Confidence intervals are so much better than p-values but at their heart lies serious problems that helped me become Bayesian:
I mean the fallacy of placing confidence in a confidence interval. See also Robust misinterpretation of confidence intervals.