Only in the quartile with exposures above 31.8 parts per billion do the odds values exceed 1, yet the figure suggests that at higher levels, probability starts to drop again.
With 10 horses running in each race, it would seem that the odds of any one district running winning would be 1:10, or a 10% probability for the Wave to win.
Anything lower reduces the expected cost of criminality, without doing anything to improve the probability of detection. (Treating whistleblowers leniently is consistent with this logic: letting them off punishment raises the odds of truth-telling, and therefore of detection.) There are plenty of arguments against ultra-high fines, however.