Goldsmithing skill: 106-110 goldsmithing skill with stall (5), shades (1), and, for new data, apron (1) and torque (2) for 112-119 skill total
Additional gear: Craftmaster's Ring and Artificer's Ring
Recipe: Avarite + Wind Crystal (Level 81, so Tier 2)
Area: Bastok Markets, no Signet, Sigil, Sanction, or Ionis
Rusk:
NQ 7386
HQ1 1983
HQ2 488
HQ3 143
No Rusk:
NQ 7466
HQ1 1920
HQ2 447
HQ3 167
The BGwiki description of how crafting works implies there are 3 independent probabilities:
1) Break or Success
I could potentially reanalyze some of my data for break rates if people particularly care, but at the moment I am ignoring the first comparison because that's not supposed to be related to what Rusks do.
2) NQ or HQ (varies based on HQ Tier, would be 1/4 for this synth)
Rusk:
26.14% (2614/10000)
No Rusk:
25.34% (2534/10000)
Chi squared p=0.196
The raw HQ rate is not significantly different between these samples.
3) HQ1 (12/16), HQ2 (3/16), or HQ3 (1/16)
Chi squared p=0.180
The HQ distributions are not significantly different between Rusk and No Rusk either. If I include NQs in the stats and assume steps 2 and 3 are not independent, the difference in the distributions becomes (p=0.164).