in my research I'm investigating whether elbow angular velocity (AV) and wrist AV predict success of a jump shot in basketball.

The linear regression analysis demonstrates a fairly high R and the p value in ANOVA is significant. However, doing a simple correlation between elbow AV and wrist AV, these are not significantly correlated.

is this possible? does a poor correlation between elbow and wrist compromise the prediction to success?

Thank you in advance.