[chris_a]

For a binomial distribution (sums of yes/no events) the standard deviation of the event after N hands is:

sqrt(p * (1-p) * N)

So if your true VPIP is 17%, then after 1000 hands your standard deviation in number of hands that you voluntarily put money in is:

sqrt( 0.17 * 0.83 *1000 ) = 11.9 hands. Divide that by 1000 and you get that your corresponding deviation (in %) is 1.19%. So after 1000 hands, you are within 1 standard deviation if you are between 17-1.19 = 15.81% and 17 + 1.19 = 18.19%

2 standard deviations would be between 14.69% and 19.38%.
3 would be between 13.43% and 20.57%

So I'd say that what you saw wasn't that out of the ordinary.