This was an interesting read with a good chunk of the messages deleted.
I spun around 520 times during the event, and averaged a gold just under every 7 spins. However, I did not refresh the wheel and always worked from wheels that were already in process, which means in some cases full circles were already closed and in other cases, I was closing circles as I burned through spins. This leads me to believe that the odds fluctuate with the amount of active circles available when you spin. So the odds start at 1/9, but increase in your favor each time a circle closes. This would seem to best explain why those that refreshed continually experienced closer to the 1/9, while more selective spinning resulted in higher success rates. Unfortunately, the confidence interval width even for my 520 spins is still too wide to say for sure vs a straightforward 1/9 chance.