is success rate formula still valid for high end combines?

[chrisian]

could be bad luck with RNG, but i tried 13 aug cuttings (351 trivials) at 303 modified skill (no mastery AA), and had 6 failures. success rate formula estimates it at about 91% success.
i know its a small sample but 54% success when i'm expecting 91% success is pretty bad. anyone else have any data?

[KyrosKrane]

I've cut about five augs total, only failed one so far, at or near 300 raw skill.

[ColdHeart]

I think very lousy luck. It is possible and if you did them close to each other you may have hit the RNG's "low" phase (Or high phase, depending on what it needs to succeed). I know there have been streaks of good/bad success rate and I'd guess you hit one.

[Verdandi]

/aside on

I think very lousy luck. It is possible and if you did them close to each other you may have hit the RNG's "low" phase (Or high phase, depending on what it needs to succeed). I know there have been streaks of good/bad success rate and I'd guess you hit one.

For the record, randomness (the R in RNG) is inherently and naturally streaky. This is not a flaw in the RNG, it's an inherent element of any truly random system. It's not a sign that you should stop, as waiting until later won't improve your chances one whit. See Trellium's excellent posts in the Thrugadin? Cursed or RNG Jacking Me Around? thread for a better explanation.

/aside off

[Bobaten]

Well, to add a data point, I failed 8 of the last 21 augment cuts I have made with my raw skill between 295 and 297 (0 failures out of 3 combines , 3 out of 6 combines, 2 out of 5 combines, 1 of 1, 1 of 2, 0 of 2, 1 of 2). On the stack of hydroabrasive before that I only had 1 fail of 20 though. I have been skilling up on a mixture of palladium and radiant cuts. I hope this is just the RNG, but I have never had that many failures on something the calculator tells me I should be at 95% success. It almost seems to me that my success rate has gone down as I got more skill, so it could be a bug, but I continue to cut augs in the hope that it is not (and because they sell well :-) ).

[Nolrog]

It almost seems to me that my success rate has gone down as I got more skill, so it could be a bug, but I continue to cut augs in the hope that it is not (and because they sell well :-) ).

I dunno. Could be a bug I guess. Maybe something in the forumla doesn't like the higher skill caps. I'll pass on and and see if maybe someone could have a look see.

EDIT TO ADD: Everyone that's providing data, please provide not only your current skill, but also any mastery AAs (if you haven't already) that you have for that trade as well, so that all the relevant infomation is known.

[Zacatac]

Prior to DoN cap raise, we had enchanter w/max skill and JCM-2 doing cuts.
I hand her 5 augs, she fails 4 out of 5.
Next guildmate hands her 4 augs, she makes 4 out of 4.

Total of 5 success and 4 fails. But completely lopsided for individual results (20% for one player vs. 100% for the other).

[GrimwoodCT]

After hitting 300 (315 modded, JCM3 Salvage3) I have experienced an upswing in my augment successes. I was hovering around 68% success rate when I was skiling from 250-300. I have seen my total success rate begin to climb into the low 70s now. I would estimate that post-300 I am in the 80-85% success range.

Overall Info: Includes Cuts From 250-300
267 Successes
376 Attempts
46 Salvages

71.0% Success Rate
42.2% Salvage Rate

[Bobaten]

I don't have any levels of JCM or salvage currently.

An 85% success rate at 300 skill with jcm is definitely not the success rate you would expect from a modified skill increase of 63 from the old cap of 252. Of course, if we only get "half value" from skill of 250+ (which is not uncommon for stats of 250+), 60 skill would only increase the success chance from ~40% to ~70% (with no mastery). Mastery would bring that success rate up to 85%. That seems worrying to me.

[chrisian]

whatever happened, it isn't adding up.
it might be specific to augs, but something is wrong. PLed a rogue to 300 make poison, went 2 of 5 on aug upgrade combines (listed as 327 trivial) with 315 modified skill. lost 2 black gold inlays and a radiant cut citrine.

asked jeweler friend with JCM3 and 315 modified skill how he was doing on radiant cuts, he said he failed 5 of 41 combines. not a huge variance there, but still, at 315 without JCM you should have 95% success if the old formula applied.

[ColdHeart]

Ngreth, my comment wasn't a complaint, just an observation and possible explanation. I myself have hit a stage of bad luck which ran over 4-5 batches of making Kaladim's. Each batch being 60 combines and hitting ~40% success rate with max skill (should have been nearer 50%). Also each batch was done on different days. It sucks but it happens.

I'd also like to see some more possitive data posted, if there is any. I'm always concerned when people post relatively small samples and it is always the people upset because they had lousey luck.

Talking of bad luck, 297 skill rogue failed to add a Mark to an Aug, triv 296? Had MP Mastery 1. Need to hunt down more slugs :-/

[Ngreth Thergn]

?

I have not posted or commented in this threead... why was that adressed to me?

[Celestya]

*giggles* Maybe he's having difficulty telling ogre from big bear....the eating habits should have given it away. Bears have some manners when slicing up their gnomes

[Keyeluke]

This is clearly a small data sample, but I've cut probably 15 gems with unmodified skill between 290 and 300, plus trophy to bring it up to whatever, and have had one failure.

The old success formula =(Skill-0.75*Trivial+51.5)/100 can't be valid any more - plug in skill 315 and trivial 351 and you get > 100%.

[KyrosKrane]

The formula could still be valid -- it's long been the case that you could be over 100% chance of success. However, the formula is capped at 95% success, so if your chance of success is over 95%, it drops down to that cap. The chance to succeed can be increased slightly with the crafting masteries (to 97.5% max), and if a combine is deeply trivial to you, the chance to fail can be reduced all the way to zero (making it effectively no-fail for you).