I've pulled the Dortmund Table 5, 180 day, data-set (37,231 trials) and dropped the outcomes into a spreadsheet.
The frequency of three numbers present at greater than +/- three standard deviations from the EV, which is worrying; one out of thirty seven of them possibly on this size of a sample, but three . . . . .
Could, of course mean the wheel was wonky but I would think this unlikely with a wheel maintained by a reputable casino. Do you know the source of this data, ie can it be traced back to local site records?