100 times, undoubtly you get double numbers. But as each roll is independant of the next one, the chance of getting all number 1~100 once is equal to getting 42 a 100 times.... And ofc, taking 20 is nice, but untrue at any time.

[iphdrunk]

But as each roll is independant of the next one, the chance of getting all number 1~100 once is equal to getting 42 a 100 times.... And ofc, taking 20 is nice, but untrue at any time.

I submitted this question for next q&a round. We are assuming independent random variables and of course a perfect random number generator. Is this true? I wonder if there are other variables taken into account as the tool HP, grade and quality of the used mats, etc.

[borg9]

I submitted this question for next q&a round. We are assuming independent random variables and of course a perfect random number generator. Is this true? I wonder if there are other variables taken into account as the tool HP, grade and quality of the used mats, etc.

Just a random thought - haven't gone into the full detail of the thread.

A fail AFAIK is when it says you have failed to make the item. This is what the % on the crafting plan refers too. Degrades are a 'different' kind of fail, you have succeded in making an item (therefore not a fail) but it isn't the QL you intended.

Would be nice to see both:

Success chance
and
Degrade chance

listed on the plan, as I beleive them to be different things!

[sehracii]

Degrade chance is the more important stat, could replace fail chance on the info screen and I would be sooo happy.

Failures mean a few wasted seconds, or maybe minutes if you're crafting a lot.

But degrades are wasted mats, and can be very valuable mats, as we all know. Knowing your chance of wasting them is what will determine if it's worth trying or not.

Although I imagine the stat shown on screen is some sort of combination of both, although I don't think it's as simple as success chance = perfect undegraded item chance.

And I would love to know if there are controllable factors that affect your chances...

[aelvana]

I think fails include partial successes. I overcraft skills I don't use for SP, and with the minimum 5% success rate, about 5% are successes, about 1/3 are partials, and about 2/3 are total failures.

[keoni]

This is my impression of how things roughly work with overcrafting... Much of this will be common knowledge for most crafters, but I'll put it here for context, and because I have a tendency to state the obvious

Your % success drops the more levels above your skill that you're attempting to craft, until you reach 5% at 30 levels or so overcraft. Your xp per mat used in the pattern also goes up the higher you were overcrafting, up to a maximum of 300xp per mat consumed, for a total success. You hit that 300xp max at 50 levels overcraft. You can overcraft much higher than that, but your max xp remains the same, and the 5% success probability remains the same.

From my observations, the more extreme you're overcrafting, the more likely that your 95% fail will be a total failure. This probability seems to continue to shift, even beyond that 50 levels of overcraft threshold. The down side of this, of course, is that as you narrow the overcrafting gap, you start to get more and more partial successes - and depending on what you're crafting those might be useless to you. So even though the ratio of perfect successes to failures (which include both partial successes and total failures) may remain 5%/95%, the ratio of perfect successes to partial successes decreases as your crafting gap narrows. Net result: you can end up chewing through a lot more mats to get a perfect success at (for example) 40 levels of overcraft, than you would at 120 levels of overcraft.

[mmatto]

There is a frighteningly easy way to "fix" this peculiarity. Compelete failure -> 0xp and mats lost

[iphdrunk]

After having tried more than 70 times (including total failures) and wasting a few mats -- aiming at q240 boots with 10% of success -- I've tried a few more times with and without boosts. It is only a partial observation and conclusion, but I seem to be more successful when I craft *without* boosts. Maybe boosts are reducing the success rate somehow, although it is not displayed?