We correlate the annual Wolf numbers W and their time derivatives Wʹ by shifting time fragments of W and Wʹ relative to each other. The most significant (up to 0.874) correlation is with 3 years shifts for fragments covering 14 years. For longer and shorter periods, the correlation coefficients 0.771–0.855 with 2–3 years shift. The most significant 9 years shift corresponds to -0.852/-0.824 anti-correlation coefficient for 14/11 years period. The other periods are less significant. To evaluate predictive estimates, we use the times series fragments of W shifted back into the past. A forecast can be made using the leading graphs based upon the derived calibration factor. Test calculations show that the most effective is the calibration factor calculated for changing the phase of the cycle. The best linear pairwise correlation coefficient of the approximation is 0.94.